FACILITATING ORGANIZATIONAL KNOWLEDGE WORK THROUGH WEB INFORMATION SYSTEMS

FACILITATING ORGANIZATIONAL KNOWLEDGE WORK THROUGH WEB INFORMATION SYSTEMS ,\Ki IhqrESTIG,-\TION OF THE INFOmfATTON ECOLOGY AND INFOItIJLATION BEHiVT...
Author: Sophia Wade
2 downloads 0 Views 17MB Size
FACILITATING ORGANIZATIONAL KNOWLEDGE WORK THROUGH WEB INFORMATION SYSTEMS ,\Ki IhqrESTIG,-\TION OF THE INFOmfATTON ECOLOGY AND INFOItIJLATION

BEHiVTOURS OF L'SERS IN X TELECOhEiTC.-IICATIONS C0hfl'A~W'

Bxïan Dedor

-A thesis subrnitted in conformity with the requlements for the degree of Doctor of Philosophy

Faculty o f Information Studies University of Toronto

O Copyright by Brian Detlor 2000

1+1

National Library d cana,

BMiotheque nationale du Canada

Acquisitions and Bibliographie Services

Acquisitions et services bibliographiques

395 Wellington Street Ottawa ON K1AûN4

395. rw Wellington ûttawaON K 1 A W CaMda

canada

The author has granted a nonexclusive licence allowing the National Library of Canada to reproduce, loan, distribute or sel1 copies of this thesis in microfom, paper or electronic formats.

L'auteur a accordé une licence non exclusive permettant à la Bibliothèque nationale du Canada de reproduire, prêter, distribuer ou vendre des copies de cette thèse sous la forme de microfiche/fllm, de reproduction sur papier ou sur format électronique.

The author retains ownership of the copyright in this thesis. Neither the thesis nor substantial extracts fiom it rnay be p ~ t e or d othenvise reproduced without the author's permission.

L'auteur conserve la propriété du droit d'auteur qui protège cette thèse. Ni la thèse ni des extraits substantiels de celle-ci ne doivent être imprimés ou autrement reproduits sans son autorisation.

FACILITATING ORGANIZATIONAL KNOWLEDGE WORK THROUGH W E B INFORMATION SYSTEMS AAN I M E S T I G A T I O N O F THE INFORhLITION ECOLOGY AND INFORhMTION BEHAVIOURS OF USERS IN A TELECOh~fUNICATIONS COMPANY

Doctor of Phdosophy, 2000 Brian Detlor F a d t y of Information Studies Lrniversity of Toronto

ABSTRACT This dissertation outlines a n exploratory research investigation of Web information systems OVIS), such as departmental intranea, corporate portals, and the World Wide Web, and th&

capacin- to facilirate orffanizational knowledge work-a

set of knowledge creation, distribution,

and use processes in a-hich information pkys a prknary role. -1s a mrans of uivestigating dus phenornenon, a case smdy eraluation of Web infomiation

systcms usage was conducted on five major sets of users at a large telecommunications Company. Emphasis was placed o n esarnining the information needs and uses of participants

and the information ecology Li which Web information systems were ualVied. Data collection involrcd a u-ide rariety of techniques, namely in teniews, Web tracking, field obserr-ations, document review, and questionnaires.

A

modified version of grounded theory sen-ed as the

prima? mode of data analvsis; descriptive statistics helped tgangulate research resulrs.

Field work hdings suggest that W I S d o support knowledge work, but in a limited way:

primanly for know-ledge creation only; and to a lesser estent knowledge dismbution and use. Different WIS are utilized for knowledge work in varjing ways to resolve diverse problem siniaaons and support disparate infomiation behavioun. Information politics, the \VIS derelopment process, and the organizaaon's information culture impact the use of WIS for knowkdge work. Furrher, a person's hnctional role, general motivation to leam, perception and interest in WIS technology influences the estent to which WIS are îdopted and u&ed.

Findïngs also suggest that to fosrer knowledge work, WIS need to: support the tailorability and organization of information; deliver relevant, reliable and timely information; provide collaborative tools; and promote user engagement witb these systems. The central contribution of this research is a mode1 of Web use showing the organïzauonal,

user, and interface factors that promote or inhibit the use of Web information systems for knowledge work. Moreover, these factors CO-existin mutual bi-directional reiationships with one mother. Implications iaclude the need for k

s to pay attention to the salient atmbutes of these

shaping factors w-hen consuucting Web înfosmation systems to help organizationd participants create, distribute, and unlize knowledge more effecavely.

ACKNOWLEDGEMENTS .As a P h D . candidate and iong-distance nuuier, I have corne to the conclusion that compleclig a doctoral dissertation is in many respects quite similv to ninnlig a marathon. Both require a certain degree of sramina, preparation, and d power to see oneself through successhil completion of the task at hand. However, what most people do not realize is that ncither is a solitaq venture. Each necessitates the support of others to guide one's way dirough the process and this dissertation has been no esception. Throughout the last four years, 1 have receked the gracious support of many coiieagues, &ends, and familu, and it is to these inciiridualç that I gratcfdy acknodedge conmbution to this ha1 product. First, I would Wce to thank the members of my dissertation committee, Drs. Chun Wei Choo, Ethel -%uster, and Andrew Clement, ail of the F a c e of Information Studies at the Cniversity of Toronto for thek patience, concern and insightfd comments, thar in the end, made this disscrtaâon more rigorous and tvodwhile than it othenvise could have been. Thanks also gocs to Dr. Pierrette Bergeron fiom the École de bibliothéconomie et des sciences de l'information, L.niversité de hlonuéai, for sening as an ertemal esamlier.

To Dr. Choo, 1 would like to estcnd my heartfelt thanks for his esteemed guidance and support over the many years that I have known him. As mj- Ph-D. supen-isor, Dr- Choo went bcyond the call of duty, providing me with ample opportunities for grou-th and professional derelopment. Xlore i r n p o m t l y , his never-ending starnina and demand for high qurility work sen-ed as an excellent esample for a wide-eyed apprentice like myself to s t m leamlig the crafi o f academe. Thank FOU,Dr. Cho-for

ereq-thing.

To Dr. ;\uster, 1 would iike to extend my thanks for her u n w a r c ~ ghelp and disceming s just the a h t guidance, especiaiiy o r e r this past y.Her candid and up-hont s ~ e s t i o n were mcdicine and sen-ed me weU at difficult rîmes. N o one couid have asked for a more suppomve

personal champion. 1 thaali you whole-heartediy. To Dr. Clement, in addition to his contributions o n my dissertation, 1 would Wre ro

acknowledge his influence in 6rst inuoducing me to the concept of user-centered design and helping m e recognize the importance of the role of users in the srstem development process. This \.as pivo ta1 in shaping my research direction.

T o other members and players of the Faculry of Information Studies, 1 acknowledge th& support and conmbution co this project. S p e d thanlis go to Dr. Joan Cherry for her helpfd adrice in how to conduct rigorous, quaiity rcsearch. To Drs. Lyme Howarth and Debra Wallace, 1 am indebted for rheir practicai suggestions, good humour, and for inrolving me in the

Iinowledge Management Insunite fiom which 1 have learned so much. T o my fellom doctoral students, 1 ha\-e valued the^ camaraderie and many words o f encouragement. F d y , to Ross Barclav, I appreùate the use and derelopment of the WebTracker software that =as utililed in this dissertaaon. GracefÙlly, I d s o would like to acknowledge Bcli Unk-ersity Labs at the University of Toronto for rheir contribution and hancial support for this dissertation as a Beli LTnivcrsiq-

Lahs initiatire. A notable thank 'ou

goes to hLike huron for his continuous support and

championing of the project from its earliest d q s , as weli as Dennis Venerus, lTince Pensato,

Eugcne Roman, and Gary Hopkins for helping me gain access to the studp sites. Gratitude also gocs to the anonymous participants who fieel. gave up tirne in their busy scheddes to take part

in the smdj-, as weii as rnembers of Beii Canada's IS/IT Usabiliq- Team, namely Janet Curris, Frank Stam, and Stanley Wu, whose help Li recruiting participants and scheduling inten-iews grcatly facilitatcd the operational aspects of the Mi study.

To the faculty at the Mïchael G. DeGroote School of Business, hlchiastcr University in Hamilton. Ontario, 1 thank h e m for being excited about this dissertation. I look fonvard to

rnany h t u r e years w o r k g together on rclated research projects. O n a personal lerel, there are man.

indisiduals that desen-e special recognition. =\

distinctire thank you goes to mu colieague, Don Tutnbuil, whose candor and personal kiendship wcre a delightfuI, positive influence. Don's help in presenting p a p a at various academic conferences and accompanirnent in the dam collection process on a related research project made these msks more enjoyable. T o r n v nt'o real-life marathon running partners, Martin Clarkson and Clément .\rsenauit, 1 also owe manv words of thanlis. hfam- pauistakingly Listened to mv trials and mbulations on a majorin of our runs together. It was estremeIy refreshing to get an esternal perspective on the whole dissertation process, and 1 thank him for bis estended interest in my work. Clément uniquely sen-ed in both capacities as a m e r and fellow doctoral coileague. His fiiendship has bcen steadfast and our 'cornpetition' to cornplere o u dissertations before one another motinted

mc to work much harder and faster than 1 w-ould have done otherurlse. 1 could not have asked for a more cornpetent race pacer.

To Joanne Johannink, 1 owe an honest 'thank )-ou' for encouragmg my pursuit of doctoral smdies. \Vithout hcr insistence, 1 wouid never have embarked on a Ph-DT o mv brothcr, Bruce Dctlor, and sisters, Liz Detlor, b t h v Anstruther, and Diane Farrell, 1

thank them for their words of encouragement and motivation. A special thank ?ou goes to mv nephews, Matthew and Mark Bird, for a p p r e c i a ~ gth& uncle's challenge. T o m . parents, Lome and Julie Detlor, 1 am eterndy indebted. They have always been thcrc for me and this dissertation has prored no esception. 1 am especially gratcful to them for tcaching me the value of w o r k g hard and the importance of doing one's best. Thesc lessons wcrc invaluable and sen-ed me \-enweli throughout the entire dissertation process. Thank !ou

Mom and Dad for your constant and unconditional support.

Last, and most importantly, 1 wish to thank my son, Ben Detlor, for being so wonderfùi and sharing a large part of the doctoral process. Ben patient. accompanied me o n nurnerous

occasions to the Universiq and kept me grounded throughout the entire four years of study. It is to Ben that 1 uuly dedicate this wotk. He was there through thick and thin. Xlarathon or not, 1 certainly could not have asked for a better support team. It made

complethg the dissertation a less painhl adventure, and bg Car, a more rewarding ex~erience.1 had a top-notch crew to help pare the way through the arduous process of devising, planning, conducting, and writing a Ph.D. thesis. Thank you team for helping me cross that finish line. 1 ccrtainly could not have done it without ?ou.

B.D.

TABLE O F CONTENTS LIST OF FIGURES AND TABLES..................................................................................................... LIST OF APPENDICES....................................................................................................................... GLOSSARY OF ACRONYNS ............................................................................................................

CHAPTER 1: INTRODUCTION

s

xx

uti

....................................................................................

1 . 1 Problem Staternent ........................................................................................................................ 1.1.1 IE"hat an Web in/onndon yrtcrrms? ........................................................................................... 1.1.2 [Vhyst@ 1Veb iinjomation gxtemx?.......................................................................................... 1.1.3 IV73 ar fiow/edge Management /Ipp/iraion~............................................................................ 1.2 Research Objective ........................................................................................................................ 1 -3 Scholarly and Practid Significance ......................... . . ............................................................... 2.3 Outline of Dissertation .................................................................................................................

CHAPTER 2: LITERATURE REVIEW

17 17 18 20

-7-7

24 24 3' -3

AND CONCEPTUAL FRAMEWORK ........27

2.1 Introduction.................................................................................................................................... 27 ................................................................... 28 2.2 Knowledge and ICnow-ledge Work ................ ...... 2.2.1 Knowiedge................................................................................................................................... 28 7 2 . 2 Knowiedse Work ........................................................................................................................ 33 2.3 Web Information Systems as Enablers of Kaowledge Work ............................................... 38 2.4 Bamers to \VIS Adoption and Use for Knowledge Work ...................................................... 4 4 2.5 The Concepnial Framework: The Beharioural/Ecological Mode1 ........................................ 48 2 Y.1 It&zvatioti Ecoiogzer, Pmbieem Situations, and Inforution BehaMozm ...................................... 32 2 5.2 The IVorkngx of the Bebavioura// Ecoiogicui Mode/.................................................................... GO 2.6 Rescarch Questions ..................................................................................................................... 66 2.7 Conclusion ...................................................................................................................................... 69

CHAPTER 3: METHODOLOGY ...................................................................................70 3.1 Introduction.................................................................................................................................... 70 3.2 Grounded Theory.......................................................................................................................... 76 3.3 Case Study Site and Research Tirneline...................................................................................... 85 3.3 Pilot Study....................................................................................................................................... 88 .. ....................................................................................................................... 8 7 3.4.1 Pilot Partznpanr~~ 3.4.2 Metbodr Used............................................................................................................................. 87 3.4.3 Lenniitrgsfmm the Pi& ............................................................................................................. 97 3.5 Pull Smdy ........................................................................................................................................ 98 .. 3.5.1 FI,// JIU+ Pon'zcpontS. .............................................................................................................. 78 3.5.2 Entry Mto the F d S t d y Site ................................................................................................. 3 9 3.5.3 &le,hodologicai Dz@emncer/mm the P i h ................................................................................... 104 3.6 Rehted Studies ............................................................................................................................. 108 3.7 Ethical Concems.......................................................................................................................... 1 11 3.8 Research Accurac)r....................................................................................................................... 112 3.9 Conclusion .................................................................................................................................... 1 15

CHAPTER 4: FINDINGS .........................

. .

. . ...........

4.1 Introduction.................................................................................................................................. 116 4.2 Information Needs and Uses ..................................................................................................... 119 4.2.1 Mqor Jtts of U.serz..................................................................................................................119 4 . 2 2 iV7SActiuify ...........................................................................................................................12i 3.2.2.1 Cornorate Portal .................................................................................................... 130 4.2.2.2Departmental Intranets......................................................................................... 142 4.2.2.3 \Veb-based Grou~ware......................................................................................... 153 4.2.2.4 World Wide Web ................................................................................................... 163 4.2.3 K y W7.F fiowi~dge 1Vork1................................................................................................... 173 4.3 \'due-Added Processes............................................................................................................... 188 4.1 Information Ecology ................................................................................................................... 207

4.5 Conclusion ....................................................................................................................................

CHAPTER 5: DISCUSSION AND CONCLUSION

230

..................... ................................ 231

5.1 Introduction.................................................................................................................................. 231 5.2 Answers to Research Questions ................................................................................................ 233 ........................................................................ 233 5-71 PmbIem Sifivations und Injomation Behuuioz~r~ j.22 I-/ma/ion Ecofogy................................................................................................................ 2-10 ......................................................................................................... 5 - 2 3 Valre.Ad.d Prvcesse~ 2 4 2 7 . 2 4 Gmrp~ of uim........................................................................................................................ 3 3 5.3 Descriptive hlodel of \VIS Cse............................................................................................ 2 4 7 5.4 Reflcctions on Practice ...................................... . . ................................................ 252 5.5 Rcflections on hiethodology ...................................................................................................... 253 5.6 Reflecûoas on Theory ................................................................................................................. 256 3 ./ Limitations .................................................................................................................................... 258 5.8 Future Research ........................................................................................................................... 262 5.9 Conclusion .................................................................................................................................... 264 r

REFERENCES .........................

....................................................................................267

LIST OF FIGURES AND TABLES Tables Table 2.1. Ut-eniew of the Components of the Behavioural/EcoloS;cal Mode1............... . . ....64 Table 3.1. Timeline of Events .......................... . . . . .......................................................................... 87

Table 3.2. Esample of Tracker Log Data ............................................................................................ 93 Table 4.1 : Cornputer and Web Background of Participants

..........................................................

