Scent as a Medium for Design: An Experimental Design Inquiry

A Thesis Presented in Partial Fulfillment of the Requirements for The Degree Master of Fine Arts in the Graduate School of The Ohio State University

By Alessandra Cerqueira Mattos, MBA

Graduate Program in Industrial, Interior and Visual Communication Design

The Ohio State University 2011

Master’s Examination Committee Professors Peter Chan, Advisor, PhD Elizabeth B.-N. Sanders, PhD Carolina Gill

Copyright by Alessandra Cerqueira Mattos © 2011 All rights reserved

Abstract

The roles of Designers and Researchers are blending together and each one is trying to work in and improve on the other’s domain. This represents an important shift for both fields, but it is fundamental that the next steps are taken by studying and acquiring more knowledge across both fields.

Thus, this study aims to adapt Experimental Design as a method to be used by Designers. The use of experiments as a method for Design Research is a relatively unexplored domain within the field of Design that has implications for positive and useful applications. With the application of rigorous methods of experimental study, results can be generalized to broader situations. Scientific methods address concern of bias, replication, and disclosure of the problem and findings. Experimentally-based findings will be well respected not only by Designers, but also by the Scientific community.

The experiment consisted of interviews with ninety people who were asked to associate the presented smells with words. Quantitative analysis was done using statistical tests to understand whether people’s responses did or did not occur by

ii

chance. This helped support, scientifically, the findings. Additionally, qualitative analysis was done by creating different types of visualizations of the data. These visualizations helped to reveal patterns and to create unique “fingerprints” for the smells.

The sense of smell was chosen as a medium for the Experimental Design. In Psychology or Neuroscience it is easy to find information about how the sense of smell works. It is a sense that hits us directly through the limbic system, which supports a variety of functions including emotion, behavior and longterm memory. Therefore, our associations with smells are incredibly strong, emotional, and cognitively bonded to memory, which shows opportunities for Design explorations. However, there are not a clear understanding about people’s psychological perceptions and associations to the odors. Understanding this will enable the use of smells in a more conscious way according to people’s perceptions.

iii

Dedicated to my family: My mother, who believes in me more than anyone else. Everything I am, I owe her. Phil, who keeps her happy giving me piece of mind. Bruno, who is and will always be my best friend, love, and life partner.

iv

Acknowledgements

During my six years of professional experience, prior to coming to The Ohio State University, I have worked creating and developing concepts, products, and ideas. However, rarely have I actually had time to research and understand the real problems to develop better solutions early in the process. Being exposed to Design Research methods during the first year of the Design Development Master Program, studying Design Research and User Experience under the direction of Professors Liz Sanders, Heike Goeller, Paul Nini, Carolina Gill, and Brian Stone, made me re-evaluate my personal interests and create a new direction for my thesis ideas, focusing more on research techniques and data analysis rather than practical Design applications.

My committee was essential to feed my interest for Design Research and keep me on track. I would like to thank Dr. Peter Chan, my main advisor, who guided me calmly through all the steps on this journey. He was open to my ideas and helped me polish them to create an interesting and useful project. His patience was essential to keep me going. Thanks to Carolina Gill who happily embraced my topic and always gave me considerable and thoughtful feedback. Last but

v

not least, I would like to thank Dr. Elizabeth Sanders who not only inspired me throughout my time at The Ohio State University, but also involved and immersed herself in this study with me.

vi

Vita

1979

Born – Rio de Janeiro, Brazil

2004

B.S. Industrial and Visual Communication Design,



Escola Superior de Desenho Industrial, Rio de Janeiro,



Brazil

2004-2009

Design and Branding Manager, Uncle K



Rio de Janeiro, Brazil

2007

MBA in Marketing



Fundação Getúlio Vargas, Rio de Janeiro, Brazil

2009-2010

University Fellowship Recipient, The Ohio State University

2010-2011

Graduate Teaching Associate, The Ohio State University

Fields of Study

Major Field: Industrial, Interior and Visual Communication Design

vii

Table of Contents Abstract............................................................................................. ii Acknowledgements............................................................................... v Vita................................................................................................. vii

Chapter 1 Introduction......................................................................................... 1 1.1 Objectives and scope....................................................................................

1

1.2 Background information............................................................................

2

1.3 Introduction to Research Methods............................................................

7

1.4 Roadmap to the study..............................................................................

10

Chapter 2 The sense of smell............................................................................... 13 2.1 Describing and identifying odors..............................................................

16

2.2 Odor classification systems.......................................................................

17

2.2.1 Zwaardemaker smell system, 1895........................................................................... 2.2.2 Henning odor system, 1916....................................................................................... 2.2.3 Crocker-Henderson smell classification, 1927......................................................... 2.2.4 Stereochemical smell theory or the Seven Basic Smells, 1952................................. 2.2.5 Conclusion..................................................................................................................

18 19 20 20 21

2.3 Designing fragrances.................................................................................

21

2.3.1 Traditional fragrance classification......................................................................... 2.3.2 Modern fragrance classification............................................................................... 2.3.3 Fragrance wheel........................................................................................................

23 23 23

2.4 Using smells as a Design element............................................................

viii

24

Chapter 3 Design Research methods exploration...................................................... 28

Chapter 4 Designing and conducting the experiment................................................ 36 4.1 Exploratory Study #1................................................................................

36

4.2 Exploratory Study #2...............................................................................

46

4.3 Final Experimental Design .....................................................................

53

Chapter 5 Experiment Analysis............................................................................. 61 5.1 Quantitative Analysis.................................................................................

61

5.2 Qualitative analysis..................................................................................

73

5.2.1 Word clouds . ............................................................................................................. 5.2.2 Gender Comparison word clouds............................................................................ 5.2.3 Categories’ comparison............................................................................................. 5.2.4 Words’ comparison................................................................................................... 5.2.5 Summary of the smells..............................................................................................

75 80 82 91 93

Chapter 6 Exploring Smell Tool concepts/ideas/prototypes......................................... 97

Chapter 7 Discussions and Conclusion................................................................. 103 7.1 Discussion: exploring and applying the research’s findings.................. 7.1.1 Checking if a chosen smell is consistent with pre-defined concepts....................... 7.1.2 Looking for smells associated to specific concepts................................................. 7.1.3 Researching associations for a new smell..............................................................

ix

103 104 106 108

7.1.4 Probing deeper into patterns...................................................................................

7.2 Conclusion..............................................................................................

109

109

Appendix ......................................................................................... 111 Appendix A. Exploratory Study 1 – transcriptions.......................................................... Appendix B. Final Experiment Answer Sheet................................................................. Appendix C. Responses to the Final Experiment Questionnaire..................................... Appendix D. Grouping Demographic Information – Final Experiment....................... Appendix E. Re-coding to Annotated Words – Final Experiment................................. Appendix F. Calculating significant words’ threshold value.......................................... Appendix G. Selection of Significant Words....................................................................

111 140 141 155 158 169 170

Bibliography..................................................................................... 185

x

List of Tables

Table 1. Experiment demographics: gender........................................................... 61 Table 2. Experiment demographics: age range......................................................62 Table 3. Experiment demographics: department...................................................62 Table 4. Experiment demographics: origin............................................................63 Table 5. Threshold values for significant words: total............................................69 Table 6. Threshold values for significant words: gender........................................69 Table 7. Chi-Square test showing p_values for comparing females and males where orange represents significant difference at 0.01 level, and green represents significant difference at 0.05 level..........................................................................72 Table 8. Percentages of the average of words chosen by a single participant....... 83

xi

List of Figures

Figure 1. Design and the Senses (Shedroff, 1994)....................................................4 Figure 2. Olfactory system (Lynch & Icke, 2007)................................................... 15 Figure 3. Henning’s odor prism (Stevenson et al., 2006)...................................... 19 Figure 4. Fragrance Wheel (Edwards, 2011)...........................................................24 Figure 5. Map of Design Research and Practice (Sanders et al., 2011)..................29 Figure 6. Map of Design Research and Practice modified......................................32 Figure 7. Box with all the smells in numbered bags............................................... 41 Figure 8. Participants’ desk: consent form, questionnaire, tasks order, stickers, cards, categories’ definitions, and coffee................................................................ 41 Figure 9. Investigator’s desk with script, smell and tasks order, and notebook....42 Figure 10. Classification of smells into good or bad...............................................43 Figure 11. Analysis spreadsheet example................................................................52 Figure 12. Experiment answer sheet.......................................................................67 Figure 13. Word clouds: size represents number of times a word was chosen (1).77 Figure 14. Word clouds: size represents number of times a word was chosen (2).78 Figure 15. Gender comparison word clouds...........................................................81 Figure 16. Categories comparison chart visualization #1.......................................85

xii

Figure 17. Categories comparison chart visualization #2.......................................87 Figure 18. Categories comparison chart visualization #3..................................... 90 Figure 19. Words’ comparison chart (detail)..........................................................92 Figure 20. Cinnamon word cloud...........................................................................93 Figure 21. Cinnamon categories’ comparion chart.................................................93 Figure 22. Ginger word cloud..................................................................................94 Figure 23. Ginger categories’ comparion chart......................................................95 Figure 24. 2D Smell Tool concept.......................................................................... 98 Figure 25. 3D Smell Tool concept...........................................................................99 Figure 26. Interactive concept (OECD Better Life Initiative, n.d.)...................... 101 Figure 27. Pomegranate word cloud for females .................................................105

xiii

Chapter 1

Introduction

“The biggest felonies in sensory science have to do with equating the physical stimulus, or early neural signals, with psychological perception (…) Researchers have mostly attempted to determine how a chemical stimulus is represented in the olfactory system, without considering context and learning.” (Dr. Peterzell, 2006)

1.1 Objectives and scope In the broadest sense, this research project examines how Designers can make use of research methods in a more scientific way. The aim is to learn what kinds of perceptions, memories, and associations smells can trigger in people, and to understand if it is possible to use scent as another Design element. This study also highlights the potential for using scientific methods with a User-Centered Design mindset.

Design Researchers, or even Designers interested in research methods or in researching new smells are the people who would find this project and its results

1

relevant and significant. The idea of understanding how smells influence people’s emotions could also inspire Design practitioners. However, the main objective of the research is to explore the use of more scientific methods, specifically in this case, the use of Experimental Design, in Design Research settings. The focus is not on the specific results of each one of the smells that were used as stimuli. Scent is being used as the medium to implement the method.

The use of experiments as a method for Design Research is a relatively unexplored domain within the field of Design that has implications for positive and useful applications. With the application of rigorous methods of experimental study, results can be generalized to broader situations. Additionally, the investigation is being done from an interdisciplinary perspective, an approach that intends to bring different schools of thought and disciplines together as new needs emerge.

1.2 Background information Design discipline is comprised of different areas, and more have been incorporated and created adapting to changes in technology, and demand. The three most common specializations that can be found in most Design schools are Industrial Design and Visual Communication Design. Some schools also include Interior Design under the same umbrella. However, the approach for this study, analysis and discussions, include only Industrial and Visual Communications Design.

2

The Design field has been changing and incorporating stimuli to other senses through the Design for Experiencing approach. However, Visual Communication Design, as the name suggests, still relies heavily on the sense of sight. Communication of content, ideas, and concepts are created by using visual aids such as color, typography, and composition. Additionally, Designers that work with time-based media, such as animation, make use of the auditory sense (hearing) as well. It enhances the communication of specific content, opening possibilities for a new way of experiencing the information. For Industrial Designers, not only the sense of sight is important, the sense of touch might be relevant as it could enhance experience and lead users to act and/or react to products in different ways. For example, in an object with different textures, the areas where users are supposed to touch or hold will be the ones with a more ergonomic grip, a lot of times with dents to accommodate the fingers, and sometimes also with a more rubbery surface for a better grip. These tactile cues might be relevant if someone is using a device without being able to see it.

Amongst Visual Communication and Industrial Designers, most of them address only three of the five senses. The sense of taste and smell have been neglected, mostly because there is not enough knowledge around what Designers could do with them to explore new possibilities.

3

tactile sense visual sense

product touch kinesthetic sense of direction

color

iconography

contrast

photography

typography

illustration

calligraphy branding

graphic design

service design

animation video/film voice

olfactory sense

music sound

gustatory sense (taste)

auditory sense

Figure 1. Design and the Senses (Shedroff, 1994)

It is the growing complexity of experiences that people expect from every product, brand, and space that explains the importance of a research comprising one more sense. The sense of smell is actually the one sense mostly related to people’s emotions. (Droscher, 1969) There are many factors that can influence how we perceive things, and things are defined by how they make one feel.

When it comes to the study of smells, aromatherapy is another field that approaches how smells affect people. It is known as an alternative medicine and

4

its purpose is to change a person’s mind, mood, cognitive function or health. Aromatherapy is considered anecdotal by the scientific community. (Ehrlich, 2009)

Smells are being used today as an important element for branding. Some of the oldest documented examples about Designers or Marketers using smells as another element to identify a brand or product come from 1965. The first example was the development of a distinct smell for the 1965 Rolls-Royce Silver Cloud. They introduced a new fragrance to the new models of Rolls-Royce cars because customers’ expectations for the new released cars were not being fulfilled anymore. Older models still had the natural smell of wood, while the new ones smelled like plastic and foam. The development of the smell for the 1965 RollsRoyce was created by mimicking the smell of the old cars (Lindstrom, 2010).

Another important example was in 1990, when Singapore Airlines decided to innovate and expand their identity to include not only colors, shapes, typeface, and logo. The company also developed a distinct smell that was applied to the perfume of flight attendants as well as blended in the airplanes’ towels and diffused in their spaces. The objective was to provide a more complete experience to their clients, helping them to recall more powerfully the experience of flying with them (Kumar, 2006).

5

Designers and Marketers already understand the potential of using scent in their projects as a branding technique in order to create another way of helping people to recall and identify brands and products. However, Designers are still not exploring other possibilities of communicating ideas and concepts using smells. Creating associations between brands and products with specific smells is already a big step, but how are the smells chosen? Is there a meaning behind the scents or is it just a matter of being a good scent? It does not appear to be based on associations that people have with those smells. The associations are created in people’s minds by Marketers and Designers: the scent is presented and at that moment people start to associate that a certain smell belongs to a specific brand.

Here is where there are opportunities for more research on smells. This opportunity offers Design Research as a path to reach that goal. The progress that Design Research has been making in recent years is trying to bring people closer and to be more involved in the Design process to help ensure their ideas, needs and perceptions are being considered, and this approach would be essential for successful results in researching smells.

This research is not proposing to treat or change one’s state of mind or mood by using smells. Instead, it explores how Designers, by designing experiments, and then conducting and integrating findings can understand people’s perceptions around smells. The ultimate goal is to make use of the powerful relation between

6

smells and memory to help enhance the communication of ideas or concepts.

1.3 Introduction to Research Methods What has been happening over the past few years in Design is the transition from the making of stuff such as products and spaces to Designing for Experiences such as serving, healing, transforming. Thus, Design Research methods and approaches are changing and improving toward that end.

The objective of this project is to conduct a Design Research project in a scientific manner to enable the application of more rigorous methods in the field of Design. This approach is not commonly seen in Design. However, interdisciplinary collaboration has been increasing, and it is important for Designers and Design Researchers to understand new ways and methodologies of research.

When Design Research was not very diffused, Designers were mere recipients of research results. Design has been changing and today Designers’ and Researchers’ roles are blending together. “Design Research is foundational to creating products, services, and systems that respond to human needs.” (Lee, 2011). This approach is changing the way Designers work and create. More information about the evolution and change in the Design Research field will be presented in chapter 3.

7

If Designers want to collaborate and eventually adapt or create their own research methods that are more appropriate to whatever information they need to discover, acquiring knowledge through various methods of research is important. However, Design Research is a fairly new area, and Design undergraduate Education does not focus on research yet.

Learning design concepts such as typography, color theory, information design, visualization, and identity systems are important to Designers and Design Researchers. But in a changing world and with the evolution of Design, it is important for students to learn more than Design practice content. This content will always be important not only to Design practitioners, but also to Design Researchers, who at some point will have to interact with practitioners. The dialog is usually easier when they can speak and understand each other’s languages. Additionally, having knowledge of visualization methods, for example, might be useful when analyzing and interpreting data. Trying different techniques might help open new ways of interpreting results and looking at them differently.

Adding relevant content such as Research Methods or Behavioral Sciences to Design Education is/will be a big challenge. Most Design programs have only four years to teach all of the Design concepts stated before. So educators face challenges and tough choices when it comes to the curricular content, however it

8

is important to keep in mind that multidisciplinary experiences are critical.

Design for experiencing, one of the new areas in the Design field, requires knowledge of the basic concepts of Design, but also requires an understanding of Human Behavior and Social Sciences, areas that use rigorous scientific methods in their research processes.

Those scientific methods are largely used in those fields because there is a great concern about how the research is being conducted and the implications and generalizations that are possible from it. Some of the most important topics that the scientific approach considers are:

1. Systematic empiricism: relying on observations to draw conclusions. The phenomena must be objective and observable.

2. Public verification: research must be conducted in such way that the findings of the research can be observed, replicated, and verifiable by others.

3. Testability: research must deal only with questions that can be tested given current knowledge and research techniques. (Leary, 2008)

Experimental Design, belonging to the scientific approach to research, was the

9

approach used in this project. The benefits of using such method is to present more solid and reliable findings. The findings are solid and reliable because external variables are controlled and eliminated through the use of experimental manipulation, random assignments and experimental control. These controls make it possible to replicate the experiment, which enhances the public verification of the results.

There are two types of manipulation: control and experimental. Control means that one of the groups being tested will not receive any manipulation and will then be compared to the group that received manipulation. Using random assignments help to balance the participant characteristics as well as control for order effects. Experimental control means that everything will be held constant across the variables: setting, environment (e.g., noise, lighting), non-manipulated materials, and experimenter behavior.

These important aspects were studied and adapted in a way they could be applied to the smell experiment. In chapter 4, it is explained how it was done through adapted Experimental Design. Explanations can be found in the same chapter.

1.4 Roadmap to the study This project starts with some background information on how the sense of smell functions physiologically, and some information about psychological effects as

10

well. Some of the existing classification systems for smells as well as research tools used in olfactory research are described as a way of informing the next steps.

In a less scientific context, but still important, information on Fragrance Design is presented. It includes how Chemists design fragrances, what are some of the aspects to consider, and how perfumes are created and used nowadays.

Analysis of how smells are being used as another Design element, as well as being the subjects of discussions amongst Designers, will also be covered on Chapter 2.

Chapter 3 presents discussion about the Design Research landscape to identify the opportunities to apply more scientific methods in Design. How Design Practice and Research have been changing needs to be addressed to understand the importance of collecting and analyzing data in rigorous ways. Design has gone beyond the practice of making the world look more appealing. It is dealing with more important and substantial issues that require a better understanding of behavioral and social sciences, for example.

