Observational and Survey Methods n

Observational Research – Description of behavior in a specific setting • “Natural” behaviors

Observational Research n

Wide variety of different settings and techniques

n

Can vary along a number of dimensions – Naturalistic observation -- Participant observation

• “Natural” settings – Human observers used to record/code behavior

– Naturalistic setting – Structured setting – Qualitative data vs. Quantitative data

– Concerns: • Observer bias / Objectivity • Subject reactivity • Control

Naturalistic vs. Participant Observation n

Amount of observer interactions with those being observed

n

Naturalist Observation – Little or no interaction

Naturalistic Observation n

– Observe with little or no contact with participants – Attempt to minimize subject reactivity • Concealed observations • Habituation

– Observations made from “outside” the group n

Naturalistic Observation

Participant Observation – Observer becomes member of group being observed

– Advantages: • Less subject reactivity -- More likely to observe natural behavior • Better chance of remaining objective – Disadvantages • Often little or no control over situation

Participant Observation n

Participant Observation – Observer becomes a member of group under study • Concealed or unconcealed – Interact as member of the group – Advantages • Insights not otherwise available • Can question and interact with group members • Feeling of being group member – Disadvantages • Influence / influenced by group • Maintain objectivity?

Naturalistic vs. Structured Setting n

Goal of observational research – Observe “naturally” occurring behavior – Naturalistic Setting -- Observations take place in “observee’s” natural environment • Advantages – presumably seeing what occurs normally • Disadvantages – may need to wait for low frequency behaviors (small amount of data) – Lack of control

1

Naturalistic vs. Structured Setting n

May use a more structured setting / situation – Create opportunity for low frequency behaviors – More control over situation

Qualitative vs. Quantitative Data n

Data from observational research can take many forms

n

Qualitative Data – Narrative record – Verbal description of what was observed

– Response to a particular situation – Still allows for participant’s natural responses n

– Can provide rich information – Complex situations • More things happening than observer can process • Objective measures?

Continuum – Depends on the types of questions being asked – More or less natural / structured

Qualitative vs. Quantitative Data n

Quantitative Data – Devise ways to quantify behavior being observed

– May focus on one or a small number of behaviors – Address specific question rather than describe all facets of behavior

Behavior Checklists & Coding Schemes n

– Record of when or how often behaviors occur – Coding Scheme – operational definitions of what constitute a “behavior”

– Frequency – Duration n n

– Intensity or degree Might use Behavior Checklists & Coding Schemes

• Description/criteria of specific behaviors (identifying characteristics)

Increase objectivity – Data obtained not dependent on who did recording

• Pretested to find ambiguities • Observers trained to recognize consistently – Should increase objectivity

– Can compare observations from different people • Interobserver reliability (% agreement) n

Allows for inferential statistical tests – Compare behaviors across situations, groups, etc.

Observation and Sampling n

n

Often – too much happening for observer to record everything Will sample to reduce the amount of information – Event sampling – only record some of the events – Time sampling – only record during certain epochs

n

Identify specific behaviors of interest

May use videotapes to record behavior

Coding System n

Method of Quantification – Frequency Method • Count of behaviors in given time period – Duration Method • Length of time for each behavior – Intervals Method • Divide time into discrete units • Record (main) behavior within time unit

– Decreases processing demands on observer – Allows multiple observers to code – Can be reanalyzed for different behaviors • Apply different coding schemes to same behaviors

2

Survey Research n

Gather data using surveys or questionnaires

n

Direct way to gather information from people Types of information:

n

– Attitudes and Beliefs • Are you happy being a student at Geneseo?

Strengths and Weaknesses n

n

• Do you believe you are receiving a quality education? – Demographic • What is your G.P.A.? – Behaviors (self report)

Strengths: – Only way to gather some information (attitudes) – Can gather a lot of information – Can ask / gather specific information Weaknesses – Accuracy of responses • May not be aware of some information about own behavior • Social Desirability – Respond in such a way to make themselves look good

• How often do you miss class? • How many hours per week do you spend studying • Are you helpful?

Question construction n

Not easy to write quality questions

n

Potential Problems – Simplicity / Clarity

Simplicity (words understandable) n

New York’s American Museum of Natural History -Surveyed public interests. – 77% interested in plants and trees

– Double-barreled questions – Loaded questions

– 39% interested in botany

– Negative wording – Yea-saying and Nay -saying

– 48% interested in fossils – 39% interested in paleontology – 42% interested in rocks and minerals – 53% interested in geology

Simplicity (clarity of wording)

Double-Barreled Questions

n

n

Roper Starch -- Poll of Americans in 1992 – 20% believed that the Holocaust never happened – 12% were not sure

n

“As you know, the term Holocaust usually refers to the killing of millions of Jews in Nazi death camps during World War II. Does it seem possible or does it seem impossible to you that the Nazi extermination of Jews never happened?”

n

Gallup -- 1992

n

“Do you doubt that the Holocaust happened, or not? – 9% doubted truth of the Holocaust

Do you believe that SUNY should lower tuition and decrease the salary of their faculty members? (Yes/No) – Asking two questions at once – People may want to respond in different ways to the two questions

3

Negative wording n

Do you believe that SUNY should not raise tuition (Yes or No)

Yea and Nay Saying n

People may respond the same way to all questions – Yes or positive to all – No or negative to all

– Can be confusing • Yes -- against • No -- for

– Can detect by constructing question such that consistent responses are unlikely – Do you believe that SUNY Geneseo is a good value? – Do you think that you pay too much for your education?

Types of questions n

Closed vs. Open Ended

Types of questions n

– Closed -- set of response alternative to choose from – Open -- participant free to supply own answer

Rating scales -- typically 5-7 points – Geneseo needs more nightlife – Strongly agree __ __ __ __ __ __ __ Strongly Disagree

– Strengths and weaknesses of each? n

Semantic Differential Scales

n

F&G: How would you rate this Uniform? – Good __ __ __ __ __ __ __ __ Evil – Strong __ __ __ __ __ __ __ __ Weak

Sampling n

Need people to take survey

n

Typically can’t test entire population – Population: All individuals of interest to researcher; everyone to which they would like to generalize results – Ex: all college students; all adults, all voters in Michigan

n

Must test a subset of population

Sampling and Confidence n

If you test entire population -- know exact values

n

– 17% of high school seniors have driven after drinking Can’t do -- must estimate using sample – Will be some error in estimate (e.g., 16%) – Sampling Error / Margin of Error – How much the estimate may be off – Decreases with increased sample size • Sample more likely to be representative of population

– Sample: set of individuals from population n

Goal: Sample representative of Population

4

Sampling Techniques n

Probability vs. Nonprobability sampling

Probability Sampling n

Simple Random Sampling

– Probability Sampling: each individual -- known probability of being sampled – Nonprobability sampling: probability not known

– Every member of population has equal chance of being selected – Need to know entire population & system for selecting n

n

Stratified Sampling – Break population into strata (groups) – Randomly sample within strata Cluster Sampling – Randomly select groups

Nonprobability Sampling n

Convenience sampling

Sampling Activity n

– Individuals selected because they are readily available

Determine what percentage of books in a given bookstore (e.g., Borders in Henrietta) have the following: – Female author(s) only – Male author(s) only – Both Female and Male authors

n

Sampling Techniques – Simple random sampling – Stratified random sampling – Cluster sampling – Convenience sampling

From Book n

Survey Methods (Pros and Cons) – Interviews – Phone Surveys – Electronic Surveys – Written Surveys

5