121

Tablc 4.2 Significant Episodes by W S Tupe: Breakdown Withm Functional Groups ............. 127

Table 4.3 Sigaificant Episodes by W S Type: Breakdown by Problem Type ............................. 130 Table 4.4. Portal-Related Sigrilocant Episodes: Predominant boa-ledge IVork Patterns ........ 132 Tablc 4.5: Portal-Reiated S w c a n t Episodes: Breakdown of fiowledge Work Activities by Functional Group........................................... 133 Tablc 4.6. Portal-Related Signifiant Episodes: Predominant Problem Dimension Patterns ....134 Table 1.7: Portal-Related Significant Episodes: Breakdown of Problem Dimensions bu Functional Group ............................................................ 136 Table 4.8. Portal-Related Sigmficant Episodes: Predominant Information Trait Patterns .........137 TabIc 1.9: Portal-Related Sigmficant Episodes: Brcakdown of Information Traits by Functiond Group................................................................. 139

Table 1.10. Portal-Related Significant Episodes: Predominant Information Use Patterns ........139 Tablc 4.1 1: Portd-Related Sigmficant Episodes: Brcakdown of Information Uses bu Functiond Group .................................................................. 141 Tablc 4.12: Intranec-Rclated Significant Episodes: Prcdominant fiowledge Work Patterns ........................................................................................... 143 Tablc 3.13: Intraoet-Related Signïficant Episodes: Brcakdown of Knowledge Work Activities by Functional Group............................... ............ 145

Table 4.1 4: Intranet-Reiated S i g d k a n t Episodes: Predominant Problem Dimension Patterns .......................................................................................146 Table 4.1 5: Intranet-Related Significant Episodes: 1 4 8 Breakdown of Problem Dimensions by Functional Group ...~...........................--..............-........ Table 4.1 6: Innanet-ReJated Sigarficant Episodes: Predorninant Informaaon Trait Patterns ...........................................................................................

147

Table 3.17: Intranet-Related Signifiant Episodes: Breakdown of Information Traits by Functional Group................................................................. 150 Tablc 4.1 8: Inuanet-Related Signifiant Episodes: Predominant Information Use Patterns.....151 Table 4.19: Intranet-Related Significant Episodes: Breakdown of Information Uses by Functional Group ..........................-.

- . - . - . . - . - . - .153 -..

Table 4.20: Groupware-Related Significant Episodes: ............................................................. Predorninant I Preferences); the second Lirolved lengthening the number of days that pages in the history Eiie would e-xpire to cover the duration of the m o n i t o ~ gperiod (the default settïng was one day). This contrasted with the more lengthy process of configuring WebTracker on individual participant machines which entailed about ten minutes of downloading and installation t h e . The third change was that addi/iona/ intemews were cond~ctedin Ihej.ï.ï rtz@ than in /hr pifut. Since the fidl studr was clmed out in association Mth BeU Canada's IS/IT L'sabiiity team, a member

from that team administered a separate inten-iew Mth each participant, asking questions aliich concentrated on the concerns and problems participants had with using a rariety of office productivitv took such as e-mail and hlS Office. Since pamcipants would be discussing their k , researcher aIso attended these esperiences with using cornputers in their da. to day ~ ~ o r the usabilin- interviews, taking notes and paying particulv attention to any comments partiapants made on their Web use activity. In addition, furrher ad hoc inteniews were conducted with

managerial staff from the Communications group responsible for managing the content on the corporate portal and the information technology group which was in charge of the hances and architectural infrastructure o f the portal. These ad hoc intenievrs werc vrangcd wide the rcsearcher \vas in the field and leamed of the importance of these nvo groups' roles in the design and implementation of the portal in the organization. Some of these inteniews were tapcd and

transcribed; others were recorded using field notes. The fourth change aras in fbe wayje/d ob5emafionswere camed UML Obsening people engaged in n-ork practice prored more difficult than esperienced durlig the pilot smdy. \Vith the pilot, the

researcher found obsen-ations to be easiiy conducted since ail the pa-rticipants were Erom the

same functional group working side by side; as a result, the researcher could sit at a nearby desk and observe the participants unobûusiveiy. This was not the case with the fidl study where participants were scatrered across two office towerç at a centrai location. Though some participants sat close to one another, the rnajorit). worked in isolation from others. The rcsearcher did not have access to nearby desks so observations had to be conducted through visits to the participants' cubicles. This caused pamupants to feel compeiied to engage in conversation wvith the researcher rather than continuing normal work behatiour. Plus, for about half of rhese \-kits, a member of the company's Usabiiity team accompanied the researcher. When this occurred, many pamapants stopped nrorking altogether. Though many insights and

ncw idcas wcre elicited through these ad hoc infornial tisîts, the method itself drffered from the more subtle field obsenrations conducted during the pilot study. Note that there was minimal bias in having the member of the company's Usabiiit). team accornpany the researcher during panicipant observations. This member was a junior analyst and knew many of the participants from his prior role in the Company as a technical support worker. -4s such, participants were cornfortable with

his presence and forthcorning with answers to his

questions. Furrher, this poreneal bias was reduced over t h e as the junior analyst becarne

invoh-ed in othcr projccts and did not accornpany the researcher during field visits. To rninLnize the risk of having participants perceive the researcher as a consultant or an estension of the LTsabilitytearn, the researcher explained and rc-iterated on different occasions (i-c., during brichg sessions and at commencement of interview sessions) that the researcher was a Cniversity researcher and the purpose and intent of the study. Also, as the project

progrcssed, participants becarne more farniliar with the researcher; for example, trust was sucngthcned through ad hoc encounters during coffee and lunch breaks. Furthemiore, p o t e n d bias in participant response was reduced through uianguiation of data collection methods.

3.6 Related Studies The rnethodological procedures uulited in this research estend and complement the methods used in rehted studies. There have been other smdies concerned nith the use of IT tooh for knowledge work, Web information systems, and the information environment of users. Howel-er, most tend to restrict themselt-es to the use of qualitaave methods, primariIy concentrating on certain ethnographie techniques, such as intenden-s and/or field observations. -.\ few studies have utilized WIS transaction logs or Web usage statistics to complement their use

of qualitative methods. In general, such smdies seem to provide a more holistic understanding of their phenornenon of study. It was hoped that a similar approach would strengthen chis dissertation's fïndings by identifjing both the estent and contest sunounding organizational Web use for kriowledge work. For instance, several studies were presented recendy in a special track on knowledge

management at the 33rd Hawaii International Conference on Systems Science. These tended to rcly stnctly on quaiimtive data collection methods. A few u&ed

grounded theory or a x-ariant

form of content andysis to anal\-te the data. Only one combuied the use of interviews nith log

file analysis. Of interest are studies bv Snis (2000), Sherif and hiand\lwalla (2000), Stenmark (2000), and Robertson, Sorenson, and Swan (2000). These studies wexe found by the researcher

to bc insightful and thorough at expIo&g and gainuig insïghts on a relatively new area of study: thc use of IT for howledge work. Snis (3000) conducted a field investigation of how a group of knowledge workers managcd

knowledge in their daily work; the goal of the study was to draw consequent implications for the design and use of IT-based knowledge management processes. The group sped;ili+ed in quality assürancc of pharmaceutical hedth care products and comprised eight participants, most of whom wcrc chernical engineers. Snis conducted the smdy over a four month period utilizing

many of the same qualitative methods as this dissertation: semi-scnictured interviews, participation at various meetings, direct observations, and ad hoc inteniews wïth key informants. Likewise, she used a similar approach in examining the collected data across key knowlcdge processes: knowledge creation and development, knowledge organizing, knoa-ledge distribution, and knowledge usage. Sherif and Mandviwalla (2000), utilizing a multiple case study approach, conducted thirt)-thrce semi-suucnired interviews with kev informants and performed document renew on a \-arien- of archiva1 material in their study on the factors affecting the adoption and use of IT to

support organizational knowledge structures. Like & dissertation, d interviews and field notes wcrc uanscribed and loaded into QSR NUDIST (the forerunner of QSR W i v o software) and

analyzed using grounded theor). methodolog).. Stcnmark (2000) studied a large group of knowledge workers at an IT service subsidiaq\vichin TTol\-o.The workers comprised technicians, system developers, content protiders, and administrators. Csing a mixture of data coilecaon instruments s i m h to this dissertation, namely semi-suucnired intemiews, e-mail questiomaires, and transaction log files, the purpose of the

investigation was to esamine the estent to which an intranet recommender system prototype could facilitate the shaxing of tacit knowlcdgc arnong organizational workers. Robertson, Sorensen, and Swan (Robertson, Sorensen, & Swan, 2000) conducted an interpreave case study on 15 organizational workers in a consultancy 6rm uglizing sernistructured intemiews to investigate how groupware technolog)., such as Lotus Notes and c-mail, could facilitare knowledge creation. Similar to this dissemaon's use of grounded theory to

analyzc the information ecology of the organization and its impact on knowledge work activity, Robertson et al. utilized content analysis on their interview data focusing on the structurai, cultural, and social conditions which supported or constrained the process of knowledge

creation. In t c m s of the feu- studies m e n *

a~ailableo n inaanet development and use, these tend

to utilize many of the same techniques outlined in this dissertation. For instance, Abraham

(1998) used both interviews and questionazires as a method of e i i c i ~ gresponses fiom a rariety of intranet stakeholders across 17 organizations. Of ioterest is a study by Lamb (1999), t ~ h o foliowed a three phase schedule in her intranet study. The &t was cross-sectional in nature where she relephoned 200 different organizations from four selected industries to queq- their intranet status. The second was comparative where she conducted a field study at four organizational sites using direct e&ation

of intranet content and usage statistics,

obsen-aâons of intranet use, semi-strucnired inteniews, and document collection. T o analyze this information, Lamb utilized constant comparative methods ftom grounded theo.

to

formulate data categories, cross-case patterns, and theoretical ieads. The third was traversal in naturc, where Lamb conducted addiaonal interviews nith hrm contacts. Though there have been relaavel- few past studies that hare concentrated o n organizational information environments, one of the more recent and notable is that by Rosenbaum (1996)".

In his work which inrestigated the information use environments of managers, Rosenbaurn relied o n semi-structured inten-iew-s t o coliect his data, and constant comparative anaiysis methods ro gain an interpretive reading of the data. One of the limitations of the study noted by

Roscnbaum is the rehnce on intelî-iews as the single data coiiection method. In responsc, Rosenbaum comments on the need for other methods of data collection to help triangulate rcsearch results (as was done in this dissertation) and suggests that his work could have becn strcngthened had the case study site @-en permission for hirn to conduct direct observations.

'"

Rosenbaum's dissertation cntided "'rianagcrs and Information in Organizations: Towards a Strucnrntional Concept o f the Information Use Entironment o f Ahagcrs" won the -4merican Soaet). o f Information Saencc (,ISIS) 1996 Doctoral Forum =iwxrd.

3.7 Ethical Concerns Sreps were d e n to protect the prk-acy, confidentiality, and anonpity o f the partiapants. Prior to data collection in both the pilot and hill study portions of the project, an inilai b+Çng mas conducted with participants at the organizational site. The five participants from the pilot and die 6v-einitiaiparticipants from the Ml smdy attended f o d briefing sessions organized by the researcher for each of thesr two groups. Indiridual briefing sessions were conducted with the ten other participants from the €distudy as the- joined the research project. At all of these

sessions, participanu were told the nature and purpose of the study, the tasks to be performed, and the procedures to be foiiowed (refer to Appendix F for the partiapant o l - e ~ e wsheet). Participants were aiso infoxned of the v o l u n t q nature of theu participation, their right to withdraw consent and their right to d i s c o n ~ u eparticipation at any time aichout penalty. They

werc asked to sign a consent form s t a ~ gthar the? understood these conditions (refer to

r\ppen&x G for the participanr consent form). Funher, at the end of each of these b r i e h g sessions, participants were P e n the opportunity to ask any questions the. might have. Prior to individual inten-iews,pmi~n'onwa.r asked /o tape mcord conc'eriatzonr.

In rems of cornputer-monitored data, participants had the oplion o / t ~ t t n i nIhe ~ We6 friding of and ikzvitzg /hr i~ümationLeing fracked With respect to IVebTracker, pamcipants could disable and enable the tracking br single clicking a radio bunon on the main WebTracker interface. They could also preriew the current dafs Web t r a c b g acavity through a pull-down menu within WcbTracker, or could peruse aii Web transaction logs by accessing the test 6les contained in a separate \VebTracket directon7 resident on thek local hard drives. Further, participants were rcminded thar rhey were being a c k e d dirough the presence of an icon displayed at the bottom

of their deskrop screens. With respect to Netscape's history tracking mechanism, pamapants

could disable monitoring (if they so chose) by either resetting the number of days that pages in

the history List espired or dearing the history list; both options are available under Netscape's Prcfcrcnces menu item. Participants were askcd to inform the researcher if the histor). list was rnodified Li an)-waf7. hlso participants were shown how to preview the contents of the &tory File in Netscape through an option (Conaol-H) under Netscape's Tools menu item-

Ali participant data were kept confidentid and anonymous. The data were coiiected

= such

a wav as to prevent any association benveen coiiected data and partiapant names. Inteniew data,

tracking logs, and questionnaires were coded with identificaaon numbers to group collected data

on individual subjects. To prevent linkage to participant names, consent forms did not include

identification numbers. In terms of reportkg of results, any reference that involved a pa-rticular participant was presented in a mariner to protect the individual's identity. Since the research ïnvolved ïuterr-iewing and observùig people in their work enlitonment, an ethical review was necessary to ensure that subject recruitment, methodoIogical procedures, intcrriew questions, and consent f o m s conformed to University of Toronto research guidellies

and that participant confidentdit). and anonymity were maintalied. In fact, n~-O ethical reviews wcrc conducted and approred to ensure that participant privacy, confidentiaii~,and anon)mity were maintained. The hrst was perfonned by thc FIS Human Subjects Review Cornmittee prior

to commencement of the pilot study; the second was conducted by the Oftïce of Research

Sen-ices at the University of Toronto as a prerequisite to securing BUL support prior to the start of the f d study.

3.8 Research Accuracy Ensuring the accuraq of the depiction of a phecomenon is a necessity in qualitative rcsearch (Buchwald, 1999, p. 45; Krk 8t Miller, 1986; Lofland & Lohnd, 1995). Cresweli (1991,

p. 157) observes that though "qualitative researchers have no single stance o r consensus on

37

in acruaiity, no participant cleared or reset theu history hle during the data collection process.

addressing traditional topics such as validity and reliabihty in qualitative studies", more recent reports from qualitative research investigators such as Lincoln and Guba (1985) and E+landson, Harris, Skipper, and M e n (1993) have established criteria such as trustwoahiness and authenticiq- to counter the need to relate more traditionai saentifïc criteria to procedwes utilited in quaIitative research. For instance, Lincoln and Guba (1985) protide a List of procedures to establish uusnc-orthiness in qualitatire reseatch. These include: keeping daily logs of acuvÏties, refle'uve musings and methodoiogical decisions to form a research audit mil; maintaining a prolonged engagement in the field to ensure research credibiliry; conducting persistent participant observation; and receicing feedback from key informants ( h o c d e d member checks) of the smdy frridings. Creswell (1994, p. 158), allied with comments from h l e h (1988) and Mies

and Huberman (1994), dso identifies the need to niangulate or h d convergence among sources of information, different investigators, or different methods of data collection as a wag- to ensure

interna1 research validity (i.e., the accuraq of the studfs fïndings and whether the)- match rcaliry).