After gathering all the background information necessary for developing initial ideas and the hypothesis for the research, it was time to start developing the procedure to be used in the research. Based on the goals and the content, an Experimental Design approach seemed to be the best fit. Exploratory studies

11

were made to understand how people react in olfactory research, what kind of language to use when talking with them, and obviously, which smells to include in the study. The objective was to identify patterns amongst people’s associations when triggered by smells. If patterns were identified, we could assume that smells could be used as another Design element. Chapter 4 will explain the process of planning, conducting and analyzing the experiment.

Visualizations were used to help understand, and interpret the data gathered. Designers are visual learners, and are able to communicate and best understand concepts visually. Exploring different methods and techniques for the visualization helped reveal new ways of looking at the results, and different patterns in the data. During the visualization process, ideas for possible smell tools emerged. Those are then discussed in Chapter 6. Preliminary explorations of possible Smell Tools were an attempt to see how the results of this type of Design Research could be made useful for Designers.

Chapter 7 presents some fictional examples discussing ways that Designers, and researchers could use the results of this project.

In the last chapter, there is a reflection about issues and relevance of this project, and what the next steps would be.

12

Chapter 2

The sense of smell

It is easy to find Psychology or Neuroscience books, for example, explaining how the sense of smell works, how receptors are connected to nerves, taking that information that was just sniffed and transformed in the brain. But there is not much about how it relates to people’s perceptions and emotions. The close relationship between smells and memories and emotions is discussed because the sense of smell is directly related to the limbic system, which supports a variety of functions including emotion, behavior and long-term memory. However, it is hard to find scientists trying to identify what smells are connected to what emotions and memories, and how that connection occurs not only physiologically, but also psychologically.

The sense of smell and taste are called chemical senses. They are the only senses that are activated by detecting chemicals and then sending that information to the human brain. Comparing smell and taste, it is also relevant to state that the most important difference is that the sense of smell can be activated even if far away from the main components. This does not happen with taste. (Rodriguez-Gil, 2010)

13

Olfaction is also one of the most powerful senses in other species where it is used for food hunting, survival, and procreation, for example. It happens when the nose detects an odor molecule in the air. Human noses can be extremely precise. For example, natural gas is odorless, so companies add a smell molecule to the gas so that humans can perceive it if there is a leak. In this specific example, human noses can detect one of that molecule added for every 10 billion molecules of the gas. (Zimmer, 2010)

The sense of smell starts functioning when one inhales air. At that moment the odor molecules reach the nostrils, and dissolve them in the mucus. Underneath the mucus, there are specialized sensory cells that detect the odor. These specialized cells send the information to the olfactory bulb, located on the bottom side of the brain. From there, messages are sent to the limbic system, which is the most primitive brain center that influences emotions and memories, and to “higher” centers where they modify conscious thought (neo-cortex). At this point, the connections between the odor that was sniffed and memories, emotions, places, people and moments are created. (Rodriguez-Gil, 2010)

14

limbic system olfactory bulb olfactory receptors

odor molecule

Figure 2. Olfactory system (Lynch & Icke, 2007)

Human olfaction, or sense of smell, is usually said to not be very accurate. It is known that dogs have a very keen sense of smell and because of that they are trained to identify hidden drugs in suitcases or to look for missing people for example. However, physiologically, humans have also the ability to identify specific smells, in similar ways as dogs. What happens is that humans are used to super-estimate two other senses; vision and audition, therefore, the olfaction sense is forgotten and people don’t pay as much attention, and it is suggested that this happens because it is not significant for humans. It is also suggested “that the present poor olfactory acuity of humans is the result of a steady decline,

15

which started when our animal ancestors came out of the water and then became upright, eventually going up in the trees. (…) Presumably, odors tend to be concentrated near the ground and therefore human beings are preoccupied with vision and audition.” (Droscher, 1969)

However, it is known is that “our olfactory receptors are directly connected to the limbic system, the most ancient and primitive part of the brain, which is thought to be the seat of emotion. Smell sensations are relayed to the cortex, where ‘cognitive’ recognition occurs, only after the deepest parts of our brains have been stimulated. If the olfaction is actually more involved in “visceral and emotional activities” (Droscher, 1969), smells can play a really important role in communicating ideas and triggering specific emotions.

2.1 Describing and identifying odors Describing odors can be a very difficult cognitive task. First, there are not adjectives or names that exclusevely describe smells. It is necessary to borrow words from other senses to be able to describe a specific odor such as sweet, sour, burning, fresh. Second, the initial experience in smelling may be hedonic, a feeling state rather than a sensation. The stress on feeling is thought to be the reason that odor memory seems to be exceptionally good. (Engen, 1982) Thus, by the time we correctly name a “particular scent as, for example, ‘vanilla’, the scent has already activated the limbic system, triggering more deep-seated emotional

16

responses.” (Fox, n.d.)

Identifying odors can also be a daunting task. “Odor naming failures are often accompanied by strong feeling of knowing (FOK) or feelings of imminent retrieval of what that smell is.” (Jönsson, Tchekhova, Lönner & Olsson, 2005) According to De Wijk, Schab & Cain (1995), when asked to name odors, people tend to guess correctly, on average, 50% of common smells (e.g. coffee, vanilla and tar). This average is rarely exceeded. However, feelings of knowing what the smell is, and that the retrieval is imminent are very common. This event was named by Lawless & Engen (1977) as “tip of nose” in similarity with the phenomenon called “tip of tongue”. (Jonnson et al., 2005)

Some speculations about people’s inability to identify or name odors come to the fact that although there are a lot of teaching/learning regarding visual and its relation to their names, however there is no such thing for smells. Associating “odorous objects with their proper names is not as formalized in society (e.g. in school) as is naming of visual objects”. (Jonnson et al., 2005) Therefore, it might be very difficult to link an odor to its name. (Jonnson et. al 2005)

2.2 Odor classification systems There are several different odor classification systems. The majority uses chemical aspects of molecules to separate different types of smells. None of them

17

deal with emotional responses people might have. Additionally, most of them look too simple compared to the almost infinite possibilities of different smells. People having to place smells into those categories without understanding the more scientific side of the systems would probably not be able to place them correctly. Some more common smells, such as vanilla or coffee could not be easily contained in one of the classifications.

Understanding how smells are seen by scientists was important to identify gaps in the science of smells and help create the hypothesis for this study.

Some of the most known classification systems are described below.

2.2.1 Zwaardemaker smell system, 1895 According to Dutch physiologist Hendrick Zwaardemaker, odors can be classified into nine primary groups: Ethereal, Aromatic, Fragrant, Ambrosiac, Alliacious (garlic or onion), Empyreumatic (roasted coffee), Hircine (sour milk), Repulsive or Foul, Nauseous (faeces).

One important aspect is that even though there are nine groups, there are still smells that “do not readily find a place in the scheme at all”. (Ladd & Woodworth, 2006)

18

2.2.2 Henning odor system, 1916 Hans Henning’s classification has six primary categories or groups. He organized those groups in a prism. (Figure 3) He suggests that any olfactory stimulus occupy a position in the three-dimensional space, and all other odors are mixtures between the primary categories. (Stevenson & Wilson, 2006)

His system was created based on an experimental design using 415 smells and semantic differential methods. More information about semantic differential on Chapter 4, page 63.

The odor groups are: Fragrant, Putrid, Ethereal, Burned, Spicy, and Resinous.

Figure 3. Henning’s odor prism (Stevenson et al., 2006)

19

2.2.3 Crocker-Henderson smell classification, 1927 Crocker-Henderson classification, proposed in 1926, also did not consider psychological effects odors might have in people (Paulkner, 1965). It was inspired by Henning’s system. They proposed only four primary groups of odors that were termed in an odor square: Fragrant, Acid, Burnt, and Caprylic (goatiness). Those four categories were based on the “four types of smell nerves which are stimulated (...) by the various chemical excitants. Additionally, they related their classification scheme to the “phisycochemical structure of the odorant”. (Stevenson & Wilson, 2006)

This system provided a sample kit commercially available. The set was composed by 32 different smells belonging to one of the four groups of odors proposed. The advantage of this classification was the idea of creating a set which helped the industry to achieve “reproducible results in examining and testing the sense of smell”. (Ross & Harriman, 1949)

2.2.4 Stereochemical smell theory or the Seven Basic Smells, 1952 This seems to be one of the most popular models describing odor quality and similarity. John Amoore’s classification proposal corresponds stereochemical classes to an odor, adding up to seven types: Ethereal, Camphoraceous, Pungent, Putrid, Floral, Minty and Musky. Stereochemical classes are based on the shapes and sizes of molecules. For example, “camphoraceous odors all have molecules

20

shaped like balls, according to Amoore’s research”. (Engen, 1982)

2.2.5 Conclusion Here, only most common or important classification systems were presented. The approaches are usually related exclusively to Science and Chemistry rather than Social Science and Human Behaviors. Hence, classifications of odors fail to give clues regarding their psychological or emotional effects. They do not “provide the means for a reliable description of odor impressions or of mental images of fragrances”. (Paulkner, 1965)

2.3 Designing fragrances When fragrance professionals talk about their creations in a more scientific way they usually use terms other people would not be able to understand. This section will provide background information for Designers interested in understanding the world of scent in a more holistic way, with knowledge that might be useful for later applications of scent as another Design element.

Fragrances can be created in several different ways: using just one main ingredient or various ingredients. There are three types of chemical ingredients: natural, synthetic and artificial. (Headspace, 2010)

In order to design a fragrance, one must decide what is going to be the

21

relationship between the ingredients, if more than one. The relationship is given by what is called notes. Similarly to music, fragrances are made out of “notes” that together will create a harmonic scent. The notes can be top, middle, and bottom or base, and they have different functions when the fragrance is being scattered, spread, and consequently evaporated in the air. Progressively, people will perceive note after note over the evaporation process. (Burfield, 2000)

Top notes are the ones people notice first. They are also the more volatile ones, which means they evaporate faster. They create the first impression that will not stay for too long.

Middle notes are the ones that give personality to the composition. This is also called the main theme of the scent. They can last for a few hours but they are only noticed after the top notes evaporate.

Bottom notes are the elements that remain after top and middle notes have evaporated. “They create the memory that makes the theme linger in your mind, and help the fragrance last on the skin.” (Edwards, n.d.)

Fragrance Design has its own naming and classification systems. Those systems are also called olfactive families. They are usually used to group and organize types of perfumes, just like in a taxonomy. However, considering perfumes can

22

have several different components, with notes belonging to different families, it is difficult to designate one scent as belonging exclusively to one family. Some of these systems are developed by fragrance and perfume companies based on their ingredients’ repertoire.

Here are some of the broader perfume classifications:

2.3.1 Traditional fragrance classification Around 1900’s, the Traditional classification, one of the first ones was created. It is comprised of seven categories or families: Single Floral, Floral Bouquet, Ambered or Oriental, Wood, Leather, Chypre, and Fougère.

2.3.2 Modern fragrance classification After 1945, due to technological advancements, new methods to produce a bigger array of perfume possibilities were necessary. Then, new categories emerged, and the Modern Classification was developed. Its categories are: Bright Floral, Green, Aquatic (or Oceanic or Ozonic), Citrus, Fruity, Gourmand.

2.3.3 Fragrance wheel The Fragrance wheel created in 1983 by Michael Edwards is one of the newest approaches to perfume classification. It is widely used in retail and fragrance industries.

23

This approach consists of five families: Floral, Oriental, Woody, Fougère, and Fresh. Additionally, these families are divided into sub-groups and presented in the form of a wheel.

Figure 4. Fragrance Wheel (Edwards, 2011)

2.4 Using smells as a Design element Interpretation of visual stimuli, such as colors, is subjective to a lot of things such as culture, lighting environment, or personal preference. The same holds true for smells. If one sees a yellow paper, usually what is going to be triggered is a visual image, for example a lemon. However, if one smells a lemony scent, not only might the image of the lemon be triggered in one’s mind, but also memories

24

related to that scent, such as cleanliness or citrus, for example.

This variety of subjective responses to smell might be the reason why not only Designers but also other researchers have not started to research or use scents as another tool. It is an unknown and extremely subjective area.

The sense of smell cannot be turned off easily, and smells are everywhere, all the time. Scent happens both before and behind all other senses and hit us directly through the limbic system, which supports a variety of functions including emotion, behavior and long-term memory. Therefore, our associations with smells are incredibly strong, emotional, and cognitively bonded to memory. (Headspace, 2010)

With a better understanding of the sense of smell, Designers will be able to design for a more emotional, expressive, and even functional experience, adding texture, depth, and richness to our everyday lives. Designers should be aware of the spatial and perceptive potential of scent to start adding enriching detail to their projects. Scents might open up new ways to learn and remember.

Designers could benefit from dealing with not only the visible but also with what is invisible. Smell is invisible information and communication. (Barbara, 2010)

25

A book called “I Miss My Pencil” (Bone & Johnson, 2010) presents the idea of using a scent diffuser in a door bell that depending on who is ringing, will dissipate inside the house the smell related to that person. The scent is associated uniquely with each person who rings the bell. This way, at the end of the day, the house will be filled with scents, that mixed will represent everyone that visited that day.

Sissel Tolaas presented several possible uses for smells in a recent symposium, Headspace: On Scent As Design, at Parsons in New York. In one of the possibilities, she proposed to use smell not only as a communication tool, but also as a navigation tool in cities. One of her projects was done for Mexico City. The purpose was to navigate the city that was divided into around 200 neighborhoods, each one of them represented by a different smell. The smells were defined by tracking down a neighborhood’s identity smell. Thus, when entering a neighborhood, one would notice it just by the smell. Besides facilitating navigation, this project aimed to trigger people to discover the neighborhoods not only by using the sense of vision or audition, but also the sense of smell.

Tolaas also talks about using smells as a metonymy, a figure of speech, where a thing or concept is represented by part of it and not the complete idea. Just like the sound of an airplane flying might represent the whole idea of an airplane, the

26

smell of the airplane, if consistent, could represent the whole idea of the airplane as well.

“If people get the message through the nose, they really get the message. The more we understand about smell, the more we will be able to use its potential for whatever purpose: science, architecture, education, communication, navigation, life. Nothing stinks, but thinking makes it so. ” (Sissel Tolaas, Headspace: on scent as Design)

27

Chapter 3

Design Research methods exploration

Over the last ten years the Design Research space has been growing and changing, by adding new perspectives from other disciplines and “approaches that share a common or related goal” (Sanders & Chan, 2007) in order to complement and improve the Design Research process. The roles of Designers and Researchers are blending together and each one is trying to work in and improve on the other’s domain. This represents an important shift for both fields, but it is fundamental that the next steps are taken by studying and acquiring more knowledge together. Researchers and Designers are even “misinterpreting or misapplying the other’s methods and tools for Design Research” (Sanders, 2006). So it is important that they learn more about each other’s domain.

The Design Research landscape map discussed in this chapter is outlined based on Sanders’ Map of Design Research and Practice (Sanders, 2006; Sanders et al., 2007; Sanders & Liem, 2011), and modified to introduce Experimental Design.

28

Figure 5. Map of Design Research and Practice (Sanders et al., 2011)

The Map is represented in two dimensions: vertical and horizontal. On the vertical axis, the top part represents areas where Designers lead using designerly approaches to Design Research. The bottom part represents areas where Researchers lead using scientifically-based research methods. The Map started as being more populated in the bottom part (Sanders, 2006). However, with more Designers getting into Research areas, the top part has been receiving much more contributions lately than the bottom one. Designers do not explore as much the more scientific side of the research process because they still do not receive scientific education related to behavioral research methods, and because some

29

might think that the research-led mindset is more connected to basic research where the results may not be readily applied. However, Designers can contribute even on the research-led side of the Map. The knowledge Designers have of the practice combined with the knowledge of scientific methods and behavioral research might help them understand and generate ideas not only for innovative products, concepts and experiences, but also to create new research methods more suitable to the research they need to do.

The horizontal axis represents the nature of the research being done: expert vs. Participatory mindsets. On the left side is the expert mindset. Here, users are treated as subjects of the research. It is also called “Designing-For”. The right side uses the participatory mindset that has been expanding, and it is where people are seen as the experts. They are the ones who know what they need and want, but they might not be able to express themselves in regard to what they really want. So, Designers create ways to help them express their needs, wants, and emotions, and ideas. This approach is known as “Designing-With”.

Additionally, the Map presents circles in different colors and each one of them represents zones, clusters or bubbles. Big and transparent zones represent the main activities of research in the Design realm. There are four of them: Critical Design, User Centered Design, Design and Emotion, and Participatory Design.

30

The second biggest circles, also know as clusters, can be found inside the bigger zones. They are the different research approaches belonging to each one of the zones. Some examples of clusters are: Usability Testing, Applied Ethnography, and Human Factors Engineering, on the User-Centered Design zone; Scandinavian Design, and Generative Tools, on the Participatory Design zone. (Sanders, 2006)

Smaller circles, which are the bubbles, also represent approaches belonging to the zones, however, they are “smaller than clusters because they are not yet supported by professional organizations”. (Sanders, 2006) Contextual Inquiry, Lead-User Innovation, and Probes are examples of those smaller bubbles.

This research project aims to add more to the ongoing changes happening on the Design Research landscape by bringing a method that is not commonly used by Designers. This method could be applied to conduct different types of research that might bring new opportunities and discoveries for Design.

31

Figure 6. Map of Design Research and Practice modified

This “new” research method for Designers, used in this Research study, is placed on the Map on the User-Centered Zone because it is treating people as subjects who are observed doing a specific task.

Practitioners by nature, Designers might have a hard time working in a scientific environment where research results and findings are not necessarily applied right away. (Norman, 2010) However, there are opportunities for Designers to embrace the changes and start looking above and beyond what they are used to. In the early days of the Design discipline, Designers were trained to work with

32

the physical aspects of products. Designers were not trained on complex aspects of behavioral science, for example. “There is little or no training (for Designers) in science, the scientific method, and Experimental Design” (Norman, 2010), and the way Design Research is expanding and getting closer to other disciplines it is important for Designers to start to understand and deal with more complex problems. The way not only the Design field is changing, but also how everything is becoming collaborative, it is important that Designers are not perceived as “shape polishers”. In the future, more scientific approaches to Design should be developed, especially when it comes to Design Research, and it is important that everyone embraces the change. The complexity of problems that Design has been taking part of requires a better understanding of Science, Technology, Behavioral Sciences, Math, and Statistics.

“Designers are practitioners, which means they are not trying to extend the knowledge base of science but instead, to apply the knowledge. The Designer’s goal is to have large, important impact. Scientists are interested in truth, often in the distinction between the predictions of two differing theories. The differences they look for are quite small: often statistically significant but in terms of applied impact, quite unimportant. Experiments that carefully control for numerous possible biases and that use large numbers of experimental observers are inappropriate for Designers.” (Norman, 2010)

33

This study aims to adapt Experimental Design as a method to be used by Designers. Designers need to keep looking for bigger impacts, but still seeking truth. Applied science can be a great help if they still want to have large impact. It is important to start applying methods that address concern of bias, replication, and disclosure of the problem and findings. This way, Designers can stop being seen as merely creators of “pretty things”, and start presenting their findings as something that will be well respected not only by Designers, but also by the Scientific community. It is not and will not be easy to make use of all the constraints science requires, but it is essential that Designers understand the issues when researching human behavior.