In tcrms of this dissertation, al1 these items were performed. Field notes of obsen-ed participant activïties were created throughout the entire data collection period; insights and ideas on cmcrgmg themes and categories and methodologicd decisions were recorded directiy into QSR Mrivo during data analysis. The researcher spent months in the field at both the pilot and full study organizational sites to l e m and understand the b ' s culture, build tmst with the

participants, and test for misinformation. Persistent obsert-ation occurred during this period to facilitate the identification of the characterisacs and elements most relevant to the phenornenon of study. Pilot, progress, and h a 1 reports were handed back to various organizational members,

namely the manager of the pilot participants, the BUL interna1 champion, and the manager of

the Usability Team to check the credibility of the study's hdings. Interviews Mth key informants in the field, such as those from the IS/IT and Communications group, were conducted in addition to participant i n t e ~ e w sto help increase the credibility of the research fïndings. Triangulation of data coilection methods, namely the use of both qualitative and

quantitative techniques, was pirotal ia helping achiere a more balanced perception of participant Web use activïty and the contest surrounding this usage. There was accuracy in tenns of the coding of the data. First, since the study- occurred over manv months, there was ample tirne for the researcher to reflect and osciliate between collection and coding of the data. Second, there were many iterations of constant compaeson of the gcncrated coding categories as seceral passes through the data occuned as the researcher identified, developed, combined, and rethought coding categories. Third, statement ralidity d e s adapted from hiiles and Huberman (1991, p. 267) by Buchwdd (1999, pp. 46-47) were used to asscss the relative value of information provided bu the participants: a statement was truc if

it

\vas made b y more than one înfomiant and not contradicted by others; it was provided by a knowledgeable informant; it was a h u d fact; or could be nîanguiated by independent sources. In terms of validitv, there was consistenq with which the data were collected and analyzed as d ~ researcher c conducted ail data collection and anaiy-sis procedures by hirnself. That is, one

researcher inteniewed all participants and took field notes, transcribed these into QSR NVivo, and identificd and coded emerging categories and themes.

on the part of the researcher to the phenornenon There wns also strong t/Ieomticco/senniiv~~ of study. Strauss and Corbin (1998, p. 16) refer to theoretical sensitivity as "haviag insight into,

' 8 Thcorctical scnsititlty refers to the persond qualities of the researcher which aiiow hirn or her to be aware o f emcrgent caregoncs found in the data. This awareness can be obtaïned through past expericnces, such as rcading, training, cducation, prior research, or through current dealings wïth present resewch. The abiiity to understand data and ro idenu- irs pertinent aspect aids die researcher in g e n c r a ~ gtheory that is grounded, conceptudy dense, and \di-integratcd.

and being able to give meaning to, the events and happeniqs in data. It means being able to see beneath the obrious to discover the ne#.

The researcher brïngs o r e r a decade of work

espenence in user systems design, severai years of exrperience in the telecommunication field, and knowledge of the developrnent and potential of orguilational Web-based technology to the research process. Strauss and Corbin (p. 17) note that professional experience is a valuable potential source of sensitivit). "as it can enable the researcher to move into an area more quickly because he or she does nor have to spend t h e gainllig familianty with surroundings or events".

3.9 Conclusion The purpose of this chapter was to present the main rnethodological approach urilized in

this research. The chapter elcplained how a case study was conducted o n several sets of distinct users at a large telecornmunications Company ualwng a variery of qualitative and quanatauve methods. The work \vas perfonned over m o phases (a pilot and full study). T h e intent was to

uirestigate

war

in which organizational participants utilïze Web information systems in their

daily work for knowledge creaaon, distribution, and use. Discussion occurred o n how the work: complements and estends past related studies; took steps to protect the privaq, confidentialiq-, and anonymity of partiapants; and ensured research accuraq.

The ncst chapter reports o n the studfs hdings o n the use of Web information systems for

knowledge work. The conceptual framework described Li Chapter 2 helps guide the presentaaon of rcsuits.

CHAPTER 4: FINDINGS 4.1 Introduction This chapter reports on the study's research findùlgs. Discussion is based on an analysis of

the participants' significant episodes of WIS activity recorded in the tracking logs, and predorninant thernes and patterns which emerged from the use of grounded theoq- techniques on interview and field note data. Overail, analysis concentrated on the use of Web infomiaaon systems from the ~ a r ï o u sperspectives of the study's conceptual b e w o r k . This helped in obraining both micro and rnacro orientations to this case study investigation of organizational

Web information system usage. Specificaily, the information needs and uses perspective of the Eramework focused atrention on the foilowing areas of investigation: major sets of users; their problem situations and information behaviours in terms of actual \VIS acti\-is this. The corporate portal is veq- read-only. It's not interacave."

P53: 'YXhat -

1 found here is that thïs [portal] is just a bunch of Web pages... I want a Web page that is dehned by what's in a database. So if I add more, then my page changes.. . my pages change... Thcre are many esarnples out there of other Web sites where things ~~~namicallu change as you go dong."

m:"It would be nice if the client.. . out clients.. . that look at these results could go in and hll in a q u e q ~vindowor menu, and then we'd have the engine in the backend pulling esactly the da ta the? as ked for. As opposed CO... these are canned reports we're showing them ~ o w . "

Inremiewer: 'What type of features would )-ou Wre seen adcied to the portal?"

P 1 1: Thac get people invol-ed! Not -

just sit there and give ?ou an interface. 'Here's the results

that give you 10,000 items, go and find it if you idce!'.

Inrem-icwer: w h t - . You want features that get you involved."

Xioreover, to be an attractive and heady-urilized application, a couple o f participants and kcy inforrnants thought that the portai should move beyond its current role as an information

publishing o r corporate cornmunications medium and towards the faulication o f everyda~work tasks and applications. For esample, one participant comrnented as follows:

P59: 'YVhy -

do 1 go here [to die portal]? Becausc 1 want to do things, but what can you do? Frankly, 1 don7twant to search ail the tirne... It's not a matter of searching. Searching is only a research actiiity. Aii the other thïngs you d o in your daily job, that's r e d y where 1 s o n of sce it [the pond] going... What 1 believe the goal should be is ali the tasks employees want to do, the. do it through the pord... It [should] d o w you to do all tasks you've got to do in your job. 1 would likc it to provide a much more clear context and interface to the activiaes you have to

perform and the information )-ou have to use that is available there. Basicaiiy, you can't do any of your tasb with it [mith the current version of the portail because it's not designed to d o w j-ou to do tasks. It doesn't give you access to the applications that might be used in the tasks you might do."

-4nother pa.rtiapant thought the portal could be more engaghg if it O ffered more interactive work-related applications, such as cornputer-based training: P l 2: "1 think what the intranet could also be utiiized for... 1 don? know about costs ... is CBTs, -

likc computer-based uaining. %, Iot of the time you could sign up for a course and go to the classroom, j-ou ho=- you leam stuff, but 1 h d that I could l e m it just as easiiy bu reading it o r going through a nitord. ,And some rutoriais won? let you pass u n d you've got a certain percentage right. So you're forced to pay attention; you're forced to l a m it. I think it c m be a ver). efkctive mcans as... some people don't have a week to take off or tn;o or three days to take off [to do a traditional course]."

4.4 Information Ecology The previous section of diis dissertation idenüfied the feanires and functions of Web information systems that influenced the adoption and use of these systems bu participants. In conuast, this section takes a broader focus towards the esamination of Web information systern usagc within the organizational contest of the hrm. Field note and interview data obtained from both participants and key informants were analvzed to detenninc the more salient characteristics of the informa5on ecology that promoted or detexed the use of Web information systems for howledge work.

Figure 4-28 below ourlines the results of this analysis with respect to the high-le\-el groupuig of coding catcgories of the Information Ecology subtree (refer back to Figure 1.1 for placement of the subaee within the o v e r d tree structure of the project). Recall that an organization's information ecology rehrs to the social and culwal contest within a corporaaon that influences ho\v information is ralued and shared across the enterprise. In generai, several broad categories \r-crc identified: the iffornafionpolilccr surrounding the design and development of these systems; thc IEI7S drr~efupmen;process by which these svstems wcre maintained; and the information ~ x h n

the organizaaon, namelv the degree and willingness of organizatïond members to share

infomiation, the cxtent to which information orerioad ocwred, the physical access to Web information sustems, the control over standards and infomiation content, and the general attirude of participants in the organization

in ternis of

che

estent to which Web-based

information is d u e d . These factors were found to impact the design, adoption and use of rzious

Web information systems, in particuiar the organrzation's corporate portal. Thus a large

pan of the discussion on infornation ecological issues pertains to this particuiar Web

information system.

lnfo

WIS Developnent

lnfo Overioad

9 I nfo Access

&re

4.28: Tree Structure for Information Ecoio

lnfo Contrd

Note that the organuation of the uee structure in Figure 4.28 is one possible interpretation of how the various nodes in the Information Ecology sub-aee could be ananged. For exunplc, there is a relationship betwecn information control and information politics; these two nodes could be grouped together in some fashion. However, the workiags of QSR NVko resect the researcher Lit0 identifjing categories and associations benveen categories within a single hierarchical tree structureo'. Thus the researcher organized the node structure according to what

he considered to be more representative of his interpretation of the emerging categories and their relacionships with each other. With respect to the information ecology sub-tree, information poiitics deais with the tensions between various WIS stakeholders; WiS derelopment pertains to the processes in place to develop and maintain WIS hctionality and information content; and information cultute refets to the generai organizational attitudes and bcliefs relating to information.

It should be mentioned that though the abore coding categories were i d e n ~ e dfrom analysis of field and inten-iew data, no specific patterns of behavïour in rems of functionai

groups were discemable from the data. The following description of information ecological chnracrerisàcs appiies equdy to administrative assistants, business analysts, project managers, middle mangers, and system developers.

With respect to information politics, there were two broad political strugg1es over ownership of the corporate portal. The first was among three intemal Communications groups

nlthin Beii Canada orer management of information content on the portal. The second was bcnveen Communications and the IS/IT division which controiled the hanciai purse strings of the portal's development.

In terms of the h s t politicai scenario, the three intemal groups comprised Communications

More discussion on the suengths and weakncsscs of QSR W'ico can be found in Section 5.5.

I''

Ontario, Communications Quebec, and Corporate Communications. Each group reported to

different cornpan). Presidents-an organizational strucniring based on federal regulations fiom h e

Canadian b d i o and Television Commission (CRTC) stipulating the creation of two sepaxate

regions for Ontario and Quebec under BCE. The Corporate Communications group prllnarily concerned itself with the communication of corporate strategies to Beii shareholders and stakcholders, while the other two groups were responsible for more product-oriented communications within th& respective regions. Of the three groups, Communications Ontario was responsible for administering the information content on the Beii Canada portal. However, since the portai serciced communications across the enterprise, the other m-Ogroups had an interest in the management of the portal content as well. In addition, both Communications Ontario and Cornmunications

Qucbec sen-iccd separate depamnental inûanet sites (cailed forums) for constituents in their regrons.

For the most part, there was duplication of sert-ices and feanires across the three sites. Scvcral participants indicated confusion over the duplication of information on these sites,

which was esacerbated by hypertest links to the two forum sites directij- from the portal homcpagc. O n e interviewee labeled the forums as 'portals u~thina portal'. A recent consuitant report substantiated the confusion by Beil Canada employees on the difference benveen the corporate portal and the two forums. .-\ key informant from one of the Communications groups discussed the political turmoil in

gaining consensus among the three groups in managing the information content on the portal. Hcr cornmenrs below highkht the need for cooperation and shared \+ion

within the

Communications groups and clirnination of duplicate information content and applications across the pond and forum sites:

1n forrnanr: ' T h e trouble is you've go t one Company, BeU Canada, but o u ' v e go t.. . two ver). strong regions which almost operate as companies unto thernselves. And therefore, each wants an on-line presence to deiiwx regional specific information and applicaaons. And yes, there is some duplicaaon right now which we're working to elrmrnate.. We're just in the process of remotuig that duplication ... in fact, we're having a meeting later this month, in cwo weeks ... no, nest wveek.. . to discuss v q seriously the relationship between the forums and the portai.. . how it should evolve and even if forums are needed. Should they even e.&t given the way the p o n d is evolving. 1 think they should ... 1 think there's a need and a desue both at the Iower Ievels and the esecutive levels.. . but dearly dehning what the portal is and what the forums are and what they should be and redu&g the duplication ïs a chief objective ught now. It's jusr going to take some cime because again consensus has got to be achieved. It talces rime."

The meeting itseif, of which the informant spoke, was a perfect case in point of the

difficdm in achieving consensus among the three groups. O r i g i n d y slated for two hours, the m c e ~ abruptly g terminated after 20 minutes of debate. During the meeting, each party coiced

different risions of the forum and portd sites. In generai, there was a reluctance by members of the three groups to give up ownership of the content they cunentiy conuolied.

In remis of the second political scenario, the Communications Ontario group reported concern orer the control the IS/IT depamnent had in detemùning portal development. T h e

IS/IT depamnent orersaw the budget for the corporate portal and had influence over the h a 1 dccision on the senices and feanires the portal protlded. The Communications group saw the

portal as communications vehicle and did not agree wlth the technology-focus that IS/IT pramoted for the portal: Infomanr: "The p o r d is not r d y within Communications. It is really at the end of the day owncd by the CIO, because they have the contract and engage CG1 to maintain the servers and host the sen-ersetc."

Informant: "The portal is owned by the IS/IT organization within Bell within the CIO's office. Ir's actually an adjunct of theu [desktop standardization initiatives] program. So the strategies overail for the p o r d are scill driven out of the CIO's office, and we lui Communications Ontario] are scen as one of the key stakeholders because we manage communications on the dau-to-day content... They're the ones who deade the strategies. We influence thm, but we don't own them." Intcn-ïewcr: "1s this a major problern? 1s this a barrier?"

Informant: T e s . In m y opinion it is because they think about the site based on technology and based on its impact on e-mail. They see it as a piece of infrastructure and they treat it iike a picce of inGastructure. We' would prefer to treat it like a living breathing environment to the view chat as the site continues to evolve and more and more people get access and the [standardizauon] project Gnishes and e v q b o d y has standardized access, this is the site where ever).one should be. This is the place on-line, this is the virtual world that Bell employees should be Living and breathlig and interacting in. So seen as a piece of infrastructure devalues its real purpose and ueating it like a piece of infrastructure devalues its real purpose. It also tends to stagnate it because th en... if you treat it like a living, breathing environment it's orpnic.. . you make changes as they're required, on an ongoing basis rather than locking yourself into arbittar); development qdes.. . the old IS/IT model.. . which is what we are locked into now.,. In my opinion... the way that it is done cunently is problematic because in the end what the CIO says, goes. So, if he wants to tum the entire site into a rag to promote the CIO's mandates, that's c,uactly what d i happen to it, as has happened recently, because the? hold the purse strings- So we don't pay for aii the stuff that haypens on the pond, the CIO's office does. h d that is problematic... It means chat ure deal with some interesring political issues."

On the other hand, the IS/IT department viewed Communications as a group that did not understand o r were concemed with technical constraints and limitations, such as bandnridth and system response tirne, in designing a corporate portal. According to one informant Erom IS/IT, Communications en\-isioned "everyttiing being on the p o d " without thinking through the technical feasibility of p r o d i n g such a solution. Admittedly, this informant agreed tbat IS/IT

was thc real owner o f the portal since the group controiied the portal's pune-strings and discussed the influence IS/IT had over the portal design as a result. Two informants from the Communications group discussed the need for a democratic

srcering cornmittee, a council composed of content representatives from rarious parts o f the

organization, as a means o f easing tensions between stakeholders int-olved in the management and dcvcloprnent of the corporate portai: Informant: "As the portal evolves, there d i be other partners that corne to the table, iike Purchasing or Provisioning.... Finance. \%en we start adding stuff Wte on-line espense reporting and bill consolidation etc... You know, 1 see the portal reaiiy as an elecuonic m k o r , to some extent, of the Company itself. Therc's not one group or department that otvns the Company. WCall contribute in our various uq-s to its ongoing management and development throughout various aspects. 1 think that the portal will cvolve in sort of the samc way. At the end of the day, there will be a cound that steers it, and as the over-al1 owners ... the CIO will manage the vcndor relauonship in terms of its up-keep, maintenance, and how it's hosted.. . but it's becorning more of a collaborative partncrship with various entities and stakeholders

internaiiy... it can't be one group, or department, o r individual that7s the owner [of the portai] at the end of the day. Because it has to be more collaborative team approach as a subset of the business itself.. . because that is how it will wolre. Ir just won't be a Communications tool or just an IS/IT platform for ce& on-iine applications. It d l be more than t h a t It will be an on-line cornmunicy. It d i be our on-line business."

Informant: "If 1 were to design the portal according to my own utopian ideai, e v q b o d y should be a player. Kre should have.. . just as )-ou d o in the magazine world where you have section edirors and owners, but then you have one master editor ... it would be the sarne wap. The corporate portai would be the master site for aii of Beli but the financing area would be owned by Finance and run by Finance. The): wouid be part o f a cornmittee that a n s the o v e r d site, but the' would run and own ail of th& own content. Ontario-speafic communications wouId be run out of Ontario Communications. Quebec's [the same]. ... so =ch staiïeholder xvould own th& own. Content management would be decentraiized, but content smtegf and overall site strategies would be run or rnauitained by one group. Preferably one group outside of Ontario Cornmunicaaons, Quebec Communications, o r Corporate Communications... So, it tvould foliow the mode1 of a lot of the Canadian banks and large manufacruring companies are starting to foliott; which is they pull together.. . they a e a t e a departmenr which is literally new media or e-business and they people it with usabilit). archicects, IS/IT people, Cornmunicaaons people etc. etc. And thefve a dual reportïng structure. So you'U report into o u r Communications group but you also, you mainly report into that e-business group. And it recognizcs the importance of aii the business suacegies eutemai and internai and because !-ou7reoutside of a specific business unit, it's easier to stay corporate focused and stay focused on what the wholc thing is about which is aeating the right e-business environrnents rather than +-kg to create a business envïronmenr based only o n Communications or q i n g to create a business environment based on only an IS/IT rnindset to iq which is problemaac because of the political aspects of it."

However, the set up and running of such a steerllig cornmittee rnay prove dïfficult to

achieve. This difficulry, as Mth the scenario of the three Communication groups described above i n getting disparate groups together and agreeing on a sharcd vision, lies in reaching consensus among the v&ous portal srakeholdcrs. The follorvuig encerpt Gom one of the Communications informants speaks to this difficulty: Inrcmiewer: ".Are there any problerns you perceive in terrns of establishing poliàes and guidelines for portal development?" Informant: ",Acheting consensus amongst the stakeholders as to what those policies and guidelines will be." In ten~icwcr:'L7LWh)r k that difficult?" Informant: 'Well, Communications..., for euamplc, takes more of a user approach. What does the user want and content approach. IS/IT looks at it more Gom a nenuork and application

perspective and just a general tech side,. .. How- long does the page d e to download? How stable is the semer? 1s it secure within the h r e d ? That s o n of .... That's the ernphasis thcy take. So ?ou have different needs and desires and different mind sets co&g to the table. So consensus is needed. Sometirnes it's easy and sometimes it just takes a Little bit o f work. Sometirnes politics and ego are involved. Even within Communications and the Communications groups, the IS/IT folks.. . there's different opinions arnongst those people

tao."

With respect to the WIS development, there were various procedures for making portal system enhancements. For minor content changes, Communications Ontario accessed a content management system a-hich allowed them to publish news items, post picnues, and conduct s u - e u s through this tool. For content that could not be modified directly by the tool, Communications Ontario c d e d the customer sen4ce help desk, filed a trouble ticket, and waited for the appropriate people in CGI, Bell's arms-length systems solution provider, to d

e the

content change. Suggestions for enhancements to the portal's fearures and hinctions were gathered and

prioritized by the Communications Ontario tearn on a continuous basis. Employees could send

their suggestions through a feedback option at the bottom of every portai Web page. Deemed changes were costed out by CG1 and brokered by IS/IT. In this capaciw, IS/tT functioned as a middle man bctween Bell Canada and CGI. As one informant from IS/IT explained, it is in Bell Canada's best interest if IS/IT brokers the relaaonship with CGI. The idea is that IS/IT is fînuliar with technologv and software development and can guarantee that the bid for work is

fiîr, representatiw, and comprehensive. According to this informant, some user groups had gone ro CG1 direcdy to get systems work done and the end result was a higher dollar price and questionable deliverables had IS/IT intervened. Admittediy, the informant stated that Bcll's rclîtionship with CG1 wvs "a bit strange and complicated" since Bell Sygma was purchased by

CG1 and Bell owns a majority perccntage of stock in CGI:

Informant: "One would think that CG1 should take better care of Bell Canada and cut a good deal since the' are owned by B A "

One participant commented on the complex relationship BeU had with CGI: Inten-iewer: 'You mentioncd that CG1.-. they're the gatekeeper or custodian of Beii.net. Do you have problems with that?'

P50: "No. -

It's just a v e q complex relaaonship in t e r m s of the partnetship. When we've got sen-er probkms we cal1 CGI. CG1 caiis Bell reps. Beli reps corne and do the work It's just one of those out-sourced hnctions chat's been out-sourced so many t h e s , you don't r e d y know

who's actualiy doing the work"

One informant described the htest version of the corporare portai to be a compromised

solution as a result of pressure from senior management within the IS/IT division to constmct and deplov a new version within 60 days. CG1 agreed to this deadline and the work was

cornrnissioned. In fact, the srandard project methodologlr guidelmes for systems development projects was disregarded for the portal redesign. The inforniant esplained that though a good attcmpr was taken at creatïng a more robust portal, only a few new features were addcd in the end, such as the classifieds section, a new search udity, and on-iine news, but much of it was a

re-pachging of the old p o d . The informant stated the result was a "compromised portal solutionM-a redesign that deiivered a new look and feel with few new features added. Other factors beyond senior management pressure influenced the design of the portal. Thrce discussed during the interview sessions were t h e , cost, and available man-power. For

esample, the p o d content team in Communications Ontario consisted of oniy 3 fuli-time equivalents (FIES), making this a constraint rowards addressing changes to the portai requested by over 30,000 Beli Canada employees in a tïmely fashion.

In gmeral, there were concerns voiced by several participants on the process by which modifications to the portal and departmental intranet sires were made. Some participants thoughr it took too long to make simple scraightforward changes or too manv playen were

inrolred in the system development process. For esamplc, one participant was a p p d e d at the slow process by which changes or modifications to the portal occurred. This, according to the participant, was because modifications, no matter how simple or complex, were fkst fed through a team in the Communications department who then informed CG1 to make the necessary changes. The participant e ~ ~ l a i n ethat d CG1 was ofien over-whelrned in terms of work to do, so changes took a long tirne to go into production. To emphasize her argument, she gave an evample whcre a

I'ice-President of the Company wanted to make simple textual changes to his biography posted on the portal. The changes took two weeks to make since CG1 was

busy at that Mie and the

participant had to quem the Communications team on what was taking so long. T o exacerbate rhis situation, there were some minor errors in the test when it was Cuialiy posted on the portal.

The IT requested these be Lucd. That second round took another week and a half to implement. In total, it took three and a half weeks to implement simple testual changes on the portal. r\ccording

to the participant,

this was unacceptable. She pointed out that

Communications Quebec had a better system for making content changes on the Quebec forum since the derelopment people sat only a couple of cubicles away fiom the people r e q u e s ~ gthe content changes. There, simple testual changes took only 10 minutes to complete. Another participant comrnented that the formal process of building depamnental intranet sites at Bell Canada was too espensive and slow since the poiicy deemed that CG1 was to bccome inrolved. The pamupant discussed how she circumvented the development process when she created her departmental iotranet site, espiaining the official development process was too costly and time-consuming to warrant irs uciiization in practicc:

Interviewer: 'Weii y o d 1 get there. You have to start somewhere.. . can't wait for eveqthing to bc perfect before roUing out an intranet site."

P64: 'Yeah, but -

not oniy that. 1 \vas racing 4me before CG1 found out 1 was doing it and slapped m e on the hand" Inten-iewer: "Oh. They wouldn't let cou do it?"

P64: "No. -

1 shouldn't even have done it. 1 should have paid them. T o give you a cornparison, anothcr site \vas created for [another department] that w-e piggybacked w-ïth. But th& site cos^ 1 don't know, NI o r 60 thousand doiiars or something of that nature. And it's a very nice site. Ir's more professionaiiy done than mine1 mean there's aii h d s of reasons why it's a good site, but it was also espensive and it was a lot of red tape to get it. Whereas mine was just ... you know, 1 did it in mu spare tirne.. here, go throw it up P o s t it on a Web server]. So.. but CG1 is supposed to, 1 understand, d o all.. I f s supposed to become a project. And chat's the problem. That's why 1 didn't go that route because 1 hm*that it would cake longer than we had to get it running."

Inten-ieu-er: "Yeah, so it's the [development] process that is the barrier."

P64: 'Yep. They're getting better though. CG1 is getting better bur they're -

a little bit process h e a q and they're s d a Little bit,. . the pendulurn hasn't swung back yet the othec way where.. . right now thefre a little too controlling of what's going on out there."

One participant gare an esample of the diffidaes in c r e a ~ gan estemai Web site since

diffcrent policies esisted on the look and feel of the site depending on whether the site \vas sponsored or not by an outside organizauon. The partiapant's major cornplaint was that she u-antcd freedom to design the site in the way she felt was preferable, and not to be constrained

by poiicy. The following is an esccrpt of the interview with that participant when the researcher

qucsuoned the context surrounding one of her signifiant episodes in her Web t r a c h g log. Intemiewer: T o u checked out the Jazz Festival site?"

PSO: "Yep.. . and Beii is a sponsor. .ind -

another thing that 1 \vas interested in is that the Jazz Festival \S%b site was designed by Bell but it didn't have to be uithin Beii policies with the way chat Beil Krcb sites look. And we're struggiîng uich that in designing our o m [external project site] site because the acmai navigation through BeU is v e q difhcult with the wap-around top rhrough bottom.. . so we actually wanted to get a business case to tr)- and position it so that we could have a Ilproject] site that didn't fit withïn the BeU wrapper." Tnten-iewer: ' D i d you use that?"

P50: "W'eil... you can only get out of the policies of the Bell wrapper when it's a sponsored site. So the [project site] doesnytfit into that caregory, which is unfortunate. So we're q u i g to

design a Web site with navigation w i t h the Beii Web navigation, which is ver)- clumsy. It minirnizes your screen space for your actual text"

hlanv users also complained about the need to re-regmer departmental intranet sites on the latcst version of the corporate portai--a serious £law in the WS derelopment process according to these participants. Users were upset chat links they previously depended on

CO get

to other

d e p m e n t d Web sites were removed from the corporate portal without notice. Some participants did not bookmark these links and were b t r a t e d by- scrarnbhg to locate or obtain the CRL address of departmentai ina-anet sites. The foUoMng escerpt from one partiapant

typifies user reaction to this rnissing tünctionality: PGO: "One -

of the difficulties, and 1 understand how we got to this point, so take the comment for what it's worth, when they mot-ed over to the new portai format, the)- went out to all of the people who had inuaner sites and told them they had to register. O u t of aii of those sites, and 1 think there were some 200 sites, there are only 30 some that are registered. AU the ones that I need aren't registered, don't fit under the i n m e t criteria that thejgve set out and unless 1 know the e-mail &JRL] address, 1 can't get at them and neither can anyone else in any of my groups... \Xihat 1 thïnk probably happened was that the form was sent out to the person who manages thc site, she wasn't sure where to put it [in the portal c1assiGcation scherne], and it just nwer gor donc cause eveq-one i i e d that it would be folded over and somebody r i Comrnunicaaons] would make that decision- So, I'm not sure where it's at- I've got somebody working on that."

Two informants from the Communications groups commented on their reasoning for

taking away the pre-e'risting links to departmental pages. First, they wanted to obtaîn only an active l i s ~ of g interna1 Web sites and second, they wanted to get the intranet owners to classe

thcir own sites accmding to the content categorization schema detised bg Communications: Intcn-iewcr. '9Vhy force the individual sires to register? Why didn't you just look at the sites that were on your old list and move them to the appropriatc arcas?" Informant: '\Xe

thought =-ho better to know a certain site's customers or users than the individuals who run and operatc this site. So we M t it up to them to categorize themselves." II*

Informant: "We didn't port over the sites €rom the old portal, because in the old portal it included a lot ofreaiiy outdated sites. It included sites that were not Beli Canada: the): wcre Bell Emergis; chcy were CGI; they were former Beii Sygma; they were other companies and not Beii

Canada speafic. O r tbey just didn't operate a n p o r e . Two thirds of the sites o n the old Belinet were dead.. . the links were dead So7 thac's why we decided to start from scratch- We ported over sites that we h e w euisted that were strategic Wce Year2000, &e IS/IT, like HR.. . things like that. And then, we provïded fadiaes for people to register th& own sites. We're also sending out an e-mail to everybody aithin the nest couple days to thc entire Company saykg 'heu, if you run a site, please corne register p u r site'."

With respect

to

information cultute, five areas were identified as being sigaificant ui

irnpacting the adoption and use o f Web information systems: 1) the ~villingoesso f employees to

share information; 2) information overload; 3) access to \Veb information systems; 4) control over information standards; and 5) the general attitude o f organizauonai participants towards the value of information.

In terms of injormahn rhan'115, most partiapana indicated a general ease of sharing

documents, plans, and reports between colleagues and project tearn members with whom they worked. Trust seemed

to

be a predominant factor in determining whether information was

shared o r not: Tntcn-iewer: "1s therc a Iot of information sharing, o r do you hnd that people hoard in formation?"

1)50:"It's rery open actudy." Inten-ietk-er:' W h y do you think that is?"

PSO: "1 guess parti-

it's mu role. 1 think it's d s o in line with people not n e c e s s d y being solely accouncablc for one task So thcy share it with everybody to get information and feedback. And ic aIways seems to be collective positions.. . collective work is wrhat we do. Ir's not .. . one pcrson accountable. So we have to share informaaon."

P63: "Certainly, -

the information sharing 1 would say is probably at an above average level. L k e

the ~~illingncss." Tnten-iewer: ' W h y do you think it's above average?"

PG3:"1 just -

think that.. . becausc it is imperative that we work as a tearn in this group, because thar is sort of what our success is based on is team work. It's not sort of a one person show. One person can't do it d because there is so many components to the execution process that you need to have d team players on board and in order to do that they need to have the information."

P51: 'Teah, yeah. -And 1 mean I've dealt with [chese] groups before. I've worked on this -

contract [project]... for three j-ears. I've known most of these guys for nvo plus years, off and on. So 1 know thern weii."

P58: ' T h e n 1 hrst staned -

lui this job], my department was sort of a new group, and we deal wlth [another group], and chey have been in Bell for quite awhile and they mitiaily were wondering who we were and what we did and we owned th& budgets all of a sudden.. . and u-e're telling them what to d o but.. . It was chdenging to get them to respond to our inquiries [for information on work pracuce] but now it's improving significandy. These guys are working wirh us. So, rhat7ssort of an example of structurai challenge we face.''

Interviewer: '?t \vas difficult, so you had to get to know them." get to h o w them and they kind of understood what we were doing.. . that we were not going to replace them.. . what the role of [my groupj in the organization is."

P58: ''You -

Participants noted less shaxing occuned when it involved people aith whom they were unfarniliar o r when there \vas a fear of protecting one's domain o r job security: Inten-iewer: ' V h y do you think there is this reluctance to share?''

P53: "I think it's -

maybe a feeling of insecurity. If 1 git-e up the information that 1 have, then you h o u - , it's my job kind of thing... When 1 said a litde bit earlier that they were a bit relucrant to give out informaaon, it's more protecting their domain- That's it. It's more bWig protective of your domain where 1 see it as y-ouPe got to share information."

Interviewer: 'Do o u frnd that these people are w i h g to share o r d o they hoard the information or do :;ou h d that it's pretty open?"

P 3 : "1 don't -

know. There's s o a of a gdd, let's say,. . people are interested in what you uTantit for, that's for sure. But, you basically ... if .ou aren't in the process of restnicninng their parücular ficfdom then they seem pretty willing to share. If you want to take it to say that thek [\vork is]... set up irnproperly and );ou wanted to rcview it, I think you would have a r d problem there." Interviewer: "Ohuyso it sort of goes both ways ..."

P54: "Righr." Inren-iewer "If there's no turf infringement then people are willing CO share."

1'54: "Esactiy." -

In tcrms of the impact on Web information system use, most people who were willifig to sharc docurncnts and ideas r i a this type of medium were more k e l y to do so if there was a

f a d t y , such as in TeamCast, a-hich resaïcted access to shared documents to pre-dehed

indi\-iduals on*. E d e r in this dissertation, Section 4.3 discussed the desire for participants to have Web-based tools for document, bookmark, and idea sharing among project team members ( d e r to this sub-section on collaborative appiications for more d e d s ) . For csarnple, one partiapant commented that in his previous job, infoxnation and knowledge was shared more readiiy since the organization had tools such as an inuanet,

newsgroups, and listsen~sto faditate this type of eschange. The participant found these tools bcneficiai in helping hïm mainmin currency in his technical programrning skilis and wished that

Bell would provide a si&

means of knowledge eschange within the hmi. hnother participant

cornmented on the general desire of employees to share knowledge, but stated the need for the companv to protide a means of facilitating that sharing: P59: "It's -

not that people don't want to share it, it's just that... it's not made available in a sh&g ... the environment doesn't make it available in a sense. There is no centrai repository

that provides you any context about tvhat information rs avdable. And 1 see that as a huge irnpediment to sharing. .And especidy. .. even if the stuff is aii out there, there's none of the things that seem to be generally available on the Web. .. where if you have a W'eb site and you've got .di these documents and eveq-thing.. . like you have a hlernes [here thc participant is ceferring to Vannevar Bush's 1945 vision of a desk-top personai infonnaaon machine; see Burke (l97-4)] ... you can search chat thing. You should be able to search this stuff and Gnd thïngs. So that's an irnpediment. N o repository. We don't know what Ïnformaaon is out there. It's bad."