“Service Design, Interaction Design, and Experience Design are not about the Design of physical objects: they require minimal skills in drawing, knowledge of materials, or manufacturing. In their place, they require knowledge of the social sciences, of story construction, of back-stage operations, and of interaction. We still need classically trained industrial Designers: the need for styling, for forms, for the intelligent use of materials will never go away.” (Norman, 2010)

The growth of new approaches to Design and Design Research is already a big and important step. Application of these new approaches do not happen overnight though. It is necessary to change ways of thinking, doing, teaching, and

34

learning. The traditional Design disciplines will always be relevant, and adding more perspectives will only help improve the field and increase recognition of Design’s relevance.

35

Chapter 4

Designing and conducting the experiment

Two pilot studies were conducted during the exploration of background information to help inform the study proposed in this thesis. The objective of the first pilot study was to understand how to use Experimental Design as a Design Research method. At the end, the first pilot study served not only to inform about the method, but helped populate the final pilot study and experiment with useful and important information, such as the list of most used words to describe odors that was later on used to help create the answer sheet for the final study.

The second pilot study was a rehearsal of the final experiment. It was important to conduct this study before the final one to understand the possible challenges that would be encountered and make small changes, if necessary.

4.1 Exploratory Study #1 During Spring 2010, in the Design 787 class, students were asked to develop a pilot study to test possible methods of Research and to learn about processes and potential challenges we may encounter when running a pilot test.

36

Objective The initial objective of the first exploratory study was to: (1) Learn how to plan and conduct a pilot study (2) Identify and understand processes and potential challenges on questionnaires (3) Identify and understand processes and potential challenges on an experimental design (4) Understand how to deal with human subjects in an experimental design method (5) Understand and observe how tasks are performed and how people react to them. (6) Understand how to control for order and carry-over effects (page 56) (7) Learn analysis methods to find similarities or patterns in participants’ responses (8) Test how much time an experimental design takes, and for how long people remain comfortable as participants

Participants Four Design graduate students were recruited to participate in the study: two males, and two females.

Structure and description The study consisted of three parts:

37

(a) A small questionnaire with demographic information.

(b) After the questionnaire, the study started with participants sniffing specific scents one at a time, categorizing them into good or bad, then into one of the seven basic smells categories: minty, floral, ethereal, musky, camphoraceous, putrid, and acrid (Chapter 2), and finally talking about personal experiences involving each one of the presented smells.

(c) At the end, there was a post-session conversation where participants commented about their experience in the smell study, and what could be improved in future experiments.

In part (b), subjects performed three tasks: (1) to categorize the smell into one of the seven basic smells (Chapter 2), (2) to classify the smell into good or bad, and (3) to talk about rational and emotional associations with the smell. Each participant was asked to perform the tasks in a different order and, at the end of the experiment, participants provided their preferred order to perform the tasks. The reason for varying the order was to learn what is the most effective sequence to get people’s responses.

To classify the smells into good or bad, participants were provided with green (representing good) and red (representing bad) stickers that were used to mark

38

their choices on a card. For the Seven Basic smells categorization, they also received stickers with each one of the categories. To comment on the smells, they did not have to write anything down. The objective was to have them talking and recorders would register their thoughts, that were later transcribed. (Appendix A)

Categories and smells Minty: Pertaining to, containing, or resembling mint. Smell 1. Crushed mint Smell 2. Twisted peppermint (oil)

Floral: Pertaining to, containing, or resembling flowers. Smell 3. Rose (oil) Smell 4. Lilac (oil)

Ethereal: Pertaining to, containing, or resembling ether. Smell 5. Nail polish Smell 6. Nail polish remover

Musky: Resembling the smell of musk; having a heady or pungent sweet aroma. Smell 7. Egyptian Musk (oil)

Camphoraceous: Pertaining to, containing, or resembling camphor.

39

Smell 8. Muscle liniment Smell 9. Moth balls Smell 10. Camphor (essence)

Putrid: Decomposed and foul-smelling; rotten. Smell 11. Hard boiled eggs Smell 12. Blue Cheese

Acrid: Sharp, pungent, or bitter. Smell 13. Vinegar Smell 14. Lemonade powder

To control for order effects, the smells were presented in a different and randomized order to each participant. The order in which the smells were presented could possibly affect the outcome. Perceptions people might have when sniffing something can be psychologically or physiologically influenced depending on previous experiences. For example, if a participant just sniffed gasoline, a very distinctive smell, and the smell after gasoline is lemon, a more subtle and light smell, the participant might have difficulty distinguishing lemon. Additionally, in Experimental Design, it is possible to have a carryover effect occur, which happens when two or more conditions, in this case two or more smells, are presented to the same participant, and the effect of the first smell “carries over”

40

to the second. Fatigue in experimental subjects is another example of carryover effects. There is a chance that participants will at the end of the experiment be tired of the procedure, and not as prone to answer the questions as well as they did in the beginning. (Leary, 2008)

To control for environmental conditions, the experiment was always done in the same setting: the Lighting Lab, on 3rd floor at Hopkins Hall.

Set up

Figure 7. Box with all the smells in numbered bags.

Figure 8. Participants’ desk: consent form, questionnaire, tasks order, stickers,

41

cards, categories’ definitions, and coffee.

Figure 9. Investigator’s desk with script, smell and tasks order, and notebook.

Analysis Questionnaire responses:

How sensitive are you to smell? . All participants responded they were fairly sensitive to smell.

What is your favorite smell?. When asked about their favorite smells, 3 out of 4 participants mentioned smells related to food.

p1: freshly baked bread p2: mint, eucalyptus, fire, cut grass p3: onion & garlic sautéing in butter

42

p4: chipotle JK, cinnamon

Participants’ classifications of the smells into the smell categories and good (green) or bad (red):

Figure 10. Classification of smells into good or bad

One of the tasks that was performed by participants was to classify the smells into good or bad. Most of them chose between good and bad for the majority of the smells. However, for some smells, the participants were not able to choose between good or bad and decided to use both stickers to demonstrate they did not want to make a decision.

Results: Interesting facts Both male participants related minty smells to being focused or organized.

The words that the participants used to describe the smells were documented to be used as “smell vocabulary” for subsequent studies. Here is a summary of words used by the participants to describe all of the smells: adverse, athlete,

43

awake, baseball field, bitter, breezy, brisk, calming, camping, chemistry, Christmas time, clean, confusing, cool, decomposed, dirty, dry, family heirloom (antique), feminine, flowery, foul, fresh, fungus, funky, garbage, gardening, green, grass, hanging out, healing, hiking, indoors, landfill, light, masculine, medical, moldy, old, organized, outdoors, outside in the summer, pine forest, pungent, putrid, relaxing, relief, repelling, sharp, sick, social, soil, soothing, strong, summery, sweet, vanilla, winter, woods.

Discussion and observations The session took thirty minutes on average. Three out of four participants thought it was a good length. One thought it took longer than expected.

The study was done in two days. The smells were presented using small plastic bags, and stronger smells contaminated weaker smells from one day to the other. It was determined that more hermetic containers were necessary.

During the study, when a lighter smell was presented after a stronger one, subjects had difficulty identifying the lighter one. Therefore, in a study like this, when one is not interested in measuring the intensity of an odor, the smells need to have a similar “weight”.

When analyzing how participants performed the tasks in the pre-defined order,

44

it was determined that the order of the tasks should not be dictated: at the end all participants had opinions about the best order, however depending on the smell, one of the tasks came out first. For example, usually when the smell was unpleasant, the first thing they talked was how bad the smell was. Smells that make people recall experiences, may make them want to comment and explain what they were thinking. Additionally, some of the smells just reminded them of something specific and people wanted to categorize it before anything else.

Participants usually used more than one category for each smell. Apparently, smells cannot be classified using just one concept. Multiple ideas were triggered by the scents. Using categories or words can be useful to help participants focus on some specific concepts, but the definitions should have been more explanatory and more prominent. Testing the seven basic smells classification by Amoore was a good starting point, however, this classification does not relate to the emotional side of what was triggered by the smells.

Smells that did not bring up memories were harder to classify and comment upon.

Next steps Learning how to conduct an experiment based on the first exploratory study was an essential experience. From that, it was possible to start planning exactly what the final experiment would be. It was clear that the categories and words

45

provided in the first study needed to connect more with people. Additionally it was important to select which smells to keep or not.

Analyzing the transcripts and videos of this exploratory study, it was noticeable that some smells were not creating relevant emotional connections. A good example of that is the nail polish remover. Participants were able to identify the smell but the only thing they would talk about was the experience of polishing nails. Moreover, some of the participants were male and had even less to comment on that smell.

Another good example were the smells in the Floral category, lilac and rose oils. Some participants thought they were smelling the same smell twice and even asked if the researcher was “tricking” them. Consequently, their associations were too similar and keeping both smells did not seem necessary.

Thus some smells started to be narrowed down. Then, planning for the final experiment started.

4.2 Exploratory Study #2 The second exploratory study was planned to be exactly the same as the final one and was conducted just a few days earlier. The objective to conduct such a study was to identify possible challenges that could be fixed in a timely manner.

46

Secondary research that was done helped to generate ideas on how to structure the final experiment. Semantic differential was an important topic studied to help understand how words and their meanings could be used to be associated with smells. According to the previous exploratory study, the most used words could be classified into categories such as: mood/state of mind, activities, memories, holidays/seasons.

The measurement of meaning: semantic differential The study of language as an index of meaning, and how people interpret the words is complex, and it will be simplified to the realm of this research project.

Semantic Differential is a type of a rating scale designed to measure people’s reactions to stimulus words, objects, events, and concepts. (Heise, n.d.) The connotations are used to derive the attitude towards the given object, event or concept.” Osgood’s (1971) semantic differential uses “pairs of common verbal opposites” to isolate the meaning of the stimulus.

Basically, if one is asked to explain what a word means, one will explain it with their own words. Depending on the word it might be easier or harder to give a definition. People will hardly ever have the same answer as one another. Comparatively, if one is asked for associations with specific words, there are bigger chances that the answers are going to be more objective and will overlap

47

“because of a common mediation process”. (Osgood, 1971)

In order to use words to be associated with a specific stimulus instead of having people emitting random words and telling stories, it is important to create a selection that will be representative.

“Rather than relying on the spontaneous emission of words relating to a particular stimulating sign, we need to play the game of “Twenty Question” with our subject: SOPHISTICATED – Is it hard or soft? Is is pleasant or unpleasant? Is it fast or slow? Just as in Twenty Questions, the selection of successive alternatives gradually eliminates uncertainty as to the object being thought about, so selection among successive pairs of common verbal opposites should gradually isolate the “meaning” of the stimulus sign.” (Osgood, 1971)

The strategy of using pairs of words to try to isolate the meaning of the stimulus, in this case each one of the scents, was used in an adapted way. Because of time constraints, and the number of smells being presented, the inquiry about the meaning of each one of the stimuli did not happen in a successive order but they were all presented at once on an answer sheet.

Based o the words vocabulary from the first exploratory study, and then from

48

Osgood’s experiment presented on the book “The Measurement of Meaning” (1971), words were selected. Most of the time it was possible to pair the word with its verbal opposite or another word that had similar value and/or meaning. For example, members of the family were important words to be considered and some of them were used. They were displayed in pairs such as [mother – father] or [grandmother – grandfather]. The purpose of using those words was the close relation smells have with personal memories.

Moreover there were words like color names, red, pink, black, and white, that do not have antonyms to be paired with. The solution was to group them according to their similarity in meaning, for example: [black – white], [violet – amber – beige – brown], and [yellow – blue – red – green – pink].

After careful analysis of the words, they started to be grouped based on their meanings. For example, words with more emotive connotation were grouped together. Some related with states of mind, or more abstract adjectives were the second group. Other ones used to describe objects formed another group. Action words were the fourth group. Finally were created four categories, and they were Memories, Sensations, Description, and Activities.

Words in Memories were related to specific times, people and places, i.e. [mother – father]. Things that are specifically related to personal memories.

49

In Sensation, words were borrowed from other senses as well as states of mind [happy – sad], or colors.

Descriptions contained more concrete words, usually used to describe things such as [small – large].

In Activities, words represented actions such as [washing – cleaning].

Objectives The objective of conducting a second exploratory study was to serve as a trial for the final experiment. Understanding and learning about Experimental Design and dealing with smells were obtained in the first study. Based on the analysis of the first pilot study, a final idea for the main study was developed and tested in this second test.

Participants Eight participants were invited to be pilot tested: 4 males and 4 females. They were mainly graduate students at OSU. One of them was a participant in the first study which was very helpful to receive comparative feedback.

Procedure Participants were instructed that they would be answering an initial

50

questionnaire and smelling fifteen different scents, one at a time, with their eyes closed. For each one of the smells they would be asked to select words from a sheet of paper based on appropriateness and relevance to the scent they just smelled. The same task was repeated for each one of the smells.

To control for order and carry-over effects, each participant had a different and randomized order of the smells and words on the answer sheet.

There was no control for environmental conditions. The experiment was conducted mainly in 372 Hopkins Hall, but eventually it had to move to a place that was more convenient to participants.

The smells were: Crushed mint dried leaves (smell #1), Lilac oil (smell #2), Cinnamon powder (smell #3), Vanilla scented candle wax (smell #4), Egyptian Musk oil (smell #5), Moth Balls (smell #6), Camphor essence (smell #7), Gorgonzola Cheese (smell #8), Lemon scented furniture polish wipes (smell #9), Sunblock Coppertone (smell #10), Gasoline (smell #11), Ginger (smell #12), Nesquik mixed with water representing chocolate (smell #13), Pomegranate oil (smell #14), Sandalwood oil (smell #15).

51

Analysis Information about the participants as well as the words associated with each one of the smells were entered into a spreadsheet to understand and analyze the data gathered. The number of times words showed up for each one of the smells was counted to reveal patterns. The objective was not to perform a deep analysis of the data gathered, but to start building analysis tools and methodology for the final experiment.

Figure 11. Analysis spreadsheet example

Discussion and observations The session took from 25 to 45 minutes, depending on the participants’ responses. None of the participants complained about the length of the study.

52

It was clear again that participants felt the need to name the scent, thus some of them wrote on the margins of the answer sheet what they thought the smell was. To facilitate this process a question “What first comes to your mind?” on each paper was added to the final experiment.

Some people circled several words for each smell and category, while others circled only one or two. This gave a good grasp on how people might behave and interact with the final experiment. There were unused words that could be removed from the final answer sheet. Additionally, there were words that the majority of participants wrote onto the sheet, and those were added to the final word selection.

Next steps After the analysis of the second exploratory study, the final experiment took shape. The smells and process were defined. Some minor modifications to the answer sheet were done, such as adding the question “What first comes to your mind?”, adding some new words, and removing others that were not used.

4.3 Final Experimental Design The research method used was an adapted Experimental Design. It is a posttest Design with no control group. The Design here is being called “adapted Experimental Design” because there is no specific response expected after each

53

one of the smells. The null hypothesis states that people perceive and interpret smells in a similar way. The treatment consists of fifteen different smells presented to each participant. They were asked to select words that describe what kind of reactions and perceptions the smells triggered in them.

Participants Participants were OSU students, and a few OSU employees, from an age range between 18 and 40 years old. No other characteristics were specified. To recruit participants, e-mails were sent to The Ohio State University database with a link for scheduling appointments.

That is what is called a convenience sample. It is a non-probability sampling when subjects are recruited based on availability, convenience, and usually proximity. Using a convenience sample makes it harder to generalize the results to the entire population because the sample used might not be representative enough. This sample was used due to time and budget constraints.

Ninety people were recruited, and each one received a $10.00 restaurant gift card as an incentive.

Stimuli Participants (32 males and 58 females) were assigned to fifteen different smells

54

presented in a randomized order. Smells were household and known smells from daily life.

The smells were: Crushed mint dried leaves (smell #1), Lilac oil (smell #2), Cinnamon powder (smell #3), Vanilla scented candle (smell #4), Egyptian Musk oil (smell #5), Moth Balls (smell #6), Camphor essence (smell #7), Gorgonzola Cheese (smell #8), Lemon scented furniture polish wipes (smell #9), Sunblock Coppertone (smell #10), Gasoline (smell #11), Ginger (smell #12), Nesquik mixed with water mimicking chocolate (smell #13), Pomegranate oil (smell #14), Sandalwood oil (smell #15).

Procedure Participants were instructed to answer an initial questionnaire and to sniff fifteen different scents, one at a time, with their eyes closed. For each one of the smells they were asked to select words from the answer sheet, based on their personal associations with the scent. The same task was repeated for each one of the smells.

The initial questionnaire asked them some demographic information including Name, College rank, Area/Department, Age, Gender, Place of birth, Place where grew up, if they were smokers, and if they were allergic to anything. If they were allergic to any of the substances relating to one or more of the smells, they would

55

be skipped. This did not happen to any of the participants.

The purpose of these questions was to group the participants during the analysis and compare those groups to find patterns. Questions such as “Do males and females perceive and interpret smells differently?” can be studied when demographic information has been collected.

The same questionnaire also asked them some questions about smells such as how sensitive they think they are to smells, what is their favorite smell, why, and if they can describe it.

In this case, the objective was to gather information that could be helpful during the analysis of the experiment. Looking at their answers and the reasons why people choose one smell as their favorite, for example, might be a useful piece of data when analyzing their individual responses for each one of the odors.

The sheet of paper they received to select words for the smells contained four different categories: Memories, Sensations, DescriptionS and Activities (more about the categories on page 66). The answer sheet can be seen in Appendix B. For each one of the categories there were different number of words. In the Memories category, there were 42 words; in the Sensations, 55 words; for Descriptions, 44 words; and for Activities, 25 words. Those words were divided

56

into word sets. Size of those word sets varied according to other words they were related to in the same category. For example, antonyms were usually placed as a word set of 2, [good – bad]. Some words that described similar characteristics, such as color names, could be presented in word sets of different sizes, [violet – amber – beige – brown]. For each smell, participants received a new smell selection sheet.

Other considerations in conducting this study are described below.

(1) To control for order and carry-over effects, the smells were presented in a different order to each participant. The earlier and later stimuli presented have different perceptions, because the subject might be more receptive and excited in the beginning of the experiment and tired at the end. Additionally, the random order helps to avoid possible effects that one smell may cause on the next smell being presented.

(2) Furthermore, the same strategy of randomized order was used with the words presented to each one of the subjects. The sets of words were in the same order to each subject among the fifteen different smells, however, they were in different order between subjects.

57

Randomization was created in a web-based application (http://www.random. org) where the investigator input numbers and the application created several different orders.