P60: "It becomes dificuit at times to get informaaon to £ind out where it's housed, chat sort of -

thing. System-ttrse, I think we're l a c b g in that way. ïVe don't necessarily have a centrai archives capability. And 1 think generally wve don't have the manpower to be able to do the education and management of that."

Interestingly, some p d c i p a n t s indicated an unuillingness to conmbute documents and idcas t o a public space for fear that there may be unexpected disuplinary actions. Participants

commented on the need for the environment to support sharing in terms o f both a physical means to do so and a contest Gee of repercussions and critical judgements.

m:":hd

hoarding does go on. And I mean that in the sense that... project per project... it's son of like &at you want to watch your bun and you don't want to let a document out there, for example a technical document, until people have approved it. . h d someone might want it, and for d intensive purposes it's ready for them to look at but you don't want to commit or you don't want it out there floating around, eqosing yourself. Someone might pick it up and say 'hey, that's not right'. In that way, there's hoarding. I can have something ready and my manager will say 'don't share it with the world just yet'. So I would say only hoarding goes on with non-authorized documents."

With respect to information ouerfoad, many participants voiced their concern in trying to manage vast amounts of information they received from disparate information sources. One n-pical frusuauon was the misuse of corporate e-mail in terms o f receiving irrelevant messages and/or krge file attachments that quickly filled up a person's inbos and prevented other e-mail messages from being received: PS8: "Well, I mean ... we get tons of e-mail from ... every morning ... from everything... everyone's probably mentioned this kom Bell.. . everything from job posrings to organizationd

changes or new phone numbers. Who really needs to know if someone's got a new phone number? Just look them up in the Directory. So, there's a lot of stuff U e that. Anydung from getting a better title on the message to sorting to whom you're sending it to [would help reduce in formation overloadj. Don't send it to everyone.. . everyone's not involved."

P65: "I &k -

therc's so much information out there that its really hard to digest it all and decipher what is really relevant for you. If you get, I don't know what kind of volume of e-mail people get, but when you're getting 50 e-mails a day, how do you find time to go through and determine what's really relevant to you and what's not."

P52: 'There's lots of m d that comes out on the system to all employees advising us of new sites and things that are of interest. But, I guess, you sort of look at it and go 'that's nice' and dclcte it! But often you don't go into the site either. You just look at the note and say 'okay, I'll have to [look ar that]'. The odd rime I'll write down the site address in my addrcss book if it's an HR site or something, that I would h o w how to reference it. But often you don't go in and browse it because you're just trying to gct through the mail and you probably got more things you haw to do that morning. You think you'll come back to it later but you rarely do."

;\nother source of frustlation was the difficulty participants had in searching information

located within the port& departmental intranet sites, and the World Wide Web in that returned

hit list items were not relevant, or the iists were too long to peruse in a efficient and tirnemanner. Several participants spoke o f the difficulties in f i l t e ~ gtheir own information G o m

these sources, for esample: Pl?: 'YOUhave to do a i l your own filtering. Sometîmes 111 get e-mail and its three pages, and only one line is relevant to me. There's too much information out there, but it's not d good

information, it's not aii usehi. So .ou couid be reading stuff that's rcally not relevant and is a was te of your tune, especially if );ou7\-egot a tight timeline."

In t e m s of IVeb infornation ytem

arcers, the corporation had a strong desire to make

the

corporate porral the primary vehicle for information dismbution and funcaon as a gateway to

Beli Canada's 2,500+ departmental intranet sites. In fact, the stated god o f the portal was 'to provide [the] interna1 employee population with a centralized on-line source for news information and applications that help them do th& jobs'. Though this was the plamed goal, in realiq- not everyone in the organization had access t o

the portal. For esample, rnanp field w-orkcrs had no access t o a computer at ail; those that did often ualized older t e c h n o l o g (e-g., Netscape Naiigator 2.0) that restricted them from accessing

the current version of the corporate portal: Informant: 'mot everyone has access or the samc h d of access [to the corporate portal], whch is why weyrecurrently running tu70ponds: the old one and the new one." interviewer: ",-\nd why is that?"

Informant: "lt's because the CIO's office did not do proper studies at the beginning to dctcrrnine what krnds of technologies are out thcre [in the fieid] and they had just sort of arbitrariiy decided that weii when they derelop the new portal, the old one would be turned off, and if you couldn't access the new one, well eventuaiiy you'd get to it. And we ['we' as in Ontario Communications] decided ... it was our stance that if you had access to the old one Y O U ' ~nced to continue having access to the portal. And they inight not have them roUed out on a [standardized] computer before the company deades to say that, for esample, ail IIS messages stop and everything goes on the portal. So if you start p u h g away tools without gwing them new ones, you'rc p r o d i n g a big problem gap. So we put our foot down and said 'No, thcre's s d a bunch of people' ... and it's becausc of the way they irnplementcd the new portal. They could have implernented the new portal based on technologies that would have aliowed the users that have older technologies within the company to acccss it. They didn't have to code it to Netscape 4.0.The. could have just as easily.. . they could have got the same

interface, all the same functionaiity, if the)- chosen to code it to Netscape 2 0 , but they didn't. They coded it to Netscape 4.0 with d kinds of wekd f k k c stuff that was not really necessary, but.. . as a result, they cut off a large segment of the population. What it means is that we have to continue to run the old one so that they s t i i i have access to [the] corpomte directoq, s d get ncws.. . it's not the same h d of environment and it's not as up-to-date"

Furrher, there were reported difficulaes accessing Bell's intemal network from home o r remotelr in the field. Bell technicians and other field workers were particulatIy susceptible to

haxulg limited access t o on-line information. A few participants commented on the slow de.

in

loading the portal homepage remotely and downloading hies. Informant: 'The fact that a Iarge number of Bell ernployees are field workers. They're not wued into a cornputer all day. Thefre not office workers. But the Company has detennined chat we should be moving to a 'webified' Company- So how do 'ou reach people who aren't on thc Web during th& work da): where it has nothing to do with what they do in th& daily job? .And that's not going to change within Bell anpime soon. They're not going to become a company full of office workers. So, that's a big challenge. If you begin to concentratc ali O W... strategies based on an o n - h e environrnent, you're rnissing a Luge part of the ernployee population and alienatïng the veq- p m of the ernployee population wve always had trouble rcaching and letung h e m know wve think they're valuable. That's a major issue."

m:

'Or sav, dialirig in from home. I don't know but accessing the portai, it must have a lot of graphic images, but it seems rery long. I'm sittïng there going 'what the heck's it doing?'."

With respect t o z~ônnation tond, standaxdization o f information seemed to be a prcdorninant theme. Both iS/IT and Communications Ontario indicated that they wanted to

rationalizc the corporate portal, though k o m different perspectives. For instance, IS/IT wanted to standardize die portal's technologicai design (e-g., the use of standard-approved Web page dcrelopment tools, the inclusion of standard buttons o n departmental intranet sites back t o the

portal home page). Communications Ontario, with its emphasis on creating an interna1 Bell presence within the company, wanted t o standarche the portal's look and feel (cg., the use of corporate colours, standard fonts, and the piacement o f Bell's logo at the top o f intranet sites'

main pages) as well as the content posted on the portal (i.e., the group currcntly cdits and standardizes the wordùlg o f content written by others in the organization before it is posted o n

the portai). Both groups indicated a need to standvdize the compmy's 'undernets'-ad

hoc

departmental intranet sites developed by ernployees that were not formaliy registered with the portal and did not u s d y foilow standard intranet der-eloprnent guidellies. To do so, the IS/IT group recentlv embarked on an initiative to identj- the undernet sites that esist in the company;

Communications Ontario indicated that with

the

latest portal release, they dropped th&

previous requirement for sites to comply to guidelines before the17 could be registered, in the hopes rhat more undemets would identify themselres so that later they could be encouraged to

be made 'more appropriate'. In effect, this relasation of standards was a sofier approach in tq-ing to sway employees to conform to company Web standards: Interviewer: If 1 wanted to crmte my own Web site, does it have to go through a formal approval process o r would you just post it? Informant: '%O. Fùght now there is no approval process. There was, but we decided to tift it to encourage people to register. But that's part of what pS/ITJ is doing in setang and dereloping standards. And we'U play a part in that You know, what's acceptable and what's not and what can or should happen. But right now, anyone can register anything with whatever standards thefre followkg because we're sirnply just providing a iink by fiULig up the site regisuauon form which is located at the bottom of each page [on the portalj. But we're also elcvatïng and lifting to the top of cach of these sub-sections."

Informant: 'There are, depending on whose numbers )-ou beiieve, 800 to 3,000 intranet sites wïthin Bell Canada. The CIO's office calls them 'undernets' because in a lot of cases thefre just one guy whose written it in Frontpage and is running it offhis N T machine at his desk. O r hlmsclf, or his little group, or his deparunent or whatever. There's dso larger sites that are more corporate sanctioned like Year 3000' and things k e that So, as a result, a lot of thoçe intnnets, we don't even h o w esist. They haven't done it through any official process because Bell didn't have an offiaal intranct policy in place for man. years. So they just went ahead and ciid it on their own. So we don't know about them ... In the new portal, u n U e the old, we havc absolutcly no criteria for registering Four site. If your site is about your own personal ski d u b for si.- people in your dcpartment and running it off yout own computer, Gne we'U register your site." J n ren-iewer: "Why the change in philosop hy?"

Informant: "l3ccause we figured if we put in a rationalization or guidelines for w-hether or not you could List your site, then we would lose aii those sites wc don't know about So what we frgured out right now is that we would gather the information. We'U get everyone to list th&

site. Thar way we know they e-uist and we c m start gentiy and graduaiij- remoulding them and reshaping chem and rationalking the kind of sites and lening the company h o w what is appropriate and what isn't appropriate. But we can't do that u n d we know about them." Intemiewer: "So it's a way of getting them in?" Informant: 'Xight. We have iP addresses.. . the CIO's office has I? addresses for most of the intranets that are out there and the' are beginning to craft a strategy that will have them rationalize those sites. But at this point, we're not there yet."

According to both the IS/IT and Communications group, control over Web design and contenr mas a n e c e s s q means for c o o r d u i a ~ gand developing a cohesive company gateway to

in forrna tion: Informant: "There always has to be a centralized point... with a Iimited nurnber o f people, that have ro calce ownership at the end o f the day with the ulcimate responsibiliy." In ten~iewer:'?.%y do o u feel that that is required?' Informant: "Because if o u don't hsve some forrn of an owner at the end of the &y, then there is no control as to how t h g s get done. It's just k e throwing a bag of candies into a room of kids. l'ou couid t d them 'ohy, divide it up f k l y and evenly'. You h o w they're never going to do it unless someone is orchestrating the process."

There was mked reaction from users on the control over Web site design. Some participants saw the necessity of setting standards: PS4: ''This is the whole thing with me. WC should be standard ... A lot of thïngs are not consisten~and we donyt foiiow standards. We should have a standard look feel. Then you know it always works this way."

m: "Yeah, we need

a standard... we need to standardize for costing savings and whatevcr

clse."

PGO: "1 appreciate the requirement and necessiq- for standards and so 1 appreaate wherc it's corning from, 1 think it's probably to have a central place where peoplc can go to".

Howcvcr, several participants were vocal in their opposition to the company's control over Web design. One participant was hstrated by her attempt to create an intranet Web site. She

had cornmissioned the work of hoscing the site to another group and ran into trouble d e n m

g to get this done since her subordinates wbo dereloped the site udized non-standard

s o h a r e . The partiapant had to make certain aesthetic changes to the site as weil as convert the site to FrontPage '98 (Bell's deemed standard s o h a r e for Web site design). The participant was frustrated that she had to convert her site to a package she saw was more inferior to the one that was previously u&ed.

In essence, she sa=- the standards as being myopic in ternis o f not being

flcsible enough to consider improvements o r changes in deemed standards: Inten-iewer: "I'm corporate standards?"

to hnd out why they reject certain sites. -&e they q 5 n g to foiiow

are. And what's rery interesthg about that.. . weU we c m get into that when 1 teii you about how the Web site went ... but there's a svuggle hem-een douig things right and doing them quickiily and efficiently. Oky? And that's always the one. Wd,1 developed this site and 1 did most of the work myself but 1 needed someone to actuaily ciean it up because 1 was an amateur but someone else to post it to an elcisting host. And it never occurred to me that they wvouldn't foUow Beii standards. As it m e d out they didn't on the software, and it caused us big problems."

PG4: "They -

Intcn-iewer: "Did you know what the standards wcre?" P64: "No, but 1 relied on them to makc sure they knew the standards. Yeah, 1 knew what the standards were and 1 bought the standard, which was FrontPage 98, and that's what 1 created it in. and 1 gave it to these guys and said 'oliay, please post it to this host' and clean it up o r do whaterer because 1 had problerns wlth Links. But a n v a y , 1 said 'please post this for me' and they said 'oh, did you h o w that they changed some thrngs?'. They specificaliy said to me on thc phonc 'your Bell Iogo or Beii icon or whatever was not standard.' They changed that. I said 'okay, right, fuie.. . they know what the standards are.' Weii the? came back and they used software that wasn't FrontPage. So they used non-standard software. So when it came back to the tcarn that was to maintain the site, Corporate Purchasing wvouldn't let us bu. the s o h a r e that \vc now needed to run the site. \Ne converted it back to FrontPage, but I'ii teii you it was Likc 18 headaches in the meantirne q%g to convince these people look, just git-e me some software, oka!-?. .. 1 don't care [about the standards]'. T o th& credit, they're q i n g to maintain standards and 1 understand that, but when something like this happens and 1 did not h o w . . . 1 m a n 1 gave it to them in FrontPage, I'm expecting it to come back in FrontPage, a.Li of a suddcn it doesn't.. . you know. Intcn-iewer: "Yeah, you don't want the process to be the obstacle." P64: "0, -

bat the sofnvare they used is bener than FrontPage. And that's obviously why thcy

used it. But, it wvasn't sundard." Tnteriewer: '%%O cares what software you used to develop the Web page in?"

w:" T h e concern is s

o b e cost And that's how Bell was before. If everybody urent and bought their own stuff... it wasn't th& hult necessarily and in some cases it was.. . but they would go and buy new tors just to hnd out what the^ were. The? were spending a lot of money. Nothing was standard. , b d weB..." Intenietver: "Okaj-, 1 understand the need for controIs but there's a Gne he..."

P64: 'Yeah. Yeah. You have to [bahce -

benveen looseness and suucnire]. .. What 1 think they should do, 1 was taking to them about this just two weeks ago when 1 was arguing with h e m over this software, 1 said 'Okay, Gne, we have to use the standard sofnvare, the FrontPage, but 17rnhere to tell you this other soft\i-are is better, so how d o 1 get that informacion across to the people who make the deckions what is standard sofnvare?' And interestinglg enough they said 'you can't.'. 1 said 'okay then!' So, there's a dictator over there sajing Y o u use this and we won't listen to any feedback'. Oliay. No, 1 don't know if that was just the person [she was talking to about using non-standard software] but 1 just dropped it. 1 had enough anyway. 1 said 'okay, okay, we'ii use FrontPage.. . fine, fine.' So."

Another user elaborated on her history svith utiliziog non-standard sofnvare in her job. She \vas frustrated at being seen as a trouble-maker in not using standard sofmare and the lack of

support she received from the Help Desk in the Company as a result. She stated that she used non-standard software to get her work done and wished the conuols were more flesiile to handle pamcular situations:

m:"1 guess m); issue from a... corporate standard is that the srandard Gts the norrn and the standard fis a benchmarked group of people, but does not make any provision for angbody outside of that norm and is not flexible to go outside of them."

In terms of a~iitndextoward~IVeb-based infmaiioion, analpis of the data hinted at the changing narure and perception of the use of Web information systems among Bell employees. One

participant noticed the oscillation of acccptance and use of Web-based information and commented on its back and fonvard nature. Shc also commented t h a t employees with longer csperience in the Company were less accepting of Web-based information sources: P53: "And -

1 t h k for people who have been in Beii for a very, very long Ume thefre set in th& ways of thinking... and 1 think sometimes it's a littie bit frightening for the people who are aircady in chere when something new comes dong. Cause change is needed, but it's also difficult for people to accept change. Inrcn-iewcr: '730?ou h d that culture is changing as new peopIe join up?"

PS3: "...What -

I notice bs] the pendulurn keeps swinging constandy. They go forward and they take a step back. And they go fonvard and they d i e a step back It's Wce for mer)- three steps that they take they're t a h g one step back."

-4nother participant commented on the influence of the portal and deparunend intranets in c h a n p g the cultue o f the organizatïoc. In her view, the introduction of access

to

Web-based

information changed company culture and promoted the sharing of informaaon within the organization: Tntcn-ïewer: 'lVhat about the culture of the organizauon?'

P23: "1 ttiuik it's changing. 1 think it used to bc that people really protected the infonnaaon, but 1 found lately within I'd sa? the last tuTo years, a big change in that. I think now information's a lot more accessible sirnply because one of the main things is because of the Intranet."

Though encouraging, another partiapant noted the difficulty in changing people's attitude and use of WeY information systems, in particulat a Web-based groupware application. In

general the participant did not think the company culture m e n t i v supponed the Mdespread use of groupware tools but was optimistic that people's artinides would e v e n n d y change:

PSS: "Tt's going to be a huge cultural shift, Gom the information I've been getting, to get people to use this [a Web-based groupware system] o n a d d y basis, going in there. You h o w , a calendar is a great thing ro see in there so that everyone o n the tcam can see when the nest meeting is. There's a deliverable m c b g mechailism on there so that when o u log in o u can sec what your deiiverable is, you can update what percentage of it you are complete. So that's kind of ncat but it's going to t a b a huge sort of cirire to get people using that and I don't think people are being encouraged to. Tnten-iewer: "Do you think that w-îühappen or do you think the culture here won't support it? it's got to happen. It's got to happen. It's got to happen. Sure the culture ... l&e even-thing else, thcre's going to be a Little perïod of acceptance."

P58: "Oh, -

4.5 Conclusion The purpose of this chapter \vas to present the study's research hdingj. -4nalysis and prcsentation of resuits were accomplished through use of the study's conceptuai framework. The information needs and use perspective focused attention on major sets of users, their problem simauons and information behaviours, and the personal characteristics of key WIS knowledge workers. The value-added processes perspective focused attention on partiapant perception on the design of Web information systems interfaces in support of knowledge work. The information ecology perspective concenuated on an esamuiation of Web information system usage within the organizational contest of the hmi.

In cotai, a comprehensive set of hndings was produced. The next chapter protides an indepth assessrnent and summarization of the results presented in this chapter. It does this by a n s w c ~ gthc studv's research questions and presenting a new mode1 of organizaaonal Web information system use based on the hndings presented above.

CHAPTER 5: DISCUSSION AND CONCLUSION

5.1 Introduction The hdings presented in the previous chapter offer a rich and descriptive account of the current use of Web information systerns for knowlcdge work and the factors that impede or promote such usage. e e r a l l , analysis of the data showed that Web information systems, to

varing degrees, were used at the case study site to faalitate organizational knowledge creation, distribution, and use and that certain characteristics of the interface design, users, and information ecology of the fkm influenced the degree ro which these systems supported such acti+. The goal of this chapter is twofold: 6rst, to siimm.rixe leamings based on the prerious chapter's hndings; and second, to offer a reahtic description and portraya1 of the utilization of

Web information systems for knou-ledge a-ork. In essence, this chapter is a culmination of the ideas, insights, and hnduigs presented in the four earlier chapters. R e c d the purpose of each prerious chapter. In Chapter 1, a general inuoducaon to Web information systems, th&

importance, and need for inrestigaaon as enablers of knowledge work

were espiained. The primaq- objective of this dissertation was described, namely to invescigate ùic

use of Wcb information systems such as corporate portais, depamnental inmanets and the

W'orld Wide iVeb and the& capacin- to facilitate organizational knowledge work.

In Chapter 2, a review of the literature relating to the use of Web information systems for

knowledge work was gn-en. Specificail!-,

an ooerrïea- of the knowledge and knowledge work

consuucts, the p o t e n d of WIS to enable knowledge a-ork, and the barriers which impede the usc of such systems were described. Three prima- tenets were expiicated and used as a basis of

inrcstigation for this research: 1) information phys a cenual role in knowledge work activiq-;

2) Web ïnfomiation sj-stems have the p o t e n d to facilitate organizational knowledge work; and

3) contesmal factors c m inhibit or promote the adoption and use of Web infoxnation systems for howledge work. The Beha~-ioioural/Ecological hfodel was presented as a conceptual frarnework that incorporated the u-orkings of these tenea. The Mode1 heiped set boudaries surroundlig the object under invesagation and served as a theoretical lens from which to glean insights and make discoveries. More specificaüy, the Behavioural/EcoIogica Mode1 emphasized

the need to inform WIS design &om three different perspectives: organizational; user; and interface. Based on these perspecaves, four research questions were fonnulated which centered on an informational analysis of knowledge work practice in organizations.

In Chapter 3, a mcthodological suategy was described which was used to investigate the ualization of Web information systems by several sets of distinct user groups at a Luge tclecommunicauons companj-. The intent was to investigate the estent to which organitationai participants adopted and utilited Web infonnaaon systems for knowledge creation, dismbution, and use. Data collection comprised both qualitative and quantitative techniques: semi-sauctured

intcnlcws, Wcb tracking, document r e ~ e w - ,field obset-~ations,and a questionnaire. Data analysis primanly involred a variant form of grounded thcory; descriptive statistics helped triangulate results. Two phases wcre employed to c a q out this endeavout: a pilot to test the

riability of the study's data collection and andysis methods; and a fidl study comprising other sets of users.

In Chapter 4, the studfs research findings were prcsented utiliting the various perspectives of the dissertation's conceptual framcwork. This helped serve in obtaining both micro and macro orientations to this case study investigation. The information needs and uses perspective focuscd attention on major sets of users, their problem situations and information behaviours in tcrms of actual WIS activity, and the personal characteristics of key \VIS knowledge workers.

The value-added processes perspective concentrated o n participant perceptions o f the design of Web information systems for knowledge work. T h e information needs and uses perspecrire involved a more holistic eiramination of WIS usage within the organizational contest of the

b.

In this fifdi uid &al chapter, ansa-ers to the studp's four research questions are presented and integrated into a new mode1 of Web use describing the organizatiod, user, and interface factors chat deter or foster the use of Web information systems for knowledge creation, distribution, and use. Furthermore, reflections are made on the research process: namely a discussion o n methodologicai, pracacai, and theoretical implications; limitations; and future research directions.

5.2 Answers to Research Questions 5.3.1 Problem Situations and Information Behaviours

Ifz lent15 of pmbfem nhmtiom and infornion bebanom, bow do myor

sets

of

z m n z~ti/iteWeb

iffomzafionytenz~Jorknowhige work?

This question concentrated o n the estent to which Web information sTstems wcre utilized for knowledge creation, dismbution, and use*

set of in formation-intensive processes that

comprise knowledge work. The question pertained to the esamination o f the types and characteristics of problem situations that drew people to use WIS for knowledge work, as well as

th& Web-related information behaiiours: namelv the Web information sources fiequented; the n-ays in which this information was displayed and forrnaned; and how this information was

typicaIiy used. In r e m s of Web information syscem usage, there were large variations in use across the four

WIS npes. Overaii, the World \Vide Web and departmental intranet sites were vïsited the most frcquentlv, alrnost to the same estent. The corporate portal experienced sizeable usage, but to a

lesser degree. Only a few participants utilized \Veb-based groupware. What nras interesthg upon e.uamliing this usage was that different problem types drew participants to unlize different Web information systems. For esarnple, the most common problem type, ~eanhfor in/omation, occurred predominantly with the Worid Wide Web and dcpartmental intranet sites. The second-most common problem type, exjtdore/cbeck ouf, occuned most frequenùy on depamnental intranet sites. n i e third-most common problem type, bmiv~e/or io/0nmfion, occurred equallv across the corporate portal Worid Wide Web, and departmentai i n a n e t sites. The fourth-most comrnon problem type, pnfom a trumacrrion, occuned primarilg on the corporate portal and depamnental incranet sites. The fifth-most common problem type,

dowdoad, occurred most fkequently on departmental intranet sites and the World Wide LVeb. The least common problem q-pe, port/rham injorfion. occurred predominandy on Web-based

groupware.

The use of different W I S for different problem situations was pivotal in shaping the anaiysis of the data. That is, since participants unlized each p a r t i c h \VIS for speufic purposes, hrther hi-cstigation of the problem simations and information behaviours was carrïed out individually for each WIS type. In fact, variant patterns of knowledge work, problem dimensions,

information traits, and information uses were identified across each WIS type.

In terms of the corporate portai, the predominant problem types were browsing for newsrclated information, exploring and checking out the hinctions available o n the portal, and searching for direcrory-related information on other workers in the organization and/or speafic dcpamnental Litranet sites. Knowledge creaaon was a predorninant activïry comprising half of the significant episodcs of activi-. This primarilj- occurred when parficipants browsed for newsrclatcd information and esplored portal features. Howerer, roughly forty percent of portal significant episodes eshibited no evidence of knowledge creaaon, distribution, or use. This

occuned when participants used the portal to obtain directory-based information, a Web address of m intranet site, or post a transaction such as a rîcation e-mail message or update employee records. Thus rwo types of problems for which participants ualuied the corporate portal cshibited eridence of knowledge work, in this case knowledge creation only: the browsine of i ~ of portal feanires. For the browsing of news-related information and the e s m l o ~ d c h e c k lout information, the predomiaant problem dimension pattern involved discouey-rrhted,famiiiar, iowri&,

ri/np/e. d@m, i//-.rtmctur~d situations. The information utilized tended to be cornplex, nameiy

cornprishg hoth hard arzd s@

abta, 60th wbat and w b diagnoriic~, ~ both dzfl~.reand p

e e

iofomation,

qzmiitative, n ~ u i t @ i . . - o i o n e ami , ~'urnnt-Fwther, the information tended to be used for en/,htenn,rnt purposes. For the e q d o ~ g of portal features, the problem dimension patterns , tended to involve un/-amibar, ~~~~~~~~~ihposed, f o w - K J ~n'@le,

and zif-r~mctundsituations. The

information urilized in these episodes tended to be bard, pren'se, q~aktative,muii$/e-rolkoned, and t r m t ~ t As . with the browsing of news information, the infornation tended to be utilized for en/&htenn~e~~t purposes.

Nrith respect to departmentai intranets, the predominant problem types were searching for s~ecificinformation such as Human Resources policies, IS/IT methodolog. guidelines for sysrcms del-elopment, and employee records, and exploring/checkin~our intranet sites to c o n h information, give feedback, lcam functions, see what's new, and/or maintain an intranet site. Participant behaviour on the inaanet sites comprised evidence of ail three knowledge work activities: knowledge creation, dismbution, and use. Howerer, knowledge dismbution was the leasr represcnted of the three with ne+

half of all haanet-related significant episodes showhg

instances of knowledge creation and use only. Most instances of knowledge creation and use occurred when participants used the intranet to search for specific information or explore/check

out a departmental inuanet site's functionaliry and then utilized information kom that site in

some way, such as to make a decirion or create a report. The second-most predominant actirity inrolred knowledge creation only; this occuned when participants browsed for information, esplored/checked out a site, or searcbed for information, and as a result, learned something new but did not disuibute or ucilLe that knowledge direct.. For the searching of information, the predomliant probiem dimension pattern involved &coq-mhted, butb new andfamikar, internafbzhpo~ed,iow-RA,n'mph, rpe+

wefLzttuct~~~pd situations. The information utilized tended to be bardI

niagrru-ch--da/,pnire, quaLifative, ~7ng/e-zofutioned,and nlmnt. Further, the information tended to be used for fa~-t~taf purposes. For the exploring of intrane t feanires, the problem dimension patterns te nded to b e dengn-nfated, new? internai,-imposed. fuw-n'A, rimple,

d'fluse, and if/-xtmctured situations.

The information utilized in these episodes tended to be bard, ppnch-e, qua/itutive, m~dt$fe-~o~u~ionerd,

und mrnnf. Further, the information tended to be utiiized for confirmoona/purposes.

In terms of Web-based groupware. the predominant problem types were the p o s t i n ~ / s h a r i n ~of information, p r i m d y to coordinate and eschange project-related information among team members, and the esdorine/checkine out of feanires and functions araikble \iithLi the groupware. In general, Web-based groupware activity comprised e\-idence of

all dirce knowledge work activities in the folloailig rank order: knowledge creation; knou-ledge distribution; and knowledge use. However, this occurred across taro knowledge work patterns: die most predominant being when participants utilized the groupware to post/share information

which inroIved knowledge creation, distribution, and use; the second most predominant being when pamcipants esplored or checked out the groupw-are applications which involved linowledge creation only. For the posting/sha~g of information, the predomliant problem dunension pattern tended to be dirrovey-maed a

intema/&iqûored, high-rük, J+o&, -'pen;flc-

godtd, uu.rl/~fnufztmi situations. No pattern of information traits was discemable, though the information tended to be used for factzxaf purposes. For the exploring/checki.ng out of

groupware feanires and hinctions, the predominant problem dimension pattern tended to mmpkx-ydz~u~e-goukd, if,-xtmcfuredsituations. The inrolve &ngn-nfatcd, nm, ~~~~~~~~inpozesrd, /orv-n3kl and information utilized in these episodes tended to be bard, p m i ~ e ,qaafitatiwf muft@/e-~ofutioned, izm-ent. Further, the information tended to be ualuied for znxtmmental purposes.

With respect to the World W i d e W e b , the predominant problem type was the searchino for information for primady project-related research purposes, though moderate usage occuned for e s d o ~ ~ / c h e c k i nout e other extemal Web sites for ideas on Web site design and the browsing of information relating to persona1 areas of expertise. Two primar}. knowledge work pattems

were identified with these WWV-reiated sqpificant episodes. The most predominant, knowledge creation, occurced for d three problem types when pdcipants were actively learning about particular topics of interest, the functionaiity offered in other Web sites, and kecping up-to-date. The second predominant knowledge work pattern occxmed for searching for information and explo~g/checkïng out Web sites. These instances involved both

knon-ledge creation and use where pamcipants leamed something upon cisiting the Web sites and then directly applied that knowledge, such as c r e a ~ ga report, making a decision, or designing the& own Web pages. Two predominant problem dimension pattems were identiued

in \Y1%V-rekted

sigmficant episodes pertaining to the searching of information. n i e h s t

pattern involrcd dirrowy-rrfated, new, estmaf~-iqûosL.d, jour tirk, coqdes, ~enzc-goa/ed,i/l-ltmctzmd

problcm siruations. The second involved problem situations &at were discoverq.-rrfated,familiar, r-\-fernaf&hpo~ed,four-tirk, simple, ~eeiic-gouled,und wel-~tnlctured.Search-related significant episodes on the World Wide Web tended to inrolve the presentation of information that was hard,

diagnostic-ivhat,p m i ~ e ,qz,altative, mufip/r-soluloned,and

~ I ~ I I Further, I .

the information tended to

be used prirnarily forfactctaf ppupose~,Mth some instances used for imtmttlentafpurposes.

Table 5.1 below providcs a summary of the knowledge work patterns identified across the

tour W I S types. The table shows in rank order the three predominant patterns of 1) knowledge creaaon; 2) knowledge ueation and use; and 3) knowledge ueation, dismbution, and use. For each pattern, the predorninant problem types and LUS tvpes associated with the knowledge

work pattern are displayed-

Rank

Knowledge work Pattern

Associated Probhm Typos

1

Knowledge Creation Only

r

Search for Information

9

ExplorefCheck Out

Portal; Intranet; Groupware; WWW

Browse for Information

Portal; Intranet; WWW

2

3

Knowledge Creation and Use Knowledge Creation. Distribution and Use

9

Search for Information

a

ExploreICheck Out

PostlShare Information

A8sociat.d WIS Types

O

r,

-

Intranet; WWW

Intranet; WWW Inbanet; WWW Groupware

Table 5.1 :01-erriewof Knowted~eWork Panerns

O r e r d , Table 5.1 shows that participants eshibited Linowledge work activity when utilizing \Veb information systems. However, the primar)r activity concentrated on knowiedge creation O+.

There was less evidence of knowledge use and lesser cvidence of knowledge distribution.