(3) To control for extraneous environmental conditions, the experiment took place always in the same setting, with appointments scheduled every 60 minutes starting at 9 AM, and with the last appointment starting at 6 PM on Mondays, Wednesdays and Fridays. On Tuesdays and Thursdays appointments were also scheduled every 60 minutes starting at 9 AM, with the last appointment starting at 12 PM.

Those measures helped to hold everything as constant as possible between all sessions and different participants.

The instructions to the participants were consistently delivered. Here is how the experiment was presented:

“This is a research about smells to understand how people perceive them, because I believe that if people have similar associations between smells and memories, sensations, we will be able to use smell as another tool for Design. For that, I am running an experiment with 15 different smells that I am going to present to you one at a time. For each one of the smells there is

58

a sheet of paper just like this. So once I give you the smell, I am going to ask you to close your eyes, I am going to bring it close to you and I will let you know. Once you could smell it, I am going to take it out, and you are going to open your eyes and circle as many or as few words as you think relate to that smell. Are you ok with this?”

If the participant agreed in performing the tasks, I would finish the explanation.

“There are no specific rules about circling the words. If you think there is no word that relates to the smell that’s good. If you think a lot of them relate that is good too. It is totally up to you and about how you feel. There are no right or wrong answers. There is coffee for in-between the smells because it helps to clear the palate, but also it is up to you. Use it when you want. Before we start, I need you to go through the consent form. On the third page there is a place where I need your signature if everything is OK. And the fourth page is a small questionnaire just about your background that I need you to fill out before we start. That is your copy of the consent form and your gift card. Please let me know if you have any questions. So, take your time. I am going to start setting up the audio and video recorders. Just let me know when you are ready to start.”

59

The study took around 20-45 minutes, depending on people’s responses to the smells, including the initial questionnaire and consent form. Each session was recorded with both audio- and video-documentation.

60

Chapter 5

Experiment Analysis

5.1 Quantitative Analysis Demographics The next tables present the demographic information of the study. In order to group the participants in the categories presented here, some adjustments were made. For example, in the Area/Department grouping, some participants filled out the questionnaire saying they were in more than one program, so they were counted twice. Similar adjustments were made in the Origin category. Participants said they grew up in more than one place, and so they were counted again in the two different locations they grew up in. The way Origin and Area/ Department were grouped can be found in Appendix D.

Gender Male

32

Female

58

Table 1. Experiment demographics: gender

61

Age Range 18-24

68

25-30

14

31-36

6

37-40

0

older

2

Table 2. Experiment demographics: age range

Department Art

16

Business

5

Communications

16

Design

22

Humanities

11

Languages/Linguistics 19 Medical Science

4

Technology

8

Table 3. Experiment demographics: department

62

Origin Ohio

67

Outside Ohio

18

Outside USA

7

Table 4. Experiment demographics: origin

Questionnaire The answers from the questionnaire showed that when asked what their favorite smell was some participants included more than one option stating that was difficult to select just one. Analyzing the answers, the smells could be grouped according to their nature and similarities.

It was noticeable that smells related to Food such as coffee, baked goods, vanilla, and cinnamon comprised the biggest group of answers. The second group of answers contained smells related to Nature such as grass, flowers, and rain. Other three groups were also created and they were: Memories, Sensations, and Artificial. They did not receive many entries. More information about the answers can be found in Appendix C.

Statistical analysis The objective of the statistical analysis was to identify how many times a word needed to be chosen in order to be considered significant to its association to a

63

particular smell, i.e. if the association was unlikely to have occurred by chance. Additionally, gender comparisons were made to study possible differences.

Words counting was done for each one of the categories (i.e., Memories, Sensations, Descriptions, and Activities) from the sheets of paper where subjects selected the words.

When using the answer sheet given for each one of the smells, some participants felt the need to annotate some of the pre-defined words. To be able to analyze the significance of those words as well, some annotated words were recoded as they were interpreted to be the same as some of the words from the given list. For example: mom cleaning* = mother + cleaning, mom’s storage closet* = mother + closet*, parent’s closet* = mother + father + closet*, mint green* = green. Words with asterisks (*) are the ones that were not on the answer sheet and were added during the analysis. For more information on how words were grouped see Appendix E.

The recoding of annotated words was done for all ninety participants on the fifteen smells presented.

During the experiment some people could not relate any words to some specific smells, or sometimes they couldn’t smell anything. In those cases, they did not

64

mark any words on the given sheet of paper. This happened three times: two times with smell #6 and one time with smell #3. The data from those participants was retained since the missing data was minor and not likely to affect the result statistically.

Defining significance levels One of the important goals of using a scientific approach to this research project was to be able to have data that could be analyzed by statistical tests. This way, relying on the data gathered, other scientists and researchers would be able to judge the research and make their own assumptions based on the facts. Interpretations could be more objective rather than subjective. (Significance in Statistics & Surveys, n.d.)

Since the objective of testing smells was to explore if people had similar associations to each one of the scents presented, it seemed clear that counting the words and checking if the count happened by chance was the first step of the analysis.

The everyday definition of “significant”, in English, is important. However, in statistics, it means “probably true”, i.e., that did not happen by chance. “A research finding may be true without being important. When statisticians say a result is ‘highly significant’ they mean it is very probably true. They do not

65

(necessarily) mean it is highly important.” (Significance in Statistics & Surveys, n.d.)

Scientists usually use significance levels of .10, .05, and .01, which represent 10%, 5%, and 1%. For this study, the significance level used is .01, or 1%. It will indicate that the findings, when significant, show a 99% chance that the relation between the variables (smells and words) found in our sample are true, i.e., did not happen by chance.

In order to calculate significance levels for each one of the words, they were grouped based on the word set size in which they were displayed on the answer sheet. The word sets contained either one, two, three, four or five words. For example, the word “mother” was displayed in the answer sheet side by side with the word “father”, [mother, father]. This means mother was in a word set size 2. Same things applies to the word “father”.

Statistical tests The assumption was made that, within a chosen word set, words being chosen or not are exclusive, which means the subject must choose one and only one word within the word set in order to be able to run tests for significance analysis. Assuming no preference for specific words, one word is chosen in probability of 1, or 1/2, or 1/3, or 1/4, or 1/5 in the word sets size of one, two, three, four, or five words.

66

Given no preference for any word set, the chance of each word set being chosen is evenly distributed to any word set in each category (memories, sensations, description or activities), thus 1/X for any word set in the Memories category, 1/Y in the Sensations, 1/Z in the Description, 1/W in the Activities. The blanks like “time: ___”, “places: ___”, and “people: ___” were treated as a word set too. In this way, X = Y = Z = W = 22, which is the number of rows in each category according to the answer sheet. word set size 2

word set size 4

word set size 1

Smell #1: What first comes to your mind? Memories

Sensations

Descriptions

Activities

row 1

grandma - grandpa

safe - unsafe

sharp - dull

smoking

row 2

city - farm

minty - musky - floral - fruity

clean - dirty

playing

row 3

babyhood - childhood

refreshed - weary

beautiful - ugly

working

row 4

forest - garden - mountains

stimulated - unexcited

rough - smooth

eating

row 5

breakfast

rational - intuitive

strong - weak

shopping

row 6

indoors - outdoors

wet - humid - dry

feminine - masculine

washing - cleaning

row 7

time: _____________________

anxious - relieved

large - small

dancing

row 8

teenage years

bright - dark

healthy - sick

camping

row 9

home

bitter - sweet - sour - salty

new - old

driving

row 10

people: ___________________

fast - slow

complex - simple

baking

row 11

summer - winter - spring - fall

happy - sad

pungent - bland

cooking

row 12

salon - hospital - beach - landfill

focused - unfocused

pleasant - painful

gardening

row 13

adulthood - elderly

hot - warm - cold

fresh - stale

relaxing

row 14

kitchen - bathroom - bedroom

violet - amber - beige - brown

natural - artificial

drinking

row 15

christmas - easter - thanksgiving

yellow - blue - red - green - pink

progressive - regressive

healing

row 16

gas station

constrained - free

fun - boring

sleeping - waking up

row 17

birthday

confused - organized

active - passive

growing up

row 18

places: ____________________

calm - agitated

hard - soft

running - swimming

row 19

mother - father

hungry - thirsty

valuable - worthless

hiking

row 20

morning - afternoon - evening

black - white

good - bad

decomposing

row 21

vacations - holiday

ornate - plain

heavy - light

socializing

row 22

friends - girlfriend - boyfriend

clear - hazy

rich - poor

hiding p1

X=Y=Z=W= 22, number of rows

Figure 12. Experiment answer sheet

67

In a given category, say Memories, and assuming no preference for any word, the probability of a word being chosen or not from a word set size 3, for example, is 1/(3*22). This calculation was made for each one of the words in each one of the categories.

In the memories category, words in word set size three were: [forest – garden – mountains], [kitchen – bathroom – bedroom], [Christmas – Easter – Thanksgiving], [morning – afternoon – evening], [friends – boyfriend – girlfriend]

In sensations, they were: [wet – humid – dry], [hot – warm – cold]

In descriptions, and activities there were no words in word set size three.

Note p0 is the underlying probability that a word will be chosen by people for some smell, assuming the smell does not actually affect which words to pick. Therefore we split the total probability evenly to each row and then to each word in the same row.

p_hat is the sample statistics, that is an estimate we got from the data for the true probability of a word chosen or not when people have a smell.

68

If p_hat is close to p0, then it means the smell does not make any difference, and people are choosing them randomly. If p_hat is significantly different from p0, then it means the smell does make a difference. The critical values we generated in the table are used to tell us how large a count will yield a significant p_hat.

Subgroup: Total total

n word set 1 word set 2 word set 3 word set 4 word set 5 90

8.68

5.33

4.05

3.36

2.91

Table 5. Threshold values for significant words: total

Subgroup: Gender male

32

4.19

2.68

2.09

1.75

1.54

female

58

6.32

3.95

3.04

2.53

2.2

Table 6. Threshold values for significant words: gender

In order to find out what threshold value to use, it is necessary to first select the word to be analyzed. Then, determine which type of word set the selected word belongs to. When you look at the values under each one of the subgroups like Total, or Male/Female, the threshold value will be found. For example: the word “mother” was in a word set 2 [mother – father], and we want to see what was the threshold value for “mother” on smell#1 for the totality of participants (n = 90). Looking at the word counting on the spreadsheet, on smell #1, for word “mother”

69

we see it appeared 12 times (Appendix G). Comparing with Table 6, crossing n = 90 and word set 2, we see a threshold value of 5.33, which means that any word from word set 2 that had counting over 6 will be considered statistically significant. This way, “mother” in this example is statistically significant. This was repeated with all words, for all smells. List of significant words for each smell can be found in Appendix G.

Comparing Male versus Female responses The objective of the gender comparison was to understand whether there is any difference between them in terms of evaluating and perceiving different smells.

For each smell and each of the four categories (Memories, Sensations, Descriptions, Activities), the words were listed in order of their importance for female, and copied for male. This means the top ranked words are most important ones for female and also possibly important for male but not necessarily. To include as many as important words and to avoid possible zero count, 15 top ranked words for female were picked as the set to do the comparison, for each smell and each of the four aspects.

The Chi-square test was used to test the level of significance and Table 7 shows the results. Numbers in orange mean males and females are significantly different at 0.01 level; numbers in green mean they are marginally different at 0.05 level.

70

NaN (a missing value) means some counts are zero and therefore the test cannot be performed. The closer the numbers are to one, means that there is very little difference between genders. Numbers closer to zero, represent a big difference between genders.

71

Smell #

Memories

Sensations

Descriptions

Activities

1 mint

0.3907

0.9053

0.3049

0.0681

2 lilac

0.1236

0.8109

0.3379

0.5522

3 cinnamon

0.4036

0.0575

0.1722

0.4824

4 vanilla

0.1299

0.4639

0.3897

0.5431

5 musk

0.2735

0.47

0.8629

0.2498

6 moth balls

0.0181

0.42

0.0099

0.0006

7 camphor

0.1272

0.7928

0.2937

0.1109

8 gorgonzola

0.6469

0.309

0.9995

0.0433

9 lemon

0.602

0.5757

0.5856

0.1049

10 sunblock

0.9121

0.0711

0.0562

0.3696

11 gasoline

0.0856

0.5443

0.4761

NaN

0.436

0.0026

0.5353

0.1961

13 chocolate

0.2379

0.3152

0.3097

NaN

14 pomegranate

0.2537

0.3772

0.0023

0.2657

0.22

0.9134

0.0002

0.1916

12 ginger

15 sandalwood

Table 7. Chi-Square test showing p_values for comparing females and males where orange represents significant difference at 0.01 level, and green represents significant difference at 0.05 level

72

Discussion Total number of subjects. The study demonstrates that most of the times smells trigger similar reactions in different people. Those findings indicate that the patterns amongst people’s perceptions related to smells exist, making them a possible tool for Designers willing to communicate concepts and ideas triggering other senses. On Appendix G, tables with significant words for each one of the smells can be found.

Gender comparison. In comparing genders, there are only a few instances where perceptions differ between males and females. Moth balls was the smell with more difference between genders; the numbers under Memories, Descriptions, and Activities showed significant differences.

Further qualitative analysis including comparison between smells, a summary of specific concepts triggered by each smell and comparison between gender perceptions can be found in the next session. Exploratory visualizations will help the qualitative analysis.

5.2 Qualitative analysis Statistically, the study demonstrates that smells trigger similar reactions related to memories, sensations descriptions and activities in different people.

73

The analysis and discussion presented in the previous chapter are extremely important and relevant to show that the words selected by the subjects did not occur by chance. This means that, when triggered by scents, people might have similar emotional interpretation of what that smell is communicating.

Those findings indicate that the patterns amongst people’s associations with smells exist, making scents a possible tool for Designers to communicate concepts and ideas by triggering the sense of smell.

However, there is just so much Designers can learn from statistical analysis and scientific results. In order to expand on the quantitative analysis, exploratory visualizations of the data were explored. Qualitative analysis and assumptions will also be discussed. Designers are visual learners and are able to communicate and best understand concepts visually. These visualizations will help to reveal the patterns and to create some sort of unique “fingerprint” for them.

So, in this chapter, visualizations for the data gathered are explored, exposed, and analyzed. In addition to which words were associated with which smells, comparisons amongst categories, individual analysis by smells and by words are presented. For this analysis, only the significant words defined in the quantitative analysis are taken into consideration. During the qualitative analysis assumptions are made which may not reflect scientific results. There is more subjectivity

74

involved in such observations, which might reduce the results’ reliability.

Psychologists have a method for assessing and increasing reliability when dealing with more subjective analysis. It is called Inter-Rater Reliability. For example, when analyzing videos of people’s behavior, more than one researcher needs to analyze the video and they must come to the similar conclusions. (Leary, 2008)

For this study, it was not possible to have more than one researcher analyzing the data qualitatively.

5.2.1 Word clouds The first step to this exploratory visualization of the results was to create word clouds. It seemed natural that in a study like this, where words represented people’s response to the smells, word clouds would be a perfect fit. The words pictured on the clouds are only the statistically significant ones per smell.

Some principles of Design, such as color coding and different type sizes, were used in the word clouds to help communicate findings.

Color coding was used to represent the four different categories (Memories, Sensations, Descriptions, and Activities) to which the words belonged in the experiment answer sheet. It was important to choose colors with similar values,

75

i.e., darkness or lightness of the color, so they would have the same visual weight. It was also important that each color could be differentiated easily, this way two shades of the same color, such as red and pink (light red) would not work as well. Decision on which colors would represent each one of the categories was based on the theory of warm and cool colors, and also on some color psychology/ symbolism. Memories, the most personal and intimate categories, is represented in blue, the color that represents introspection, and depth. Sensations, which can be very active, but still relate to personal emotions is represented in magenta, which is a mix between a lot of red and some blue. Red is the color of strong emotions, and blue has the introspection side. Descriptions is the most hermetic and least personal category. A cool color, in this case green, was chosen to represent the category. For activities, a more vibrant and warm color was chosen, because of the nature of this category. (Smith, n.d.)

Differentiation on the type sizes of the words in each word cloud represent the number of times a word was selected.

This visualization presents a clear overall comparison amongst all the smells, and some preliminary assumptions can be made based on the sizes of the clouds and words. Another thing that will be observed is how smells can be interpreted as more recognizable or not depending on the characteristics stated above. Additionally, the overall size of the cloud will be analyzed as a way of observing if

76

a smell had more or less associations with words.

At the end of this chapter, a summary of two of the smells containing all visual aids created will be displayed in a larger format so more details can be noticed.

Figure 13. Word clouds: size represents number of times a word was chosen (1)

77

Figure 14. Word clouds: size represents number of times a word was chosen (2)

78

The first observations that can be made by looking at the clouds are that some of the words stand out because of their sizes. For example, for the smell “gasoline”, the most prominent concepts are “gas station”, and “driving”. The same happens with mint (“minty”), cinnamon (“Christmas”, and “baking”), and chocolate (“baking”, “eating”, and “sweet”). Since the words people associated with those smells are closely and obviously related to them, one can assume that those smells were identified. Furthermore, it is possible to see that those smells had a high level of consistency amongst the words selected, i.e., a considerable amount of people selected the same words to represent each one of those smells. Those are also the word clouds with bigger sizes, which means that people had more associations with those smells when compared to the others.

In the second level of consistency are smells like vanilla, gorgonzola, lemon, sunblock and sandalwood. Notice how the size of words still varies, but much less than on the previous examples. It is possible that people still recognized some of those smells based on the selected words. For example, in sunblock, words like “vacations”, “outdoors”, and “summer” stand out, vanilla cloud has “warm”, “baking”, and “sweet”, while lemon has “clean”, “cleaning”, and “fresh”.

The third level of consistency has very little variation on the sizes of the words. It is noticeable also that the overall size of the clouds is much smaller when compared to the other two groups. Smells in this third group are: lilac, musk,

79

moth balls, camphor, ginger, pomegranate. It is possible to assume that the consistency level amongst people’s response might be directly related to the level of recognition of the smells. For example, if someone is asked to think of the smell of lilacs, camphor, or musk, it is possible they would have a hard time trying to recognize what those smells really are, and to make associations with them.

Word cloud analysis summary: 1. Bigger clouds = more associations 2. Bigger words = more associations and consistency 3. Larger number of words = less consistency 4. Smaller variation of word sizes in a cloud = less consistency

5.2.2 Gender Comparison word clouds Figure 14 shows some exploratory visualizations still using the word cloud method to compare two smells: one with great difference between genders (moth balls), and one with little difference between genders (lilac). Note that female word clouds are bigger than male because there were more female subjects.

Word clouds were not enough to reveal patterns for gender comparison. It is clear that there are differences in the selected words amongst male and females, but there are also words that are similar. Additional analysis and visualizations are necessary to be able to discuss gender differences.

80

81

Figure 15. Gender comparison word clouds

lilac

moth balls

male

female

5.2.3 Categories’ comparison Different visualizations to compare categories were explored. Each one of them presents the information differently and facilitates the interpretation and assumptions that can be made by analyzing the visuals. They were deliberately exploratory and the main objective was to explore ways of looking at data and draw conclusions based on those visualizations.