Further, variations in knowledge work occurred across the four \'VIS tlpes: ail WIS types were udized for howledge creation, chough this was the only activity for which the corporate portal \vas utilizcd; boch the \Vorld Wide Web and depamnental intranets were uscd more estensively

in terms of knowlcdge work; and Web-based groupware was utilized across the fidi spcctrum of

lino~vledgework proccsses, though in actuality it was u&ed

by p&cipants the least fiequently

of alI the \VIS types.

Table 5.2 below displays the predominant problem dimensions, information traits, and

information uses by problem type. These problem types are the ones listed in Table 5.1 which showed evidence of knowledge work acavity. The problem dimensions, infornation traits, and information uses displayed in Table 5.2 are a composite list of the more common or teauring characrerisacs identified across the four WIS types. The table shows that different panems of problern dimensions, information traits, and information uses were associated uith each problern type-illustrating

the diverse and comples problem situations for which participants

mm ro \VIS, the varied and rich information sets displayed to users in their quest to resolve these problem situations, and the diverse uses of the i n f o m t i o n obtained off these sustems.

Problem Type

Prodominant Problom Dimensions

Prodominant Information Traits

Pndorninant Information Uses

Search for Information

Discovery, new, low-risk, specific-g oaled

Hard; diagnostic-what; precise; qualitative; current

Factual; Instrumental

ExploreICheck Out

Design. new, intemallyimposed, low-risk, diffusegoaled, illstnidured

Hard; precise; qualitative; multiple soluüoned; current

Enlightenment; Confimational: Instrumental

Discovery, familiar, low-

Hard; diagnostic-what & why; diffuse 8 precise; qualitative, multiplesolutioned. current

Enlightenment

No discemable pattern

Factual

Browse for Information

risk. simple, diffusegoaled, ill-stnictured PosUShare Information

Discovery, familiar, intemally-imposed, highrisk, simple. specificgoaled. well-structured .

-. .. - -

Table 5.2: Predominant Problern Dimensions. Information Traits. Uses bv Problem T m ç

52.2 Information Ecology How dors an otgani~attion 2 infomation eco/ogyferrer or i ~ p e d ethe urr of IYfeb i ~ ~ i sytemr o n for RnowMge work ?

This question dealt Mth the degree to whkh the ~r~anizational contest, from an informational perspectire, inhibited o r promoted the use and engagement of Web information systems for knowledge creation, distribution, and use. In t e m s of this case study investigation, sereral hctors were identified as being more salient aith respect to the use of Web infomation systems for howledge work: information politics; the WIS derelopment process; and the information culture of the organization. First, nith respect to information politics, nvo struggles orer the management of the information displayed on the corporate portal at the case study site highhghted the influence of d of Web informarion systems policical tensions orer information in the design and e ~ - e n use b>- organizational participants. This occurred when intemal groups MrhLi the depamnent

responsible for portal information content could not coilectively agree on the delivery of information and when one stakeholder group controiled the portai's hancial purse suings. This

led to a thal portal design that consisted of duplicated information over t h e e Web sites and O\-er-reprcsentation of services and features on the portal tdored to rhe needs of the stakcholder group controiiing the portal's budget Both of these factors irnpacted the adoption and use of the Web information system by organizational participants; some users indicated thewere conhsed to the duplication of information on the portal and sasr Little value in some of the fcarures proridcd on the portal that were deemed mandatory by the group controlling the portal's budget. Second, in terms of the W I S developmeat process, there were several factors which irnpacted the resultant design of Web information systems and the acceptance of these systems

by organizational participants. For both the portal and depamnental intra.net sites, a thkd party systcm solutions providcr was responsible for f a c i l i t a ~ gthe design and layout of these sites according to pre-detexnined poliaes and design considerations. Users were dissatisfied with the liigh financial cost, slow response time to get modifications doue, and the requirement not to dei-iate from Web design standuds. Pressure Gom senior management for quick-tumaround redesigns, insufficient manpower, and fmanaai Lunitations were identified by informana fvniliar

\\r 33nl Annuai Hawoü Infenzationaf Confmmr on .Fyi+emSrienctv (HICSS), Januaq 4th-7th. hiaui, Hawaü: IEEE. Strauss, -\. (1987). ~zta/itatiwana@ for s o d scientist.s. Cambridge, United Kingdom: Cambedge Cnirersity Press. Strauss, .A., 8: Corbin, J. (1994). Grounded theor)- methodology: An oreniew. In N. De& & Y. Lincoin (Eds.), Handbook o/ 2 u a f i t a t z ~Re.sennb @p. 273-385). Thousand Oaks, California: Sage. Suauss, -A ., 8: Corbin, J . (1998). Bancs of qztafdative m e a d : Techniquex and prvcedz~mfor dedoping grou~tdedrheory. (2nd ed.). Thousand Oaks,California: Sage.

Suchman, L. -\. (19 87). Pfanr and shated adom: The probftm York: Cambridge L'niversitq- Press.

q-hziman-machine communication. New

weahh: Managing and measunhg Rnodedge-baed arseh. San Svciby, K. E. (1997). The new orga~~i;atiotla/ Fransico, California: Berrett-Koehler.

Taylor, R. S. (1986). L'due-aded pmce~Jefin in/omafr'on ytem.s. Norwood, New Jersey: Ables Publishing. Taylor, R. S. (1991). Information use enrironments. In B. Denin & M.J. Voigt (Eds.), P ~ J E JinJ Commmication Stiencer (Vol. 10, pp. 21 7-255). Nom-ood, New Jersey: -1blex Publishing. Tesch, R. (1990). Qztalitattive ~.seamrh= Ana&Lropes and roJwarr toof.. New York: F a h e r .

ToCfler, A. (1990). Powenhifr: Knowledge, weaith and viofence at the edge oJhe Zlst cen/ury. New York: Bantam Books.

Turoff, M.,& Hiltz, S. R. (1998). Supercomectivity. Communicafiom the A C M , 41 0,116 . Yen?. (1999). The Verity corporate portal: Organize your intranet the way you organize your business (available at h~://~.verity.com/solutions/cportal.htfnl). Last accessed January 2000. \'on Glaserfeld, E. (1988). The con.rtm*tion of knowfedge: Contrilurion5 fo ronctpiuai semanhx Salinas, California: Intersystems Publications. \?on Krogh, G. (1998). Care in knowledge creation. CafqomiaManagement ReMéw, 40(8), 133-155

Walsham, G. (1993). In/erprrling infomafion gftems in oqani~a~ions. Chichester: Wiley. \Valsham, G. (1995). The emergence of interpretivism in IS research. Injonnc~tionSystems Re5earh, 6(4), 376-394.

Wikstrom, S., & N o m , R. (1994). Knowfedge and value: A new pu-ectzie on corpurait franrformaliion. New York: Routledge. Williams, F., Rice, R. E., & Rogers, E. M. (1988). Re~ean-hmerho& and the new media. New York: The Free Press.

Wuthnow, R., & Shnun, W. (1983). Iuiowledge workers as a new chss: Structural and ide ological convergence among pro fessional-technicai workers and managers. Work and Onrrpafiom,1 O(l983), 47 1-487.

S & m Quurter-., 26, 58-65. Yin, R. (198 1). Thc case study crisis: Some anm-ers.Adrnini~tra~ive Yin, R. K. (1994). Case stu4 mean-h, dengn und mefhodr. (2nd ed.). Newbury Park, California: Sage Publications. Zubo ff, S. (1988). III the age of the srnari machine. New York: Basic Books.

APPENDIX A: DEFINITIONS Axial Co-

A procedure ualized in grounded theory which estabiishes relationships bem-een categories and subcategories of data.

Case Study: A reseatch strate= which focuses on understanding the dynamics present within a single setting. Involves the observation, description, or reconstruction of a phenomenon of interest. Grounded Theory: A speciflc qualitative coding technique developed by Strauss and Corbin (1998) which identifies key categories and themes surrounding a phenomenon of interestInformation: hleanuigfui data. A commodity capable of yielding knowledge. Information Behaviours: The practices and habits of individuals and groups as they go about obtaining and using information to resolve problem situations. Information Ecology: An interna1 information environment comprising many interdependent social and cultural svstems that influence the creation, flow, and use of information- This c m includc an organization's information system goals, information management plans, information culnue, information politics, and physical settings. Dac-enport (1997) identifies six component of the information ecology: information strate=, information politics, inforrnation culture, information staff, information processes, and information architecture. Information Culture: Comprises a b ' s information behaviours and amtudes towards ~aluing, s h a ~ gdisclosing, , and u&ing information. Information Politics: The human struggle over the management of infonnation. Information Trait: How information is dispiayed. Special atmbutcs that define the way informarion can be idcntified and presented. ?vfach.lullin and Taylor (1984) identifj- nine information trait continuums. Information Use: How individuals utilize i n f o n t i o n to resolve their problems. Taylor (1991) identifies eight classes of information use. Information Use Environment: The set of elements that affect the flow and use of information messages into, within, and out of any definable entity; and determine the criteria bp \\-hich the value of information messages are judged. Taylor (1991) describes four categoxies of the IVE: sets of people, their problems, typical settings, and problem resolutions.

Interpretivism: An underlying philosophical research perspective which assumes realiq- is socialiy constructed through language, consciousness and shared meanings. Under this perspective, researchers attempt to understand a phenomenon under investigation through meanings that people assign to them.

Knowledge: Comprehended information that can exist both tacitly in the rninds of people and csplicidy in formal products and procedures.

Knowledge Creaaon: A knowledge work process in which an individual ueates new knowledge or learns somethiag new. Knowledge Distribution: A knowledge work process in which an individual receives knowledge from others, disseminates knowledge to others, or gathers information to distribute to others.

Knowledge Use: A knowledge work process in which an indiridual uses information to perforrn an action (i.e., create a repon, make a decision), or where an individual utïlizes hrs/her own knowledge to perform an action.

Knowledge Wotk The production and re-production of information. The application of comprehended information. A set of knowiedge creation, distribution, and use processes in d i c h information phys a primaq- role. Mass Information System: A global sjrstern that supports online information retrieval and routine tasks bg way of self-semce for a large number (thousands or millions) of occasional users who are spread over various locations. Open C o w A procedure utilized in grounded theorv where the researcher names and categorizes a phenomenon tfrrough close e.uamination of field data.

Positivism: A n underlying philosophical research perspective which assumes realiq- is objective and can be described by measurable propemes independent of the researcher. Problem Dimension: Characteristics of a problem beyond specific subject matter. Xiachhllin and Taylor (1984) identifj~22 distinct types of problem dimensions. Problem Situation: An instance or contest in which individuals have a perceit-ed information need requiring resolution. Selective Coding: A procedure ueilized in grounded t h e o l where the researcher identifies the core category of the study and systematicaiiy relates it to other major categories identified by the rcscarch.

Value-Added Process: The fanires or hnctions arailable nithin an information sj-stem which signal, ampli.,

and cstend the value of information.

Web Information System: A speafic subcategory of mass information systems that relies on h!pcrrcsr functionality and transfer mechanisms of the World Wide Web.

APPENDIX B: FIRST INTERVIEW QUESTIONS The purpose of this interview is to gain an understanding of the participants7 organizational information ecology, typical problem situations and information behaviours, 3s weil as th& csperiences and perceptions of using Web information systems for knowledge work activity. Interview duration is approsimately 45 minutes to one hour in length.

A) Problem Situations, Information Behaviours, and Information Ecology Tell me about the work vou nmicailv d o in vour iob? PROBES: \mat are your roles? Responsibitities? Duaes? O Who are the colleagues that vou work with? What other groups do you do work with in the organization?

A2) From that descriution. what would vou sav are the main tasks vou nmicallv iierform in vour dailr- work?

A3) In rems of carrvine out these tasks Irepeat these tasks to the participantsl. describe some of the tpical problems vou face that iirompt vou to look for and use information to helii vou resolve these problems? -44) Describe the information sources vou use and the tme of information vou look for and

prefer to heln vou solve these problems? r15) Describe how vou micallv use the information eathered from these sources? PROBES: rr What do you do ulth this information once you found it? How successhl are you in uskg information fkom these sources to resolve your problems? o m a t is the predominant use of this bformation?

-

AG) How does vour work environment affect vour abilitv to access. eatlier. and use this in formation?

PROBES: tyrhat plans or approach does the organization have in managing its information resources and senices? Dcscribe the information culture in tenns of rr information shaxing/hoarding information overload / W r e ~ g r, bureaucraq (bureaucrac). reduces the effect of new information) m information attitude (the degrce to which info iç valued) 0 dornain of information interest (the type of info concentratcd upon) O How is information managed in the organization? (Information politics) 0 How does the physical setting affect your access, gathering, and use of information?

accessibiiity to information priva. of the layout (traffic, noise) size of the s e k g (srnail, medium, large) Who are the people in the organization that help acquire, store, hlter, sort, and distribute information? (information staff) IT people information content people (intemal and extemal) What rules, r o u ~ e sor , processes are in place to help maintain an organizaaonal memory? (e-g., archives, knowledge repositoq, Lonis Notes)

B) Perceptions and Use of W e b Infornation Systems B 1) W i c h Web Information Srstems do TOU u&e? corporate portal World Wide Web depamnental intranet Web-based groupware B2) \mat are vour erso on al motives for usine these Web Information Svstems? PROBES: \mat prompts FOU to use these systems? \mat senices/features do you use? \VIq-? 0 What information do you usuaily use these \ ! S for?

B3) lY/hat features or hnctions would pou like added/chaneed to these srstems that would încrcase vour own motiration to use them? B4) What ~ o a l s / ~ u m o ds eo these \Veb Information Svstems serve? How do thev h e l ~the organization achieve its mission?

C) Web Information Systems and Knowledge Work Activity Cl) M a t would vou sar are the areas of expertise \-ou offer the ornanization? PROBES: (i-e., 'ou may be a computer e-vert, information speQalist, telecommunications ~IU-U) C2) How do vou maintain or foster this ememse? PROBES: How do J-ou keep up to date in this area of espertise? Read? Take courscs? T a k to coiieagues? Library? O Do you coilaborate or communicate with others in the organizaaon about this area of espertise? r Who are these other people with which you share or use your expertise? How do FOU c~rnmunicate/c~Uaborate with these other people (face-to-face? Intranec? E-mail? Workshops? Lunch and L e m s ? )

C3) To what extent have Web Information Svstems hel ed vou in deveIopine - or maintaining this area of esnertise? PROBES: Do you use the kitranet/WWW/exttanet/local intranet serrer to obtain information on this area of expertise? Why/Why aot? 1s thL information up-to-date and pebent? Do you use these WIS to cornmunicate/collaborate with other about this area of expertise? Why/Whv not? In what ways, has \VIS influenced or changed the way you collaborate witb Four colleagues? Has it enabIed or Irmited this collaboration? Esplain. C4) \mat Çeatures o r hnctions wouid you like to see added to Web Information Svstems to supnort vou in vour emett role?

PROBES: To help you leam more about your area of espertise (gain new insights) To help you share your expertise wïth others in the organization To help you develop new products and services in your area of expertise

APPENDIX C: SECOND INTERVIEW QUESTIONS The researcher identified significant episodes of activig in each participants' tracking logs pnor episodes were periods of t r a c h g behaviour that satisfied at to the second interview. S@cant least one of the ioliowing critePa: 1) consumed a substantiai amount of time in the a c b g logs relaare to other episodes; 2) were recurrent activities; or 3) were indicated bg the participant as being significant during the interview itseif. At the beginnuig of the interview, the researcher gave participants the opportuniry to pemse a fomatted peintout of th& Web tracking logs. This prlitout sen-ed as a prop to structure and guide the interview process. For each simiificanr e ~ i s o d ethe followinn questions were asked: Brieflv describe the context behind this episode? Why did )-ou vïsït the sites shown in this episode? \%at problem was being solred? What type of information were you looking for! How did you use the information once it was found?

9

A t the end of the ïntemiew. the foliowine auestions were asked: Can .ou offer some cornments on your Web usage. 1s it u s d y this low/hrgh? How q i c a i was the usage recorded in the logs during the tracking period? Did any unusual events occur that affected your tvpical Web usage pattern?