In this analysis, the number of words selected in each one of the categories for each one of the smells was counted. Considering that the categories did not have the same number of words, all numbers were converted to percentages so they could be compared (Table 8).

The list of words belonging to each one of the categories can be found on the asnwer sheet in Appendix B, or on page 85.

82



memories

sensations

descriptions

activities

1 mint

4.58%

6.22%

6.47%

4.80%

2 lilac

6.59%

4.87%

7.52%

4.58%

3 cinnamon

9.44%

6.06%

6.11%

8.58%

4 vanilla

4.58%

3.50%

4.37%

7.69%

5 musk

3.66%

3.43%

6.97%

3.20%

6 moth balls

5.40%

3.92%

8.26%

2.13%

7 camphor

5.84%

4.93%

7.02%

3.56%

8 gorgonzola

2.67%

3.86%

7.96%

2.89%

9 lemon

5.82%

7.19%

8.36%

6.09%

10 sunblock

7.49%

4.44%

6.64%

6.49%

11 gasoline

5.34%

3.21%

7.90%

3.78%

12 ginger

4.10%

4.42%

6.44%

3.69%

13 chocolate

6.72%

5.58%

4.80%

8.93%

14 pomegranate

5.71%

6.69%

7.83%

4.40%

15 sandalwood

5.00%

4.52%

7.70%

2.89%

Table 8. Percentages of the average of words chosen by a single participant

These percentages represent the average of words chosen by a single participant. For example, on average, each participant chose 4.58% of the words in the category Memories for the smell Mint. Since there were 42 possible words, it means that 1.92, on average, were chosen per participant.

83

Categories’ comparison visualization #1 The next chart (Figure 15) below presents a readout of which smells had bigger association to each one of the categories. If one needs to use a smell to trigger more personal feelings, one can look for smells that have bigger circles under the memories category for example.

It also provides readout on how categories compare for one specific smell. Musk is a good example of a smell that has a lot of Descriptions associated with it. Descriptions is the less personal category, so one can assume that people either did not recall what the smell was and/or could not personally relate to that smell. Thus, it was easier for them to try to describe the smell.

84

Figure 16. Categories comparison chart visualization #1

85

Categories’ comparison visualization #2 From this next chart (Figure 16), one can learn which scents had stronger associations in each one of the mentioned categories.

That information might be helpful when one is trying to develop a fragrance that, for example, has the objective of triggering more personal feelings, which would be found on the Memories category.

It was the second comparison chart created. As stated above, here, the aspect of comparing categories for smells individually is more prominent. However, it does not provide an easy readout on how categories vary between smells.

86

Figure 17. Categories comparison chart visualization #2

87

Categories’ comparison visualization #3 The final chart (Figure 17) was developed from chart #2. The objective of this one was to create a visualization that could be used to compare the smells based on categories. For this visualization, note that the smells are presented in a different order. They are grouped based on the similarities of the colored circles. Using this strategy, it was possible to create six different groups.

In group 1, cinnamon and sunblock were placed. Both smells have similar configurations differing only in the total number of associations, and on the descriptions category which is the second largest on cinnamon, but third largest on sunblock. They had a large amount of associations with the Memories’ category.

Lemon and mint are part of group 2. Their configuration is again very similar, differing on their total sizes and proportion between the categories.

On group 3, are chocolate and vanilla. Although they have similar configurations, it is not like the previous ones. The proportions are much more different between the categories. Additionally, their configuration of sensations and descriptions is inverted.

Group 4 is the biggest group. Descriptions was the biggest category for all of the smells. Those were the smells that people had a harder time identifying because

88

they were either unknown, or participants could not personally relate to them.

Group 5 and 6 have a lot of similarities with group 4. They are still being presented separately to emphasize some of the differences all those groups have.

On group 5, ginger and sandalwood were placed. Those smells previously stated have less consistency. Again, Descriptions was the largest category, showing that there are less personal associations to those smells.

Gasoline comprises group 6. Despite its similarities, gasoline was the only smell with more descriptions associated, and sensations as the smallest circle. Additionally, as stated in previous analysis, this was one of the smells with the most consistency, and recognition level. Gasoline was a very strong smell. It looked like most of the people could identify what the smell was based on the most used word which was “gas station” (see word cloud on page 97). Descriptions was a category less related to personal emotions and feelings. So, apparently gasoline smell does not trigger emotions, but people can easily describe/explain the essence of the smell.

The two groups where activities (orange) had a bigger count were 1 and 3. Sensations (pink) did not exhibit drastic variations amongst all smells.

89

90 group 6

group 4

group 2

Figure 18. Categories comparison chart visualization #3

group 5

group 1

group 3

5.2.4 Words’ comparison This next visualization (Figure 18) was exploratory. It would only work on a digital, interactive environment. It will be presented here as an initial step of a project that can be implemented by Designers in collaboration with computer science engineers.

The objective was to provide a way of organizing the smells using the words from the experiment as the starting point.

From this comparison, Designers can select keywords and find smells that relate to those words. They can also look at how many times a word showed up. Additionally it is possible to check how consistent the communication of the concept will be, depending on how many smells were associated with that word. Examples are presented with more detail in chapter 7.

Principles for Digital and Interactive Design can be used to facilitate the manipulation of information. By using mouse over, the word could be magnified and reveal more information about it. By clicking on the word or smell, new information could also be revealed. Application of these visuals will be discussed on chapter 6.

91

000: wor yellow 070: active [camphor lemon sunblock gasoline ginger mint lilac cinnamon] 000: adulthood 048: afternoon [su nblock chocolate mint mothballs gin 045: agitate ger lilac pomegra d [mothbal nate] ls gorgonzo 036: amb la camphor ga er [cinna so mon van line] 038: an ill a sandalw xious [g ood pom asoline 238: a egranate mothb rtificia alls cam ginger] l [gaso 012: phor g line m baby orgonz hood othba 079: ola] lls mu [lilac bad sk cam ] [gor 164 phor s gonz :b unblo ola g 184 aking [ ck san a : ba cho solin dalwo cola thro e 045 m od lila o te, c om t : be h c lem b inna alls c [lila 105 ach on min amp mon c : [ l t ging s e b unb hor mo , van 048 eau er van n lock i tifu c l : l illa go a a b ] mp 04 e l l rgonz i d [ l h pom ac] 8: b roo o ola ch r mu 03 m e e ocola g i s ge 2: k [ r te] p a g n o i 08 bir nge [ a m v t a e r po san 02 5: bi thda nilla egran m d e t y alw gra ate [ch min 01 9: bl ter [ nat ood s tg e su 04 6: b ack gorg ocol org anda van nbl l a l [ w 6 o i a o m ock t l la m 00 : b nd ev oo nzo nz o san dm t o a 6: lu u l h [ n a dalw l s am km illa ch o bo e [ gor bal t ood hb o i ] ls n o c rin sun go t su alls t o min g h l a n a ba nbl b g te t go mu lls [m loc zola soli o p rgo s c o k k ne ga me km nzo v l ot m e a mo s n c la m g u hb o i h r l i la c nc lin an oc oth all nt s sk] inn ate e a ball o m g an lat s] am p s i s] n a h on] e] da nd ge or] alw rc lw ho oo oo co dm d lat lila oth ec cc ba a mp lls am sun ho ph blo r mi or ck] nt m cin us na kg mo in ge nl em rl em on ] on po m eg ra na te ]

er]

]

or

ph

d

illa

sa

nd

oo alw

o

am

c

]

ne

oli

s ga

ing ng

d

oo alw

nt mi

r ho

lem

d]

mp

ca

on

lem int

oo lw

sa

a nd

d er an m an tv ing es ate g t in a l an la m r l o i r g e oc an ge ch om in sv ck n] la p all l o kg i l b c n mo th nb lo va nna o u i b c c s a m n lil un alls ck ola cs thb mo ] nz blo mo ila na l n o a n l l i u ood i g e s ] alw at or kc van or d s n g n n ] d h a u o o p ra es ate m on wo am eg olin cam m te ran dal gas lls eg d cinn om nna na san e a t a m n e a r b t i p n po th rie t [ on ana gra e c meg int yf fas mo egr me lat o om em on t po bo eak [lem oco ac p t p n i a : r m l t r 0 h ] il co na ck m pho 00 1: b righ n [c illa l lock blo cin am cho sun lls] ck c 02 6: b row [van unb olate illa s a l o l n l b a a s c nb ] 12 8: b lm thb oth g[ nv ho t su lilac ck] mo 11 8: ca mpin od [c namo zola m nblo illa min nilla n a r a v e n n 9 a o on su v r i g o o r m 1 9: c ldh c n h g [ le o i r p h g s nzola mp cam 00 : chi stma ine go musk orgo r ca usk 1 i l ger g 17 : chr gaso n lilac nm nge i in o g g m e [ 2 c solin 09 : city [lemo on lila inna sk ga int c 032 : clean g [lem d mu nm o o o n m lw 320 cleani nger le anda : i nt] alls s 180 clear [g phor mi othb m c la] : la onzo 091 old [cam mphor li s gorg c a ll : c [ a 2 b x 2 h n] t 0 ple int lemo sk mo com lls] ginger m d [mu othba 094: la e o s m z e n fu o n rg on soli nilla go 021: c ed [ga mon va nstrain te cinna la o c 015: co o h oking [c ood ginger] 162: co phor sandalw ng gasoline cam e at ol oc ch 000: danci s hball gonzola mot 067: dark [gor la] nzo rgo [go ing pos 021: decom 099: dirty [gasoline gorgonzola mothballs]

Figure 19. Words’ comparison chart (detail)

028: drinking [mint lemon] 055: driving [gasoline] 035: dry [min t cinnamon m othballs cam 032: dull [g phor] orgonzola musk mo 027: east thballs gin er [choc ger] o 178: e late lem ating [c on pom egranate hocola 043: e te cinn vanilla] lderly amon [moth 099: vanilla even balls gorgo musk ing [ 043: nzola mint camp fall [ mint g cinn hor sa c 029 inger le amo n : far innamo d a mon] n van lwoo m [g nm 009 d lilac illa c othb org : fas ] amp onz alls c 027 t [g hor c ola amp a : hoco cinn hor 242 fathe soline late g amo vani ] r orgo lla p nm flo : femi [gaso nzola ome oth ral nin line ging b g foc a r e [p anat l c er sa l s a g m e] aso om ndalw p for used h line egr or m ood e ] pom ana fre st o thb egra t e e alls nate fre san ] musk d s alw fri h gaso ood line] fru end l i lac ity s mu sk s un blo ck van illa le

92

5.2.5 Summary of the smells In this section, a summary of two of the smells will be discussed. Their visuals and short conclusions are presented. The two smells selected were Cinnamon and Ginger. They are both related to food, but have completely different “fingerprints”. The summary will show the type of information collected and how all visualizations work together to help the audience understand and analyze smells.

Cinnamon (powder) word cloud word cloud

Figure 20. Cinnamon word cloud categories’ comparison charts categories’ comparison charts

Figure 21. Cinnamon categories’ comparion chart

93

Looking at the word cloud, the more prominent words are “Christmas”, “baking” “pleasant”, “cooking”, “kitchen”, “eating”, and “warm”. From there, it is possible to assume that there is a big chance that people could identify what the smell was since it is a spice used in cooking, and very present in Christmas time.

From the word cloud is also possible to learn that there was no big word that had a unpleasant connotation. Consequently, Cinnamon is a smell more related to good/pleasant associations.

From the categories’ charts, it is noticeable that Memories and Activities were the categories with more associations. Even though one might think cinnamon would be a good smell for an active sport brand because of the big number of associations in the Activities category, it is important to go back to the word cloud. Notice that the Activities’ words associated with cinnamon are related to cooking, baking and eating and not to any kind of physical activity.

Ginger word cloud

Figure 22. Ginger word cloud

categories’ comparison charts

94

categories’ comparison charts

Figure 23. Ginger categories’ comparion chart

From the Ginger’s word cloud, one can see that there is a big amount of words and not much variation on size. It happened because there is not a very common sense amongst people about this smell.

The more prominent word is “clean”, but it is still considerably small when compared to the words in the Cinnamon’s cloud. In this example, it is already noticeable that there is a lot of green words, which are from the Description category. Ginger is a spice used for cooking, however, there are not many words related to what the smell really is. It is possible that people did not identify what the smell was, and also did not have personal associations. Consequently it was easier to use more descriptive words.

Looking at the charts, and as seen before in the word cloud, it is proved that the category with more associations was Description. The other three categories had a similar number of words associated with the smell. Memories and activities, which are the two categories more related to personal emotions/

95

feelings, considering they are the ones where more concrete things/words were presented, are the two categories with least words. This, again, may have happened because people either did not identify the smell and/or could not relate personal emotions to it.

This was also a smell with associations not related to unpleasantness. There were not big words related to negative connotation. But also there were not many prominent words that relate to pleasure/good. So it is not possible to conclude if it is a pleasant or unpleasant smell.

96

Chapter 6

Exploring Smell Tool concepts/ideas/prototypes

After analyzing the data, one of the most natural outcomes would be to create a tool with the smell information gathered from the research to help Designers and Researchers to apply what was learned here. However, the explorations in this project were experimental and the main focus was to inform Designers about other methods they can use in their research as well as to test if smells trigger similar associations.

The development of possible concepts for the “Smell Tool” was exploratory. Three ideas for presenting the findings were roughly analyzed and discussed. They will not be presented in detail and that is deliberate.

The models are divided by their characteristics. The first one is 2D, in the form of cards. The second one is a 3D model, in the form of a cube. The third one is an interactive tool. These are not finalized layouts or concepts. They are explorations of how all the information gathered from the research can be presented and used in such a way that Designers might get more interested in using the information.

97

2D model

In this model each smell would be represented by one card containing all the information gathered. There is the possibility for color coding the cards based on the nature of the smell, or even on how it was perceived by people in terms of its categories, for example. Also the real scents could be presented with “scratch and sniff” technology, i.e., the paper of the cards could be treated with a microfragrance coating related to the scent they were presenting.

Figure 24. 2D Smell Tool concept

Advantages of this concept are that they can be easily manipulated and, if there are more smells being researched and more information being gathered, it would be easily added to the existing smells and cards. They are also easy to store, and manipulate.

98

Disadvantages would be that the information is static. It might be difficult to look at the data from different starting points, such as words. Additionally the space reserved for the information might not be enough to explore all the topics one might find important about the smells.

3D model This model presents the information about the smells in the six faces of the cube. Compared to the cards, there is already the advantage of having more space to display the information. Additionally, the scent being represented visually in the faces of the cube can be contained inside of it. Not only the smell can be contained inside the cube, but also photographs that relate to the words in the word cloud of each smell.

Figure 25. 3D Smell Tool concept

99

A concept like this might be useful for Generative Research that will be dealing with scent, for example.

This model has more space to display the information. However, there is not much manipulation, interaction, and comparison of smells’ information that can be done. The information is still static.

Storing the cubes can also be a disadvantage. There are only fifteen smells presented in this research, so there would be fifteen different cubes. However, if more research is done on other smells, storing all those boxes might be problematic.

Interactive model This model would present the information as an interactive application. It could be a Web-Based or a Stand-Alone Application for computers or tablets. It would be an easier way to disseminate the content to other Designers or Researchers. Just like any other digital/online application, it would hold out possibility of ondemand access to content any time, anywhere, on digital devices.

In this concept there is interactivity between user and information. Information is not static. There is the possibility of in-depth exploration by the user. Software intelligence can cross all the information and easily present it in different ways.

100

Figure 26. Interactive concept (OECD Better Life Initiative, n.d.)

Figure 25 shows 4 different screens of an Information Design website about Life Index in different countries. There, it is possible to select topics according to what is more relevant, and then the organization of countries will change depending on those topics (screen 2). Additionally, there is mouse over options that provides more information about whatever is necessary in a fast way (screen 3). Then, clicking on specific countries, in this case, the user can get even more detailed information.

This is just an example of what an interactive application could look like and

101

offer. Necessary adaptations to the Smell Tool would be made. User could sort the information by word or by smell, for example. There is also the possibility of having a search bar, where one would easily find smells relevant to the keywords searched. Mouse over could also be applicable where the user would have quick access to more information.

One of the main disadvantage would be to not have the real scents linked to the visuals being presented. It also do not allow a more holistic vision of how complex or simple the information is, for example, how in depth it is possible to navigate or to compare more than one information visually, because there is a limitation of showing one screen at a time.

102

Chapter 7

Discussions and Conclusion

7.1 Discussion: exploring and applying the research’s findings In this section, the research project findings’ will be explored and discussed in a more practical way. The objective is to present how to utilize the information provided by the Experimental Design and also how to add more research data, and findings into the already existing database.

The Experimental Design explored how consistent people’s perceptions, memories and associations were when triggered by specific smells.

With all the qualitative and quantitative analysis, it is necessary to step back and analyze how all this information can be used by Designers. For that, in this chapter, three examples will be discussed to show how and what Designers and Researchers can do with the information gathered in this study.

103

7.1.1 Checking if a chosen smell is consistent with pre-defined concepts A group of marketing and branding executives of a retail company are reevaluating their identity: from visuals to Service Design, and in-store sensory experience. Their objective is to check if they are communicating consistently the brand’s concepts, vision and mission.

The brand sells fine accessories and gifts for women from handbags and clothing to jewelry, shoes, stationery, glasses, baby and home. Utility, wit and playful sophistication are the hallmarks of the brand. The shops need to always be warm and inviting. The main concepts to be communicated are: feminine, fun, sophisticated, warm, and inviting.

The smell they use in their store has as its bottom note (section 2.3) a Pomegranate oil scent. To understand if this smell is appropriate, they decide to check the Smell Toolkit Looking at some of the characteristics of Pomegranate smells it is possible to understand the associations women make when triggered by that specific scent.

104

Figure 27. Pomegranate word cloud for females

Looking at the summary for Pomegranate, it is noticeable that words like “feminine”, and “pleasant” are very prominent. “Floral”, “relaxing”, and “shopping” are also relevant in this word cloud. There are no words that would go against the brands’ concepts.

It would be interesting to test their scent in a similar way smells were explored by this study, or even compare several scents to see which is the closes to their intent. When re-testing smells, the answer sheet can be adapted to include some of the words they want to find out if people associate with the scent. Quantitative analysis would have to be done in a different way if the number of words, word set sizes or categories changed, for example.

105

7.1.2 Looking for smells associated to specific concepts Design researchers are hired by a Interior Designer company to provide insights on how smells could be applied as a differentiation for specific Interior Space projects for one nursing home.

Not many Design Researchers have started using scents in their Participatory method ToolKits, and they do not know how to choose from an almost infinite database of possible scents and scent combinations.

To facilitate the application of scents as another element of their tool kit, they consult the Smell ToolKit in a search for associations they want to trigger in people but also associations that they do not want. This way they can be sure they are in the right track.

As an example, below is a list of keywords that would be associated with the nursing home’s concepts. Based on them, they will see which smells they should add to ToolKit.

List of words: Calm, Bright, Beautiful

According to the Words’ comparison chart (page 112), it is possible to find, by word, which smells are associated with those three keywords. The numbers

106

represent the counting of how many times the word was selected. Additionally, the smells appear from most associated to least. However, all those smells were somewhat associated with those words.

198: calm [vanilla, lilac, pomegranate, musk, cinnamon, sunblock, chocolate, sandalwood, mint, camphor, lemon, ginger]

126: bright [lemon, pomegranate, lilac, sunblock, ginger, mint, vanilla, sandalwood, camphor]

105: beautiful [pomegranate, sandalwood, vanilla, musk, mint, sunblock, lemon, cinnamon]

The information shows that those three keywords were selected a considerable number of times for various smells. The smells more related to Calm are Vanilla, Lilac, and Pomegranate. For Bright, the most associated scents are Lemon, Pomegranate, and Lilac. Beautiful’s associations of smells are Pomegranate, Sandalwood, and Vanilla.

Based on those associations with the selected keywords, it is possible to narrow down some smells to be added to the ToolKit. Pomegranate was the smell common to the three keywords. Additionally, Vanilla and Lilac were cited in two of the smells.

107

7.1.3 Researching associations for a new smell Psychology students at OSU have been looking for existing research where Experimental Design is used in an applied research context. Their objective is to try to add some of their knowledge to new research and collaborate with other disciplines, and also improve on their knowledge about Experimental Designs.

After looking at the Experimental Design done with fifteen different smells in a Design context, they decide to add fifteen new smells to the database.

In order to replicate the experiment as close as possible to the original it is important to look at all the details. Keeping everything as constant as possible from the previous study will help to not create noise in future findings, and then both results can be compared and added to each other.

Aspects such as the demographics of people who were recruited are important to keep constant so they would have similar audience. It would also be helpful to keep the same number of smells, so the length of the experiment would be similar. The answer sheets need to be the same and can be found in Appendix B. The speech given by the experimenter also needs to be similar (page 75). The statistical tests need to be the same as well.

108

On Chapter 4, all this information can be found to help them replicate the experiment.

7.1.4 Probing deeper into patterns There are opportunities to explore other interesting findings from this study. Here, the discussion was about individual smells, visualizations to inpire Designers, and possible applications for smell as another Design element.

However, differences between gender were found but not explored. Another idea for further exploration would be trying to understand the relationships between categories in each one of the smells. For example, is there any relation between smells that had a larger number of associations in Descriptions and less in Memories? Why did Gasoline have so many associations in descriptions? Why did most of the smells not have many associations with activities? The ones that did were related to cooking activity.

7.2 Conclusion Some of the examples discussed in this chapter show the opportunities for using smells in the Design field. The scenarios were addressed to different professionals: Designers, Marketers, Design Researchers, and Psychologists.

109

However, it would be interesting to have them all working together in one team, a scenario where they have to communicate to each other and learn from each others’ domains. It is important that they do not assume their role as the “Designer” of the team if they have been educated in Design, for example, but rather assume a more holistic approach to the problem. This would let each professional bring to the table their knowledge but in a more open attitude to listen to others’ opinions to “their knowledge” and do not assume an expert role. Designers can learn new Research Methods and approaches with Social Scientists, while Social Scientists can learn from Designers new ways of looking at data to reveal more information and patterns, as well as relevant and useful research questions.

Collaborative scenarios are more common and different disciplines need to learn how to work with each other. More importantly, they need to learn from each other. And Design is no different. Adding more topics to Design education is not only about making Designers better practitioners or researchers, but also creating a bridge between disciplines that can be so close together such as Design and Social Sciences. It is important to build upon the knowledge not only designers have but other disciplines do too.

110

Appendix A. Exploratory Study 1 – transcriptions

Participant 1:. smell 8: Muscle liniment . P1: That was minty. I like it. I: Does it remind you of anything? P1: Toothpaste. I: Try to think about a verb for the smell. P1: I like to think of it as fresh. Freshness maybe. I: Do you remember of any emotions? P1: Not necessarily emotions, but it usually makes me feel awake, perhaps. Cause I associate that smell with toothpaste, and waking up in the morning I: One image for it P1: Maybe green, grass, the outdoors, fresh

smell 6: Nail polish remover. P1: Probably musky, I would say. I think it wasn’t bad. I: Try to comment on the smell, if there are images associated. P1: It made me think of something that has been put away for a long time, a

111

little bit of musky, but it was still not bad. So maybe something of like a family heirloom. Something that has been passed on from generations to generations. I: Any image for that? P1: Actually, whenever my mom opened her purse it was that smell. So maybe that’s what it is.

smell 7: Egyptian musk oil. P1: What is camphor? (investigator gives one example). This one again I think it was musky, but I didn’t like it very much. I: Did you remind of anything with that smell? P1: It reminded me of really bad cleaning agents, weird cleaning substances. I: So is it dirty or just bad cleaning? P1: Yeah, just kind of bad cleaning. It reminds me of like kind of a dirty mop, but they try to put like mint water or something in it, and it is still kind of dirty, so that’s what it reminds me.

smell 12: Blue Cheese. P1: Oooh… that’s weird. Oh man… what was that? It was bad smelling. I am pretty sure it is cheese, which I like but that’s just what it reminds of. Very overwhelming. It just reminds me maybe… because I related to cheese, it reminds me of fungus or something like that. Really bad smell but it probably tastes all right.

112

I: Any memories for it? P1: Probably my grandma who used to eat very stinky cheese. Yeah, maybe it reminds me like tea parties she used to have where she served crackers and cheese. She always had like really stinky ones.

smell 4: Lilac oil. P1: Floral. It was not too bad but… can I put both of them cause it is like in between. It was ok. I: If you think it is more one than the other, please make a mark. P1: It was kind of a little bit maybe sweet, that kind of flowery thing, but I think it might be that I wouldn’t maybe use. So it wasn’t bad, but it is not something that I would buy or that I can relate to me. I don’t know if I can associate with images or anything to that but it was just there. Maybe like bathroom air fresheners.

smell 10: Camphor essence. P1: I think it was both minty and also musky. That one I didn’t like as much. But this one really reminded me of like old closet put away stuff. It might be that smell that you know… if you wore something one time, and you like put perfume to go out and you know it was a nice place and you come back but you don’t wash that and you just put it in the closet. It knife of reminded me of that, just like something put away, aham… That’s all I can think of. Old closet with stuff put away.

113

smell 1: Dried crushed mint leaves. P1: That one again was minty. I need to put both (good and bad) on there cause I do like the smell of mint, but it seemed like it had something else in there that kind of contrasted a little bit, so I am gonna put like more… It just reminded me of maybe gardening and soil, just being out. I: Any other memories? P1: A little bit of the outdoors, the smell of plants and soil together. So you know, just good times outside, then camping and all that.

smell 11: Hard boiled eggs. P1: Humm, ok (making a bad face). That wasn’t very good. Kind of putrid but it still had some musky to it. It just smelled like rotten egg. I don’t know what else was in there, but that was very prominent. I: Strong? P1: I wasn’t as bad as the one I thought it was cheese for example, but it was still sort of there. So maybe, you know something, some food that has been stored for a while, just not good cleaning up the fridge.

I: Any images? P1: I don’t know if I have any other than bad food, maybe… in my house for example, sometimes when you open a faucet that hasn’t been opened for a while, it gets that smell like sulfur, and just bad. But I just imagine a container in the

114

fridge that you open and then “aaahhh! What was that?”

smell 14: Lemonade powder. I: This one is not that strong. P1: A little bit. Oh yeah.. it’s really light. I can sort of smell something. I can sort of smell a little bit now. Humm… The only thing I could smell a little bit of minty, just really really small or that and floral but just a little bit. But then, it was also maybe a little bit of acrid in a way. I: Like what? Would you compare this smell with anything? P1: It seemed a little bitter, like it seemed maybe some sort of perfume, but that… it was a little bit minty and floral, but still kind of in the bitter side. I can’t really say what it reminds me of, cause it wasn’t very strong. But I think it was confusing sort of, in a way. But I didn’t like it very much.

smell 2: Twisted peppermint oil. P1: That one did reminded me of that camphor. That an minty. I think this one reminded me a lot more of toothpaste than the other one, the first one. But not bad, I like it… minty. I think once again it reminded me sort of fresh, outdoors… I: Some specific activity? P1: Actually, maybe camping. Because although the other one really reminded me of toothpaste, I think this one reminded me of like the cheap toothpaste I used to buy to go camping. So I don’t know, it’s just like… it was a little bit not as

115

strong, maybe a little bit more towards the bitter side, but it was different. I: And do you like camping? Camping is a good experience for you? P1: Yeah, yeah. Camping and outdoors.

smell 13: Vinegar. P1: Hum, ok… Uh. That smells like vinegar. Probably acrid, and … I think that was just what it reminded me of… vinegar. Maybe, like a chemistry lab, i think. That kind of stuff that is always on beakers that you put away for a while, but just experimenting with different things. I: Did you use to go to labs and things like that? P1: Yeah… to do chemistry, yeah, all that. That’s what kind of what reminded me of. But… stronger.

smell 5: Nail polish. P1: I think this one is ethereal. Ahm… Just like alcohol, rubbing alcohol or something like that. I: But does it remind you of bad experiences, or you just don’t like the smell? P1: I think it just reminds me of any time I get hurt, perhaps, that I have to wipe my wounds or anything. Specifically one time that I was running and I fell, and I hurt my leg really bad. I had to use a lot of this cleaning stuff and that just kind of reminds of that. Just alcohol.

116

smell 3: Rose oil. P: That was a bit floral. Not too bad. I think it just reminds me of hand lotion or something like that. Not one I would buy, but just some kind of hand lotion I think. I: Do you know people who would use that hand lotion? P1: Kind of reminds me of old people again, and mane it is because my grandma had all this stuff. It is just really floral hand lotion that I don’t really use, cause I use non-scented ones but it’s just that, just hand cream or body lotion.

smell 9: Moth Balls (moth balls). P1: Uau, that’s interesting. That was definitely musky. So now this one, again, old put away stuff I think that’s just cause it’s musky. It reminds me of maybe what it smells like when a carpet or a fabric gets wet and it never really dries, like maybe a little bit moldy. That’s probably what it reminds me of, maybe just like abandoned places probably smell like that. I: Do you like musky smells? P1: Not as much. Some are ok, but this one was I put acrid on it, like kind of a little bit sharp, kind of bitter, not very good.

Participant 2: smell 12: Blue Cheese. P2: It smelled like a locker room to me. It smelled gross. Actually last night at

117

Trivia, we were next to a table of frat guys and they smelled horrible, and that smell reminded me pretty much of that. Like really like, adverse, not clean, that sort of thing. I don’t know … salty taste on my mouth. I would definitely categorize it as bad.

smell 13: Vinegar. P2: It had sort of a light smell, like kind of flowery. It wasn’t too heavy. It wasn’t necessarily my favorite smell, but I don’t think it was bad. I think it was just sort of in the middle. So I guess I will classify as a good smell. I don’t know… I don’t have any thoughts on this.

smell 4: Lilac oil. P2: I think that one was stronger than the second one, but it had more of a floral scent, and it smelled a lot like lavender to me. I like that smell a lot. I think it is nice, it is sort of calming, relaxing I: Does it remind you of anything? P2: It reminds me of like... hiking. More like being outside in the summer.

smell 8: Muscle liniment. P2: I like that smell a lot. I really like minty smells. They smell fresh, clean and smell really organized to me. When I smell mint it smells sharp... definitely organized. And that totally smelled like icy-hot.

118

smell 7: Egyptian musk oil. P2: I am not sure that this smelled just like number 3. No... number 2. Yeah... I don’t mind it. Its alright. I guess I would describe it the same way that I described the second smell. It sort of have a floral scent to it, but kind of light. Sort of mixed... It couldn’t make sense of one smell or the other. I: Does it remind you of anything? P2: Reminds me of this coffee. Ahn... Not in particular. I think it kind of reminds me of maybe being in a department store where there are a lot of smells and you cannot make sense of each one individually.

smell 2: Twisted peppermint oil. P2: Totally down with that smell too. Smells like a cigar shop, sort of what it smells like. Which is good. I like the smells of a cigar shop. So it kind of remind me of a drinking club, not like a bar. Some place nice and five star. Kind of like a gentleman’s club. There is a place in NY city called Keens, apparently one of the oldest steak houses in nation, and it smells like that place. I: And do you like it? P2: Yeah, I do like it. I think it’s a nice smell.

smell 11: Hard boiled eggs. P2: (Laughs) That’s awful. I dont know what that is, but I dont like it. It’s terrible. I feel harmed. I gotta tell the IRB.

119

I: Does it remind you of anything? P2: It reminds me of the left over food that gets left in your refrigerator and you forget about it. And it keeps getting pushed further and further back. And a couple of weeks later you pulled it out, and you’re like: “What did I do here?” then you go to throw it away. And you almost just wanna throw away the tupperware, and the refrigerator itself too because it is just bad. Yeah... it has like an awful smell like rotten eggs smell, but no... I cant really say rotten eggs cause it’s an even stronger smell than that.

smell 6: Nail polish remover. P2: (Coughs and laughs) Let’s start with the red tag on this. That smells like rubbing alcohol. But, rubbing alcohol that has been on something too long. Soaked on rubbing alcohol. I: What is rubbing alcohol? P2: It like straight up alcohol. You could buy it on the store and use it to get your nail polish, you can use it to clean things. It is a really strong smell like vinegary. It definitely reminds me of being at the hospital, or doctor’s office. Having medical procedures done. It is not a good smell. That’s my least favorite smell yet.

smell 14: Lemonade powder (The participant could not smell it) P2: Ahn... It was pretty similar to those. I caught a whip of these, of camphor. And I caught a whip of mint. But I don’t think... I feel like those aren’t the

120

right answers, like that’s not what I should be smelling. It was just ok. I will use both dots.

smell 9: Moth Balls. P2: It smells like spearmint gum. It reminds me of like... a baseball field. I feel like people eat bad food in the baseball field and then they chew gum to make that smell go away.

I: Do you think this smells like outside because of that? P2: I think I am feeling this because it’s when you are so close to somebody else in the stadium, and then they are talking to you, and you can’t help but smell this thing. It kind of smells like a fresh package of diapers. I know that sounds weird, but Jamie’s nephew, like when they come home from the groceries and I am there, it smells like the diapers that they buy. I don’t know what kind of diapers they buy. Maybe not sweet, but it has a musky smell to it.

smell 1: Dried crushed mint leaves. P2: Hum... I am confused by that smell. I guess... I don’t know how I would put that, like a... what does that remind me of? It smells like cut grass, but like dry cut grass, like cut grass in august, no cut grass right now (June). Ah... It has an outside, woody smell, like being in a pine forest with all these needles on the ground that dried out. Kind of like hay. How would I classify it...? It was ok.

121

smell 10: Camphor essence. P2: (Cough) Let me smell it again. Ahn... That has a really strong icy hot smell. It kind of remind me of vicky’s vapor rub. Which I think is a little bit different than icy-hop. I: Add what does it remind you of? P2: Being sick. Like... having a chest cold, cause that’s where you put it right. Yeah... being sick, not feeling well. SO based on my own description I will call that negative.

smell 5: Nail polish. P2: Hum... WHat does that remind me of? Can I smell it one more time? Ahn... it reminds me of a beauty salon.. Ahn... like that mixture of getting your nails done and like hair products. Wherever I would go with my mom when I grow up... thats what it smells like. I: Is it a good experience or... just something that you remember of? P2: I mean... I think it is fine, but I think it’s one of those smells that when you first smell, you kind of go (bad face), but then you get used to it really quickly, like vinegar. I think it shocks you, and then you just get used to the smell. So i dont think it is necessarily bad, but I am not sure it’s my favorite smell. But yeah... I think you get used to it pretty quickly.

122

smell 3: Rose oil. P2: That reminds me of my grandma’s garden, cause it smells like lavender. And she has a lavender bush. Anh... so it is definitely good cause it reminds me of being at her house. Because she would bring in cut lavender and then her whole house would smell like cut lavender. SO I think it’s a good smell.

Post-session conversation with Participant 2:. Experience: I rely on my sense of smell a lot. I tend to see things and smell and smell at the same time, so if it’s a situation that has a strong smell, I will associate it with that smell for some time to come, like my grandma’s house. So I think it is really important, in my opinion. The lenght of the experiment: I think it was fine, I just think that it probably could be better with the coffee, cause I know this stuff helps. If the coffee helped to clear the scent out a little bit more, because I think that my nose got confused about what I was smelling. Difficulties: No, not at all. Improvements: The coffee. I was almost thinking that you could keep it on a tin, something that is not permeable. And... I kind of like the set up (stickers and cards). I think, yeah... in these definitions, I relied on them a lot, it might be advantageous to

123

define ether, camphor and musk. Only because, I know those words, but I dont think I could understand without you explaining what they are. I wouldn’t think of a smell that goes along with them. Best order for the task: Good/Bad first, Category and then comment, but they went hand in hand cause I was talking when I was doing.

Participant 3:. smell 5: Nail polish. P3: Yeah. I really hate that smell. I mean, I think that it was nail polish remover, and it is one of the worst smells ever, it is just a necessary thing that you have to use sometimes. Not just for nail polish but, you know, other things, like taking the stickiness off of like a bottle or something. And I really hate that smell. I: Does it remind you of any experience or is it just a bad smell? P3: Ahn... Yeah. It is just the bad smell. I guess it reminds me of taking the nail polish off. Since I don’t wear nail polish very much then you know I don’t have to smell it very often.

smell 2: Twisted peppermint oil P3: I don’t really know what camphoraceous is. I: Camphor is usually that Vicks Vapor Rub has. P3: Ok. Can I smell it again to comment about it? ok. I guess it doesn’t bring up

124

any experience for me either. It is a nice smell. I: Any food or? P3: No... because it smells a little more floral to me than food normally is, I guess. I: And why do you think it is good? P3: I think it is good because it is a little bit sweet. And I think probably sweet smells are generally good smells, because everybody loves candy or most people.

smell 14: Lemonade powder. P3: That’s what you said was this, I think. I would put... if I didn’t know what this was... oh I wanna put two on there. And I think it only get bad ahn... even though it smelled a little bit more minty than those icy-hot things. But it did smell a little bit like that. So that’s why I put good because it was more minty and less of the camphor smell, but it did remind me of... SO thats interesting... It reminded me of like cough drops and thats like you’re sick which is a bad thing but it was a nice smell cause it was more minty than coff drops normally are.

smell 12: Blue Cheese. P3: The problem is that I try to think of what it is. Can I smell it again? Humm... it is really familiar smell. I am gonna put that it is good because I think it smells like food. Ahn, however, I wouldn’t want like my whole to smell like it. You know I wouldn’t want me to smell like it or something to smell like that all the time, but

125

since it smells like food, and I love to eat, that’s why I put that’s good.

smell 1: Dried crushed mint leaves. P3: Ahn... So I just really like mint in general. I: But does it remind you of anything? Images? P3: Yeah, I think it reminds me of Christmas time because that’s when I have mint the most. Although it can be that mint is different than peppermint, maybe, so thats more summery like you put, like Kelly made that mint and pineapple that was really really tasty. So I think if it’s peppermint minty it is Christmas but if it is... I think this is more summery minty. Like in Mojito, although I don’t drink it very much.

smell 13: Vinegar. P3: I like the stickers thing. Ahn, can I smell it again? Ok. It is not strong as I expected to be, although I think it is vinegar. And, so... I guess the memory with vinegar as I only really use vinegar for cooking, I guess sometimes for cleaning, although I still think it is bad smell. It is just necessary for cooking, so it is a necessary evil almost. I: And you don’t like the taste? P3: No I don’t really like the taste. Ahn... only if is in like,... if it is mixed, like this balsamic vinegar with the food that’s good but I feel like this was not the good kind of vinegar, like white vinegar or something, which I don’t like the taste.

126

smell 11: Hard boiled eggs. P3: Ow... (Bad face) I don’t know what that is but it is bad. I want to put this on on here too (previous one). Ahn... I don’t know what that is, and I don’t know... it doesn’t bring up anything, but I thin it is bad.

smell 4: Lilac oil. P3: Ok. I think it smells like clean sheets, and I love clean sheets, especially when they were hanging out on the line to dry. I: SO do you think this is an outside smell, like because they are outside? P3: No, I think that this smells more like the soap that you would use, but then when I think of... I think this for some reason specifically smells like sheets, and not like other clothes. Ahn, but maybe it is like the combination of the soap that you use and then being hanging outside that makes it smell like this. But yes... I love the smells of sheets hanging out on the line.. My mom’s linen closet smells like this. I don’t know why cause like in the winter time she doesn’t hang sheets outside. to dry, but I love the smell of my mom’s linen closet, and that’s what it smells like. I wish I had a linen closet and it smelled just like that, but I don’t.

smell 7: Egyptian musk oil. P3: I am gonna want to smell that one again, so keep it out. Ahn, can I smell it again? It is a little bit of this, just very slight though. Since it is so slight, it wasn’t too bad, but I think I will still put bad. And I guess that why it reminds me of

127

being sick, that’s just why I put bad. You know what also smells like is that lip balm stuff in the yellow tube (Carmex), I don’t remember what it is called. I: Burt Bees? P3: Not it is a different one in a yellow tube with a red cap, like bright yellow. Burt’s bees is a little light. I don’t know what it is called... Carmex! I: So it is a sick thing? P3: Yeah.

smell 9: Moth Balls. P3: That’s similar... Let me smell it again. This is kind of hard. I will still clip this, and this... Ahn... Still was sick. I don’t know why they all reminded me of that smell on being sick. I: Does it remind you of being outside, inside? P3: Ahn... This one... I’d say more inside. Yeah.

smell 10: Camphor essence. P3: I think that this smells ahn... like in herb, or something like that, like a plant outside that like if you were trimming it you would get that kind of smell out of it, so ahn... and I like doing garden work so that’s what it reminds me of, but it is being outside and in the garden. Not like a flower garden, not like a vegetable garden.

128

smell 3: Rose oil. P3: Ok. This one... This one also reminded me of clean things, so maybe like the sheets again. But also it smelled more like vanilla which I really like. I: How would you describe this smell? P3: I’d say it’s ahn... like really like vanilla but a little bit sweet. Clean. But that’s not any floral... (looking at her classification) But I guess that’s ..ok... So I chose floral because that’s the closest thing to describe the smell of vanilla that are on the choices over here. And I saw in my head like a... I don’t know... I had this soap that have a vanilla bean with the flower, like next to it, like the flower from the vanilla plant, I suppose. And that’s why I chose floral. My nephews love that smell. Every time they come over, they have to wash their hands... they are obsessed with washing their hands. And they are like... “Where is the flower soap?” “It is still there, I just took the sticker off.”

smell 6: Nail polish remover. P3: Ow... (Bad face) That one smells like polish remover too, and I think like you asked me before, the experience from the... other one I thought that smelled like nail polish remover, I think I just, since I don’t wear nail polish… I don’t wear nail polish very much cause I hate dealing with it, like… I like the colors but I hate taking it off. It is just like a torture. So that’s why I don’t like that smell.

129

smell 8: Muscle liniment. P3: Ok… That smells exactly like those pink little candies. Are they spearmint or something?… Maybe not spearmint… it’s a little minty but it reminds me when I was a kid my grandmother was in a nursing home and this woman down the hall from her alway had these candies and when we walked by she yelled at us from her room to come in and give us these candies. I don’t really like them very much… I mean, every once in a while just because it’s a fond memory to go visit my grandma but every once in a while I eat them, but when I was a kid I did not like them, so.. I hated going in there, but you know, you have to like… mom said: “go get some candy. just take it”. So, even though I don’t like the thing, it is kind of a nice memory to me, so that’s why I am gonna put the green on there.

Post-session conversation with Participant 3:. Experience: I thought it was good. Not a lot of smells brought a specific memory back for me, so some of them was kind of hard. But the ones like that last one… that one was instantly, I know what that smell is, I know what this is. I think it is because some of the smells are kind of maybe not everyday smells, but I’ve smelled them many different times, so it is hard to pinpoint one particular experience for me. The lenght of the experiment: No, I think it was fine Difficulties:

130

No,… there were just a couple that, since they were less pungent, and it was kind of hard to smell it, but since you can smell it a few times, I think it was fine. Improvements: Maybe these words… I think it is a pretty good range, but this, since you had to explain it to me, I think it is a very specific type of smell, and the rest of them are a little more broad. Like ethereal and floral...

Participant 4:. smell 14: Lemonade powder. P4: (Coughs) It reminded me... I am from the country, West Virginia grabble roads,... and it reminded me of my grandma’s house. I don’t know why... like the woods, like the... I don’t know. I don’t know what it is. I: Outdoors or indoors? P4: Indoors. Like... something in her house, something she used to have, I don’t know what. I: Was it a good experience? P4: Aham.

smell 8: Muscle liniment. P4: I don’t consider this on bad at all, because it has a distinct smell, because I was a student athlete for the majority of my life, and when I smell that, that meant that the relief was coming. Before, like track meats, after track meats, you

131

rub icy-hot or whatever and out it on my leg, when I was bad, and had cramps on my leg. That was the relief. (Laughs) I: So it reminds yo of physical activity? P4: Yeah. Definitely.

smell 1. P4: Reminds me of chewing gum. Mints on chewing gum. And that was equated to good times, passive like kicking there, hanging out, relaxing, even maybe social. So ... I: Does any image come to your mind? P4: Like iceberg, like cool, almost like “briskal” breezy. And it maybe even spark relaxing, like outside, chilling on a hammock.

smell 13: Vinegar. P4: Uhhh... (and move head away). Let me smell it one more time. (Coughs and laughs). Ok. This is not a good one. I: Why is it not a good one? If you had to describe the smell to someone what would you say? P4: It is not a soothing smell, it is like a pungent, like odor distinctive. It almost like repels you. It is like arghh... you know like... It feels like it is up in my head right now. I related to like, almost like garbage. I dont know why. Related with garbage or landfills, or chemicals or whatever. I don’t know. It is just not very

132

positive to me. Not a nice minty lotion...

smell 10: Camphor essence. P4: What is that? It is almost like minty or... I don’t know what it is but it is like soothing smell, almost reminded me of the whole icy-hot thing a little bit. Like a soothing, relaxing, like... I: How is it different than the icy-hot? P4: It is not as strong. In my opinion it is not as strong. It is like, almost like, if icy-hot is, i think about that smell, reminded me of track, something is injured or whatever. That almost like a vapor, vicks vapor rub. That’s what reminded me. Putting in your chest and comes up a healing type of thing. Little minty but not as strong. I: So the experience that you recall is also healing, but is it also good? P4: Yeah. It is not bad. Definitely good.

smell 6: Nail polish remover. P4: (Bad face) Yeah, I need to smell it again. I: Is it good or bad? P4: I don’t know. I did not feel a good thing about that one. I don’t know. It is like on the border. I have to figure out which way it is gonna go. I: Why would it be bad? Why would it be good? Try to think about it. P4: You know, the first thing that came into my mind was calamine lotion. In

133

calamine lotion, in essence is, supposed to be soothing and healing, to feel better whatever of a skin irritation, but it reminds me of when I was a child and... when I was a kid I could look at poison ivy and get it. And it was the most irritating horrible thing whatever but the calamine lotion make me feel better. But the experience of the smell was still associated with... calamine lotion is associated with the torment of poison ivy, even if the calamine lotion was the relief, I still think of it as negative or whatever because the experience that is associated with it. So, it smells really close to that or whatever. Bad! I am saying bad only because the relief was good, but what is directly associated with it is very painful, even though it was a relief. I think the painful experience precedes that it was the smell of relief. If that’s what I was smelling.

smell 2: Twisted peppermint oil. P4: It almost smells like peppermint. It has some kind of mint to it, but more like peppermint. I don’t know. It makes me feel like I am smart. I: Why is that? P4: I don’t know, just like... maybe because when I used to do standardized tests, they used to give us peppermint. And they said: peppermint stimulates your mind or whatever, you will do better in tests or whatever. So whenever I smell peppermint, it reminds me of coolings, like I am cool, relaxed, FOCUSED. Whenever I smell peppermint I feel very focused. Plus, I always have peppermint with me when I go out on dates whatever too, so... I feel cool, I feel relaxed. I feel

134

smart, and focused. I: Is it good or bad? P4: Definitely good, cause I want to be focused and smart. And that’s my inspiration... to be cool, smart and on top of my things.

smell 3: Rose oil. P4: Reminds me of some type of lotion. Like a ... I: Is it good or bad? P4: It is good. It is the type of lotion that you get that has the minty, not minty but at the bottom they have the herbal flower, herbal leaf, but it is really strong for really dry cracked skin or whatever, I think about that. I think about... in winter time, my skin gets extra extra dry, super dry, like horrible, and I gotta go get the extra,... the bottle with the leaf one or whatever, so it has extra treatment for my hands and my skin or whatever. This is what that reminds me. I: Winter? P4: Winter. Cause winter call but also really dry skin and the special type of lotion that I use that is a relief.

smell 11: Hard boiled eggs. P4: (Lifting eyebrows and laughs). BAD! Images? I: Adjectives, words, verbs? P4: Funky. Not ... I don’t know. I felt nothing good in this. Just kind of putrid,

135

kind of... I don’t know. I: Any experiences? If not that’s fine. P4: I can’t think of any. Something that has been left out, kind of food. Decomposed, yeah.

smell 12: Blue Cheese. P4: (Coughs and laughs) Oh God! Awful! I will put two of these red dots. Awful! It is like foul. Musky. Decomposing, putrid. Oh it was so bad. It was like... I: How would you describe the smell? P4: The greatest breeze from a landfill. (Laughs). Terrible. My stomach turned a little bit. I don’t know why it was so bad for me.

smell 4: Lilac oil. P4: Some of these are so similar. That reminds me of some type of lotion too, like herbal, like lotion, cream, soothing though, soothing smell, relaxing. it is a lot like... It is not the best though. It is not a bad smell. It is not very different from some of the other ones. I felt like a lot of them are like stuff that... lotions companies and perfume companies are using in order to make their stuff smell like soothing, relaxing. Definitely that. And I feel like they are using the smell mint a lot because something about minty, fresh is supposed to make someone feel relaxed, ahn... mental state relaxed, like a state of mind. Feeling good for someone else. I feel that it is more prominent in certain types of lotions that more

136

women use, because in our society, the overwhelming minty, like refreshing smell is characteristic of like feminine. So, if I say... usually when I say minty, they all have like a very minty, feminine, you know, smell. As the opposite, maybe like masculine or musky.

smell 7: Egyptian musk oil. P4: I don’t know what that is. I: Do you want to smell it again? P4: Yeah. I: Do you like it? P4: Yeah. I dont feel bad about it. Just I don’t have this thing reminding me of anything. I: How would you describe it then? Try to borrow words from other senses, like from taste or hearing. P4: It is not quite yet bitter, but it is knocking on the door to being bitter, it is almost there. It is on the... not perfumy, but... still relaxing, but on the border of relaxing almost like it is something some smells are relaxing and some smells wake you up. It is like right there, between wake you up and almost being bitter.

smell 5: Nail polish. P4: This gets the big red dot. Kind of reminded me of... musky ahn... like. I: Do you want to smell it again?

137

P4: Yeah. I dont know what smell is, but it is not like stink old, I am about to die smell or whatever. But ugh... I dont have a memory associated with it. I: And a description. P4: Almost like... it is a smell that grabs you, and... courses you in the wrong way, not like a positive minty pleasant feeling, but like a negative, not pleasing to your senses, not something that you have in your house or on a regular... ah... but it is not garbage. It is not yet garbage, it is not garbage. Just like, yeah... I don’t know.

smell 9: Moth Balls P4: I don’t know it makes me think of moth balls, or something like that, you know. Moth ball you put on your closet to keep moths away. That’s what kind of remind me. Not very pleasant. But I associated with that. Just something that I would find in my closet. Little balls, or whatever, from my childhood or whatever, I found sometimes. I: Do you still use it? Do you still have this smell around? P4: No. But I can distinctively remember that from when I was a child. So, finding them and smelling them or whatever. It is amazing what smells you remember or whatever. It would be kind of musky I guess, just like pungent, but it is not gonna kill you though. You are not gonna pass out, but it is not something that you would want to formulate into your house. But definitely indoors, closet.

138

Post-session conversation with Participant 4: Experience: It was fine. It wasn’t bad at all. You had it set up well. I knew what I was going to do. The lenght of the experiment: It was longer than what I expected. Uncomfortable in any way? No. The only discomfort that came when I smelled something bad. It shocked me or whatever. The minty stuff, and that comes down to the smells, senses, the minty or the good is inviting but then the poor negative is repelling, so the only uncomfortable that ever happened was when repelling feeling that I got my body jerk back, is not normally accepted. Difficulties: Understanding definitions and classifying the smells. Some of the smells were very close together making it more difficult, and I did not have anything more to say. Improvements: Some of the words are more in the intellectual side, but you had this laid out here. Best order for the task: I had the best order.

139

Activities smoking playing working eating shopping washing - cleaning dancing camping driving baking cooking gardening relaxing drinking healing sleeping - waking up growing up running - swimming hiking decomposing socializing hiding

Descriptions sharp - dull clean - dirty beautiful - ugly rough - smooth strong - weak feminine - masculine large - small healthy - sick new - old complex - simple pungent - bland pleasant - painful fresh - stale natural - artificial progressive - regressive fun - boring active - passive hard - soft valuable - worthless good - bad heavy - light rich - poor

Sensations safe - unsafe minty - musky - floral - fruity refreshed - weary stimulated - unexcited rational - intuitive wet - humid - dry anxious - relieved bright - dark bitter - sweet - sour - salty fast - slow happy - sad focused - unfocused hot - warm - cold violet - amber - beige - brown yellow - blue - red - green - pink constrained - free confused - organized calm - agitated hungry - thirsty black - white ornate - plain clear - hazy

Memories

grandma - grandpa

city - farm

babyhood - childhood

forest - garden - mountains

breakfast

indoors - outdoors

time: _____________________

teenage years

home

people: ___________________

summer - winter - spring - fall

salon - hospital - beach - landfill

adulthood - elderly

kitchen - bathroom - bedroom

christmas - easter - thanksgiving

gas station

birthday

places: ____________________

mother - father

morning - afternoon - evening

vacations - holiday

friends - girlfriend - boyfriend

Smell #1: What first comes to your mind?

p1

Appendix B. Final Experiment Answer Sheet

140

Appendix C. Responses to the Final Experiment Questionnaire

College Rank:. freshman

10

sophomore

13

junior

18

senior

34

graduate

9

not a student

6

Area/Department:. Aerospacial Engineering

1

African-American and African Studies

1

Anthropology

1

Arabic

2

Art

2

Art & Tech

1

Art Education

3

Art/Music Performance

1

141

Art/Painting

1

Business

3

Business, Finance and Accounting

1

Cinematology

1

CMRS

1

Communications

11

Computer and Information Science

1

Computer Engineering

1

CSE

1

Dance

3

Dental Hygiene

1

Design

13

Electrical Engineering

1

Engineering

1

English

7

Enrollment Services/Editor (job description)

1

Finance

1

Fine Arts

1

French

1

German

1

Hebrew

2

History

4

142

History of Art

1

Humanities

1

Industrial Design

3

Integrated Biomedical Science

1

Interior Design

2

Italian

1

Journalism

2

Linguistics

6

Marketing

1

Mechanical Engineering

1

Molecular Genetics

1

Nursing

1

Painting and Drawing

1

Philosophy

1

Photography

1

Political Science

1

Psychology

1

Public Affairs

2

Russian

1

Spanish

2

Special education teacher (job description)

1

Statistics

2

143

Strategic Communications

2

Theater

1

Visual Communication Design

4

Age Range:. 18-24

68

25-30

14

31-36

6

37-40

0

older

2

Gender:. Male

32

Female

58

Where were you born?. Abilene, TX, USA

1

Akron, OH, USA

3

Amberg, Germany

1

Ankara, Turkey

1

Athens, OH, USA

1

Barrington, IL, USA

1

144

Berlin, Germany

1

Burnsville, MN, USA

1

Carthage, MO, USA

1

Centerville, OH, USA

1

Chicago, IL, USA

1

Cincinnati, OH, USA

3

Cleveland, OH, USA

4

Coldwater, OH, USA

1

Columbus, OH, USA

17

Coshocton, OH, USA

1

Dover, OH, USA

1

Elyria, OH, USA

1

Euclid, OH, USA

1

Evanston, IL, USA

1

Fairview Park, OH, USA

1

Gallipolis, OH, USA

1

Gomel, Belarus.

1

Grove City, OH, USA

1

Houston, TX, USA

1

Hudson, OH, USA

1

Ithaca, NY, USA

1

Kansas City, MO, USA

1

145

Kaohsiung, Taiwan

1

Kettering, OH, USA

1

La Grange, IL, USA

1

Leominster, MA, USA

1

Lima, OH, USA

1

Madeira, OH, USA

1

Manchester, NH, USA

1

Mansfield, OH, USA

2

Martinsburg, WV, USA

1

Medina, OH, USA

1

Minneapolis, MN, USA

1

Mt. Vernon, OH, USA

1

Nanjing, China

1

North Canton, OH, USA

1

Orlando, FL, USA

1

Parma, OH, USA

2

Piqua, OH, USA

1

Pittsburgh, PA, USA

2

Port of Spain, Trinidad & Tobago

1

Sandusky, OH, USA

1

Santa Ana, CA, USA

1

Santa Monica, CA, USA

1

146

Seoul, South Korea

1

Shady Grove, MD, USA

1

St. Petersburg, Russia

1

Sumter, SC, USA

1

Sylvania, OH, USA

1

Taipei, Taiwan

3

Tashkent, Uzbekistan

1

Tiffin, OH, USA

1

Topeka, KS, USA

1

Waterville, OH, USA

1

Westerville, OH, USA

1

Zanesville, OH, USA

1

Where did you grow up?. Abilene, TX, USA

1

Akron, OH, USA

3

Albion, NY, USA

1

Amberg, Germany

1

Barrington, IL, USA

1

Bellaire, OH, USA

1

Bevea, OH, USA

1

Bexley, OH, USA.

1

147

Bowling Green, OH, USA

1

Brookeville, MD, USA

1

Brooklyn, NY, USA

1

Carthage, MO, USA

1

Centerville, OH, USA

1

Cincinnati, OH, USA

2

Cleveland, OH, USA

2

Columbia, SC, USA

1

Columbus, OH, USA

14

Coshocton, OH, USA

1

Dover, OH, USA

1

Dublin, OH, USA

2

Euclid and Akron, OH, USA

1

Export, PA, USA

1

Fairview Park, OH, USA

1

Gahana, OH, USA

1

Gallipolis, OH, USA

1

Gambier, OH, USA

1

Garretsville, OH, USA

1

Grafton, OH, USA

1

Grove City, OH, USA

1

Houston, TX, USA

1

148

Hudson, OH, USA

2

Istambul, Turkey

1

Kaohsiung, Taiwan

1

La Grange, IL, USA

1

Lancaster and Vincent, OH, USA

1

Lebanon, OH, USA

1

Lima, OH, USA

1

Logan, MN, USA

1

Manchester, NH, USA

1

Mansfield, OH, USA

1

Mason, OH, USA

1

Medina, OH, USA

1

Mt. Vernon, OH, USA

1

New Milford, CT, USA

1

North Canton, OH, USA

1

Olmsted Falls, OH, USA

1

Orlando, FL, USA

1

Parma, OH, USA

1

Perrysburg, OH, USA

1

Philadelphia, PA, USA

1

Pickerington, OH, USA

3

Piqua, OH, USA

1

149

Russia, Israel, Louisiana, Ohio (11)

1

Saint Louis, MO, USA

1

Sandusky, OH, USA

1

Seoul, South Korea

1

Strongsville, OH, USA

1

Sunbury, OH, USA

1

Sylvania, OH, USA

1

Taipei, Taiwan

1

Taipei, Taiwan/Cupertino, CA, USA

1

Tiffin, OH, USA

1

Waterville, OH, USA

1

Wellston, OH, USA

1

West Chester, OH, USA

1

Westerville, OH, USA

2

Wilmington, OH, USA

1

Zanesville, OH, USA

1

Do you somke?. Yes, regularly.

2

Yes, socially.

8

No, never smoked.

70

No, but used to smoke.

10

150

Are you allergic to anything?. Yes.

19

No.

71

If yes (allergic), to what?. Penicilin

3

Pollen

3

Bee stings

2

Cat

2

Codine

2

Seasonal

2

Aspirin, Ibuprofen,Lorabid

1

Eggs, Milk

1

Hazelnut, Grass

1

Medications

1

Mushroom

1

Mold, Pet hair, Everything under the sun

1

Sulfa drugs

1

Dust

1

Grass/tree pollen, and cats. No more symptoms.

1

Blue Cheese

1

151

What is your favorite smell?. Coffee

9

Vanilla

9

Grass

5

Rain/After rain

5

Baking/cooking something

4

Fire (bonfire, campfire)

4

Cinnamon

3

Cookies

3

Fall

3

Flower

3

Coconut (beach related)

2

Lilacs

2

Pine trees

2

Pumpkin/Pumpkin pie

2

Bath and Body Works “White Citrus”

1

Bergamotte/Ralph Lauren

1

Bread

1

Brownies

1

Citrus

1

Clove

1

Coconut Lemon Verbana

1

152

Cologne

1

Easter Lilies

1

Febreeze Laundry

1

Flowers (rose, hydrangas)

1

Food in the oven

1

Food smells

1

Fresh

1

Fresh scents

1

Fruit

1

Fruity fragrance

1

Fruity smells

1

Good perfume on a girl

1

Home cooked chinese food

1

Husband’s cologne

1

Laundry detergent

1

Lavender

1

Lemongrass

1

Marzipan

1

My old house in La Grange

1

Nature (water, grass, trees, etc)

1

New car smell

1

Ocean, Sea

1

153

Old Book Stores

1

Oranges

1

Popcorn

1

Spearmint, Garlic

1

Spring (flowers blooming/rains frequently)

1

Strawberry

1

Sweet and warm smell

1

Swimming pool

1

Thanksgiving dinner

1

The river smell when white water kayak

1

The smell of Cajun Food

1

The smell of the blanket I had as a baby

1

Warm vanilla sugar from Bath and Body Works

1

How sensitive are you to smells? 1 (Not sensitive)

1

2

10

3

31

4

42

5 (Very sensitive)

6

154

Appendix D. Grouping Demographic Information – Final Experiment

Department:. Art. Fine Arts, Painting, Drawing, Art Education, Photography, Dance, Art History, Theater, Art & Tech

Business. Marketing, Business, Finance and Accounting

Communications. Coomunications, Strategic Communications, Journalism and Cinematology

Design. Industrial Design, Visual Communications Design, Interior Design

Humanities. Anthropology, Philosophy, African-American and African Studies, History, History and Public Affairs, Political Science, Medieval and Renaissance Studies

155

(CMRS), Psychology

Languages/Linguistics. English, Russian, Spanish, French, Hebrew, German, Italian, Linguistics

Medical Science. Integrated Biomedical Science, Nursing, Dental Hygiene, Molecular Genetics

Technology. Aerospacial Engineering, Mechanical Engineering, Computer and Information Science, Electrical and Computer Engineering, Statistic

Origin:. Ohio: Akron, Bellaire, Bevea, Bexley, Bowling Green, Centerville, Cincinnati, Cleveland, Columbus, Coshocton, Dover, Dublin, Euclid and Akron, Fairview Park, Gahana, Gallipolis, Gambier, Garretsville, Grafton, Grove City, Hudson, Lancaster and Vincent, Lebanon, Lima, Mansfield, Mason, Medina, Mt. Vernon, North Canton, Olmsted Falls, Parma, Perrysburg, Pickerington, Piqua, Ohio (11 years old - but also Russia, Israel, Louisiana), Sandusky, Strongsville, Sunbury, Sylvania, Tiffin, Waterville, Wellston, West Chester, Westerville, Wilmington, Zanesville

156

Outside Ohio. Illinois: Barrington, La Grange Minnesota: Logan Missouri: Carthage, Saint Louis

South Carolina: Columbia Florida: Orlando Texas: Abilene, Houston Maryland: Brookeville Pennsylvania: Philadelphia, Export New Hampshire: Manchester New York: Albion, Brooklyn Connecticut: New Milford California: Cupertino, CA, USA (9 years old but also Taipei, Taiwan)

Outside USA. Taiwan: Kaohsiung, Taipei South Korea: Seoul Turkey: Istambul Germany: Amberg

157

Appendix E. Re-coding to Annotated Words – Final Experiment

Words that participants added to the answer sheet are listed below. To be able to count these words, it was necessary to group them, since people mentioned similar concepts using different words. Words with asterisk reprent a selection not listed on the asnwer sheet that was added by participants. Words in bold represent the group in which the other words were placed. Dentist*. my dentist dentist dentist office

Morning. early in the morning* mornings making* = morning + cooking

Girlfriend. girl who made me lemon squares* girlfriend’s house*

158

Garden. flower garden*

Adulthood. adults* early adulthood*

Evening evenings eating* = evenings + eating

Mother. mom cleaning* = mother + cleaning mom’s storage closet* = mother + closet parent’s closet* = mother + father + closet

Summer. summer afternoons = summer + afternoons

Afternoon. summer afternoons = summer + afternoons

159

Grandma. grandmother’s kitchen* = grandma + kitchen grandpa+grandmaa’s house* = grandma + grandpa grandparents’ basement* = grandma + grandpa + basement grandma’s house* grandma’s*

Grandpa. grandpa+grandmaa’s house* = grandma + grandpa grandparents’ basement* = grandma + grandpa + basement

Kitchen. grandmother’s kitchen* = grandma + kitchen under kitchen sink*

Babyhood. baby* babies*

Yard*. yard* backyard*

160

backyard at home* = yard + home lawn

Elderly. old people* Coffee shop*. Panera bread*

Childhood. children* Childhood Ice Cream Parlor* = childhood + Ice Cream Parlor get out of the car as a child to smell* = car + childhood Childhood friends = childhood + friends 10 years old* childhood cookies*

Car*. in the car* in car* get out of the car as a child to smell* = car + childhood

161

Bathroom. public restroom

Farm. farm in Minerva* = farm + Minerva Home. current apartment backyard at home* = yard + home living room at home* = living room + home

Friends. Childhood friends = childhood + friends best friend* my best friend* an old friend* classmates* best friend, Morgan* Kathy* Jessi* Jason H*

162

Living room*. living room at home* = living room + home

Basement*. grandparents’ basement* = grandma + grandpa + basement friend’s basement* Auto Shop*. AutoZone* lube shop* car shop workers* car shops* Sick (under descriptions). In sickness* (under memories)

Evening. Night time* night* bedtime* School*. elementary school* = school* + childhood Jr. High* = school* + teenage years

163

Teenage years. Jr. High* = school* + teenage years teenagers* 16 years old*

Dinner*. dinner time*

Working (under activities). when I was working in a Italian restaurant* (under memories) = working + Italian restaurant* work* (under memories)

Theater*. Movie Theater*

Holiday. celebrations*

Summer. summer camp dining hall* = summer + camp dining hall*

164

Pool*. poolside* public pool*

Garage*. Uncle Mike’s garage* = uncle + garage

Outdoors. Outside*

Neighbor*. my neighbor’s house*

Salon nail spa* hair salon*

Mall*. Polaris* shopping mall*

165

Christmas. Christmas time baking* = christmas + baking

Candle shop*. Dad’s candle shop* = father + candle shop*

Aunt*. my aunt*

Roommates*. roommate* roommates Stef*

Sister*. sister’s house* sisters* my sister*

Drinking. drinking tea* drinking juice* drinking orange juice*

166

Cooking. mornings making* = morning + cooking

Cleaning. sanitizing* mom cleaning* = mother + cleaning

Eating. evenings eating* = evenings + eating

Swimming. learning to swim*

Washing. washing clothes*

Baking. Christmas time baking* = christmas + baking

Humid. moist*

167

Amber. orange*

Green. mint green*

Red reddish burgundy

Cold cool*

Beige. cream* light beige

168

Appendix F. Calculating significant words’ threshold value

Calculating the threshold value for a word to be significant considering the total numbers of subjects, no subgroups, so n = 90:

The sample statistics p-hat = # of choices made on a word /# of subject in a particular group = m/n; z = (p-hat- p_0)/ sqrt(p_0(1-p_0)/n), n is the # of subject in that particular group (total subjects in this case).

If p-hat is larger than 1.64*sqrt(p_0(1-p_0)/n) + p_0,, i.e, m is larger than (1.64*sqrt(p_0(1-p_0)/n) + p_0) * n, then we can say this is m a significant number (significantly larger than 0).

So, for word set size three: p_o 1/(3*22) that equals p_o = 1/66, so if m > 3.26, i.e., if any word listed.

Threshold values were calculated for each one of the word set sizes and then defined for subjects into different groups.

169

Appendix G. Selection of Significant Words

First number represent the amount of times the word was selected. Number under parethesis represent the wordset the word was in.

170

171

172

173

174

175

176

177

178

179

180

181

182

183

184

Bibliography

Barbara, A. (2010, March 26). Space, Time and Scent. Lecture presented at Headspace 2010: On Scent As Design in Parsons, New York. Retrieved August 31, 2011, from http://www.youtube.com/watch?v=nc4iV1Za4y8

Bone, M., & Johnson, K. (2009). I miss my pencil. San Francisco, CA: Chronicle Books.

Burfield, T. (2000). Top Middle and Bottom Notes: The End of the Road for a Theory? The Magazine Page. Retrieved August 31, 2011, from http:// www.users.globalnet.co.uk/~nodice/new/magazine/top_middle_and_ bottom_notes.htm

De Wijk, R.A., Schab, F.R. and Cain, W.S. (1995) Odor identification. In Schab, F.R. (ed.), Memory for Odors. Lawrence Erlbaum Associates Inc. Mahwah, NJ, pp. 21–37.

185

Dröscher, V. B. (1969). The magic of the senses; new discoveries in animal perception. New York: Dutton.

Edwards, M. (n.d.). Note News: The Fragrance Making Scents Right Now [Interview by S.]. In Sephora. Retrieved August 31, 2011, from http:// www.sephora.com/browse/tt/index.jhtml?categoryId=C8123&tt=notes

Edwards, M. (2011). The Fragrance Wheel [Digital image]. Retrieved August 31, 2011, from http://www.fragrancesoftheworld.com/the-fragrance-wheel/

Ehrlich, S. D. (2009, September 7). Aromatherapy. Retrieved August 31, 2011, from http://www.umm.edu/altmed/articles/aromatherapy-000347.htm

Engen, T. (1982). The perception of odors. New York: Academic Press.

Fox, K. (n.d.). The Smell Report: An Overview of Facts and Findings (Rep.). Retrieved August 31, 2011, from SIRC website: http://www.sirc.org/ publik/smell_human.html

Gilbert, A. N. (2008). What the nose knows: the science of scent in everyday life. New York: Crown.

186

HEADSPACE 2010: On Scent As Design Part 1 | The New School - YouTube [Video]. (2010, April 7). In YouTube - Broadcast Yourself. Retrieved August 31, 2011, from http://www.youtube.com/watch?v=0XJnOEpMUok

Heise, D. R. (n.d.). The Semantic Differential and Attitude Research. Indiana University. Retrieved August 31, 2011, from http://www.indiana. edu/~socpsy/papers/AttMeasure/attitude..htm

Jönsson, F. U., Tchekhova, A., Lönner, P., & Olsson, M. J. (2005). A Metamemory Perspective on Odor Naming and Identification. Chemical Senses, 30(4), 353-365. Retrieved August 31, 2011, from http://chemse.oxfordjournals. org/content/30/4/353.full#cited-by

Kumar, A. (2006, September 22). Ntering the realm of branding through Senses: The Phases in Branding History. Management Canvas. Retrieved August 31, 2011, from http://www.managementcanvas.iimindore.in/ icanvas/index.php?option=com_content&view=article&id=86:ente ring-the-realm-of-branding-through-sensesthe-phases-in-brandinghistory&catid=34:marketing-and-branding&Itemid=56

187

Ladd, G. T., & Woodworth, R. S. (2006). Elements of physiological psychology: a treatise of the activities and nature of the mind, from the physical and experimental points of view. Whitefish, Mt.: Kessinger.

Lawless, H., & Engen, T. (1977). Associations to odors: Interference, mnemonics, and verbal labeling. Journal of Experimental Psychology: Human Learning & Memory, 3(1), 52-59.

Leary, M. R. (2008). Introduction to behavioral research methods (5th ed.). Boston: Pearson/Allyn and Bacon.

Lee, P. (2011, February 6). Design Research: What Is It and Why Do It? | Reboot. Reboot. Retrieved August 31, 2011, from http://thereboot.org/ blog/2011/02/06/design-research-what-is-it-and-why-do-it/

Lindström, M. (2010). Brand sense: sensory secrets behind the stuff we buy. New York: Free Press.

Lynch, P. J., & Icke, I. (2007, March 4). Olfactory system [Human Olfactory System]. Retrieved August 31, 2011, from http://en.wikipedia.org/wiki/ File:Olfactory_system.svg

188

Norman, D. (2010, November 26). Why Design Education Must Change. Core77: Industrial Design Magazine & Resource. Retrieved August 31, 2011, from http://www.core77.com/blog/columns/why_design_education_must_ change_17993.asp

OECD Better Life Initiative. (n.d.). Retrieved August 31, 2011, from http://www. oecdbetterlifeindex.org/

Osgood, C. E., Suci, G. J., & Tannenbaum, P. H. (1971). The measurement of meaning. Urbana: Univ. of Illinois Press.

Paulkner, E. (1965). Success and Failure of Odor Classification as Applied to Reactions to Erogenous Odors. Journal of the Society of Cosmetic Chemists, 16(9), 515-526.

Peterzell, D. H. (n.d.). Latest, Greatest book on Olfaction [Review of Learning to Smell: Olfactory Perception from Neurobiology to Behavior]. Retrieved August 31, 2011, from http://www.amazon.com/Learning-SmellOlfactory-Perception-Neurobiology/dp/0801883687/ref=sr_1_1?ie=UTF 8&qid=1314758236&sr=8-1

189

Rodriguez-Gil, G. (2010, September 1). The Sense of Smell: A Powerful Sense. Retrieved August 31, 2011, from http://www.tsbvi.edu/seehear/ summer05/smell.htm

Ross, S., & Harriman, A. E. (1949). A preliminary study of the CrockerHenderson odor-classification system. American Journal Of Psychology, 62(3), 399.

Sanders, E. B. (2006). Design Research in 2006. Design Research Quarterly, 1, 1-8. Retrieved August 31, 2011, from http://www.maketools.com/articlespapers/DesignResearchin2006_Sanders_06.pdf

Sanders, E. B., & Liem, A. (2011). The Impact of Human-Centred Design Workshops in Strategic Design Projects. MakeTools. Retrieved August 31, 2011, from http://www.maketools.com/articles-papers/LiemSanders.pdf

Sanders, Elizabeth B.-N., and Peter Kwok Chan. “Emerging Trends in Design Research: Changes Over time in the Landscape of Design Research.” International Association of Societies of Design Research, Hong Kong, 12–15 Nov., 2007.

190

Shedroff, N. (1994). A Unified Field Theory of Design. Nathan Shedroff’s World. Retrieved August 31, 2011, from http://www.nathan.com/thoughts/ unified/19.html

Significance in Statistics & Surveys. (n.d.). Creative Research System. Retrieved August 31, 2011, from http://www.surveysystem.com/signif.htm

Smith, K. (n.d.). All About the Color Blue. Sensational Color. Retrieved August 31, 2011, from http://www.sensationalcolor.com/color-messagesmeanings/color-meaning-symbolism-psychology/all-about-the-colorblue.html

Wilson, D. A., & Stevenson, R. J. (2006). Learning to smell: olfactory perception from neurobiology to behavior. Baltimore: Johns Hopkins University Press.

Zimmer, C. (2010, May 17). The Brain: The First Yardstick for Measuring Smells | Senses | DISCOVER Magazine. Discover Magazine. Retrieved August 31, 2011, from http://discovermagazine.com/2010/may/17-the-brain-firstyardstick-for-measuring-smells

191