APPENDIX D: CODING BOOK FOR SIGNIFICANT EPISODES The following table describes the coding schema used to categorize sigrilficant episodes of Web usage identified from the participants' Web tracking logs and/or Netscape History files. An episode was considered "significant" if the episode either consurned a substantiai amount of cime in the tracking logs, was 3 recurrent activig in the logs, or was ident&ed by the participant in the second intemiew as being stgnrticant. Note that the categories (attributes) and th& associated values emerged as the coniag was conducted. hfany of the categories and associated values were based on esisting theoretical constructs such as problem dimensions, information traits, information uses (NacMuilin & Taylor, 1984; Tavlor, 1986). A s d e r percentage of the categories were self-developed by the researcher such as knowledge process types (knowledgc creation, knowledge distribution, knowledge use). Descripaons of the categories are noted in the table. ATTRIBUTE

DESCRIPTION

VALUES

\SIS T I T E

The type of Web Information System for which a significant episode occurred.

1) corporate p o n d 2) depamnental intra.net 3) groupurare 4) World M d e Web

\YIS FE-A'IZ;RE / SITE

The feature utilized or site visited for a significant episode. These d u e s correspond to those whcre W S D T E = "corponte portai".

1) classifieds 2) company faces 3) company sw-ey 4) dialogues 5) directory 6) employee product promotions 7 ) feedback 8) main page 9) news 10) search 11) site categories 12) tirne reporting 13) tools 14) work groups

These values correspond to those where W S TlTE = "departmental intnnct".

1) communications 2) h a n c e 3) human resources 4)

IS/IT

5) marketing 6) orher departmental site 7) project related site 8) purchasing

ATTRIBUTE

DESCRIPTION

VALUES

These values correspond to those where

1) group devdoped tool 2) Teamcast

WTS TITE = "groupware"These values correspond to those where = 'World %'ide Web".

1) association site 2) communications site 3) company related site 4) consumer portal site 5) educationai site 6) govenunent/regulatory site 7 ) marketing site 8) news site 9) project management site 10) technolog?- site 11) telephone directory site 12) trainhg site 13) vendor site

P R O B E N TiTE

The general type of problem for which the WIS was utilized for 3 significant episode.

1) browse for information 2) search for information 3) post/share information 4) e..qdore/check out 5) download 6) perform a transaction

PROBLEM REASOS

The morc specific reason for urilinng the Web Information System for an episode.

tWS n T E

These values correspond to those where PROBLEM TYTE = "brou-se for informaaon". These are episodes where pmidpants perused information for general sense-making and/or knowledge building purposes.

These values correspond to those where PROBLEM TYPE = "search for information". These are episodes where pamapants looked for a specific piece of information to help resolre a problem.

1) association info 2) career info

3) company people/event info 4) company products and senices 5) cxecutive speeches/messages 6) field services info 3 h a n c e info 8) how to optirnize e - m d use 9) news - general 10) news - indus tq-/ business 11) project management info 12) t-endor informauon 13) S X infornation 1) check stock prices

2) do research 3) h d a business/organization 4) h d a person/e~pen 5) h d a speufic site 6) h d out how to get adrets 7 ) foliow methodolog'. guidellies/templaies 8) get company product info 9) get corporate discount info

ATTRIBUTE

DESCRIPTION

VALUES 10) get I T standards info 11) get policy info 12) get vendor product info 13) how to handle crisis situation 1 4 how to renm- mesnbership 15) learn about the Company 16) puü empioyee records

KSOWLEDGE CRE,+TIOX?

These values correspond to those where PROBLEM Tk'PE = "post/share information". These are episodes where partiapants added information to a \VIS.

1) coordliate wlth team 2) give feedback on portal

Thcsc values correspond to thosc where PROBLEM TkTE = "expIore/check out". These are episodcs where partiapants exunined a Web site for noninformation g a t h e ~ purposes. g

1) c o n h info/give feedback 2) Ieam hmcaonality/see what's new 3) maintain/buiid a site

These values correspond to those where PROBLEAl TkTE = "download. These are episodes where partïapants eumcted non-information entities off a \%S (as opposed to dowdoading white papers, reports etc which were categorizcd as cither information browsing or searching).

1) get screen shots (e-g., clipart, iUustrations, screen saver) 2) get software (cg., anti-rirus, Web browser)

These values correspond to those where PROBLE'rl TlTE = "perfotm a transaction". These are episodes where participants perforrned a database uansacuon.

1) arrange location move 2) maintain employee records 3) order a book 4) post an ad 5) request conference c d setup 6) set e-mail vacation message 7) sign-up for training

Knowledgc creation occurred if the partiapant in thc episode: 1) ucilYied the W S to create new knowiedge; 2) used to the K'IS to Iearn something new.

1) "true" implies knowledge creation has occurred 2) "false" irnplies knowledge crmtion has not occurred

Knowledgc distribution occurred if the participant in the episode: 1) uscd the VIS to receive/disserninate knowledge from/to others; 2) ualized the W S to gather mformauon to transmit to others

1) "uue" implies knowledge distribution has occurred 2) "fdse" implies knowlcdge distribution has no t occuned

ATTRIBUTE

DESCRIPTION

VALUES

KSOWLEDGE USE?

Knowiedge use occurred if the partiapant in the episode: 1) used information Gom the PC?S to create sornething (e-g., a report) or perfocm an action (e-g., make a deasion); 2) used his/her own knowiedge to critique a IWS site/tool

1) "me" irnplies knowledge use has occurred 2) LLfalse" ïmplies knowledge use has not occurred

DISCO\'ER\?-"

Design/Discovery problem dimension ~lacMullin&Taylor,1984;TayIor,1986).

1) "me" implies that the problem situation penainstodiscove~(i.e.,hndingsomething) 2) "false" implies thar the problem situation pertauis to design (i-e., inventing somethlig)

Familiar/Neur Pattern problem dimension (3Ia~,\lullïn8r Taylor, 1984; Taylor, 1986).

1) "me" implies chat the problem situation is familiar (ie., procedural, well-established) 2) "false" impiies that the problem situation is new (i-e.,unfamiliar, novel)

ISTERX.ILLY IMPOSED?

Intemal/Extemal Imposition problem dimension PlacJfullin & Taylor, 1984; Taylor, 1986).

1) "me" implies that the problem simation is irnposed internaiiy by the organization 2) "f'se" impiies chat the problem situation is imposed extcrnaiiy outside the organization (Le., competitoa, govemment regdation)

ALiGNITUDE O F RISK GREiT?

Magnitude of Risk G r e d N o t Great problem dimension (Maclfuilui Rc Taylor, 1984; Taylor, 1986).

1) "me" inplies that the consequences of fading to resolve the problern situation are high 2) "false" impiics that the consequences of f&g to resoh-e the problcm situation are low

SIMPLE?

CompledSimple problem dimension (lkc3fuiiin & Taylor, 1984; Taylor, 1986).

1) " m e " impiics that the problem situarion has an easily d e h e d solution path 2) "fdse" implies that the problem situation has numerous and interacting variables

SPECIFIC GO-ILS?

Speafic/-bnorphous G o d s problem dimension ( ' r f a c l f u h & Taylor, 1984; Taylor, 1986).

1) "true" implies rhat the problem situation has well-defined and directed resdts 2) "faise" implies that the problem situation has amorphous, unclear direction

\\EUSTRCcTLXED?

Well-structured/Ill-stmctured problcm dimension (3facMullin & Taylor, 1984; Taylor, 1986).

1) "me" irnplics that the problem situation can be soIved by application of logical or algorithmic processes 2) "false" implies that the problem situation has variables that are not weii undersiood and cannot casily be solved through routine acaon.

4' . i subset of the problcm dimensions dichotomies outiïned bu MacJIullin & Taylor (1981) and Taylor (1986, 1991) were ualizcd to code the significant episodes identified in the tracking logs and discussed by participants in thc second intewicw. The problcm dimensions used were ones that were discemable €rom the data.

288 ATTRIBUTE

DESCRIPTION

VALUES

Parr of the data continuum (MacAIullin & Taylor, 1984; Taylor, 1986).

1) "me" implies rhat the information utilized in the W S episode is empVically derived. 2) "false" implies that the informaaon urilitcd in the iViS episode is not empi8cally derived.

Part of the data continuum (MacLlullLi & Taylor, 19W, Taylor, 1986).

1) " m e " impiies that the information u&ed in the W S episode is drauri from inference2) "fdse" implies chat the information utilued in the WIS episode is not draun from in ference.

Parr of the causal/diagnostic continuum

1) " m e " implies that the infomiation u&cd in the W'IS episode tells what is happening. 2) "faise" impiïes that the information ualized in the WTS episode does not teil what is happening.

(Madlullin & Taylor, 1984; Taylor, 1986).

Part of the causaI/diagnostic continuum (MaclIullin & Taylor, 1984; Ta)rlor, 1986).

FOCCS-DIFFUSE? Pan of the focus continuum (Mac,\lullin & Taylor, 19W, Taylor, 1986).

Pan of the focus continuum (Maclfullin & Taylor, 1984; Taylor, 1986).

Part of the quantitative continuum (lIacMuh & TayIor, 19%; Taylor, 1986).

''

1) "tme" impties that the information uulrzed in the W S episode tells why something happens. 2) "false" implies that the information ueilized in the W S episode does not tell why someching happens.

1) "true" implies that the infornation uuiized in the W S episode is broad and multiperspective. 3) "faise" impiies that the information ucilized in the W S episode is not broad and mulaperspective. 1) " m e " implies that the information utiiized in the WTS episode is exact. 2) "faise" irnpLies that the infomiation utilized in the WIS episode W not =act.

1) "uue" irnplics char the infomiation utilized in the \Y'IS episode is descriptive. 2) "faise" implies that the information utilized in thc \VIS episode is not descriptive.

Separare attributes in the code book are made for each side of Taylor's (1986, pp. 45-47) continuum of information traits since an information item may have characteristics of both (e-g., contain both hard & soft data, quaharive and quantitative informanon etc.). -1s with the problem dimension coding, the information traits utilized wcrc thosc that wcre discernabIe from the data.

ATTRIBUTE

DESCRIPTION

VALUES

QU.LYIIT-3TIl'E;

Psir of the quantitative continuum

1) " m e " impiies that the information utiiized in the W S episode is measurabIe and represrnted numericaily. 2) "fdse" irnplies that the information utilized in the WTS episode is not measurable and represented numcricall-

(Mac~iullin& Taylor, 1984; Taylor, 1986).

Part of the solution continuum (XacJIuiiin & Tarior, 1984; Taylor, 1986).

1) "me" implies that the information utilited in the W S episode provides a number of solutions to the problem situation. 2) "faise" impiïes that the infoxmation unlized in the WTS episode does not proride a number of solutions ro the probkm siruacion.

Part of the solution continuum (Nac,\I& & Taylor, 1984; Taylor, 1986).

1) "me" impiies that the information utilized in the WIS episode provides a single desued best solution. 2) "false" implies that the information utilized in the W S episode does not provide a single desired best solution.

Part of the temporal continuum (Nac,\Iullin & Taylor, 1984; Taylor, 1986).

1) "me" implies chat the infomiaaon utïhed in the WIS episode is present-based. 2) "faise" impiies that the information utiiized in the W S episode is not present-based.

Part of the temporal c o n ~ u u r n (?clacMullin& Taylor, 1984; Taylor, 1986).

1) "me" implies that the information utilized in the W S episode is future-based. 2 "false" impiies tha t the information utilized in the W S episode is not future-based.

TE'rTPOR-U,HISTORIC?

Part of the temporal continuum (3IacJluüi.n & Taylor, 1984; Taylor, 1986).

1) "me" implies that the information urilized in the W S episode is past-based. 2) "fdse" implies that the informaaon uulized in the \WS episode is not past-based.

C0SFiR\lJ3=

Pan of the confirmational information use &ss VayIor, 1991).

1) "me" if the information in the 1WS episode is used to verifj- another piece of info. 2) "false" if the infornation in the W S episode is not used to veri. another piece of information.

ESLIGHTES?

Part of the enlightenment information use class (Taylor, 1991).

1) "uue" if the information in the W S episode is used to develop a context or make sense of a situation. 2) "faise" if the informaaon in the \VIS episode is not used to develop a context or make sense of a situation.

J'Thcse catcgorks are based on Taylor's (1991, p. 230) taxonomy of eight classes of information use. These c3tegories are not murualiy exlusive.

290 ATTRIBUTE

DESCRIPTION

VALUES

Part of the f a c d information use ciass

1) "txue" if the information in the WS episode is used to determine the facts of a phenornenon o r event. 2) "fdse" if the information in the W S episode is not used to determine the facts of a phenornenon o r event.

(Taylor, 1991).

PROJ ECTI\'E?

Part of the instrumental information use class gaylor, 1991).

1) 'mue" if the infonnaaon in the W S episode is used to figue out what to do or how to do something. 2) "falsel' if the information in the W S episode is not used to figure out what to do or how to do something.

Part of the moarationd information use class (Tay4or, 1991).

1) "uue" if the information in the W S episode is used to initiate o r sustain persona1 involvernent. 2) "fdse" if the information in the W S episode is not used to initiate or sustain personal involvernent.

Part of the persona1 or policicd information use class (Taylor, 1991).

1) "me" if the information in the KT3 episode is used to det-elop relationships, enhance status, reputation, or persona1 hilhllment. 2) "fdse" if the information in the W?S episode is not used to develop relationships, enhance status, reputation, or personal W ulfment.

Part of the prob1em understanding information use class (Taylor, 1991).

1) "me" if the infonnaaon in the W S episode is used to develop a better comprehension (in a more specific uïiy than enlightenrnent) of a problem situation. 2) "false: if the information in the W S episode is not used to develop a better comprehension of a problem situation.

Pan of the project information use class (Taylor, 1991).

1) "me" if the information in the \VIS episode is used to predict what is iikely to happen in the future. 2) "false" if the information in the \XTS episode is not used to predict what is likely to happen in the future.

APPENDIX E: QUESTIONNAIRE Participant Number:

YOUR JOB

F I What

is your job title?

What departmnt 1 functional area of the orgaoiution are you in?

F I How long have you been working in your area of occupation?

Yn

YOZTR COMPUTER SKILLS

1 BI

HOWwould you rate yourtclf with respect to your computer background with respect to the other rnembers of your organization? Novice,,

1 BZ 1

Novice/inte~mediate~~ htermediateT3 IntermediateExpertL ExpenZs

How would you rate yourself with respect to your Web expertise witb respect to the other members of your organization? Novice 1,Novicehtemediate

intermediateZ3 htermediate/E~pert=~ ExpertZs

1 ~ 3 1 Horv long have you been using the World Wide Web?

F I

How long have you been using the corporate portal ? YOUR TlME

F I How m n y hours do p u work in a typical work week? 1 ct 1

w

hours

How many hours do you spend using a cornputer in a typical work week?

ho urs

How many hours do you spend using the World Wide Web in a typical work week? hours

F I How many hours do you spend using the corparate portal in a typical work week?

-

Thank you for anmvering the questionnaire

*-

hours

APPENDIX F: RESEARCH OVERVIEW SHEET

UTILIZING WEB INFORMATION SYSTEMS FOR ORGANIZATIONAL KNOWLEDGE WORK STUDY Prinapal Im-esqator: Dr. Chun Wei Choo, clioilfific;.utc~ro~itc~.c~, 416-978-5266 Field Resmcher: Brian Detlor, Ph-D. student, d e t l o r a tis-utoronto-ca, 41 6-871-5873 Faculty of Information Studies, Unirersity of Toronto

Overview of the Research Purpose The iotended research calls for an eoaluation of the use of Web information systems such as intranets and the World Wide Web for organizational knowledge work. The study esamines the information needs and uses of major sets o f users, the organizational environment in which Web information svstems are utilized, and the estent to which the>-facilitate knowledge work activity-namely the sharing of insights and bestpractices, the generation o f new ideas and know-how, and the creation of new products and senices through group coliaboration. Proceduses Sereral different sets of users from \-arious groups wiùiin the Company will participate in the study. For each of these groups, the folloaiog data collection procedures will be utilized: The first utllizes one hour remi-~fn~cf~tred i~ttewzew~ to understand the organlational information environment for each group of Web information systems users being snidied and the& experiences aith using these systems for know-ledge work actirïty.

T h e second analyses indiridual Web usage logs. T h e logs the frequenq of Web usage and the content visited.

n.di

be analped to determine

The third involves field obsen-atioris and documcfit review. The researcher w4.i irnmerse h s e l f in the work contest of each of the groups being studied for a two week period. Obsen-ations of work practice and analysis of corponte documents d be recorded ria field notes. During dus ~o week period, participants may be asked questions on work pracace by the field researcher or request documents for reriew. Such interruptions d i be kcpt to a minimum. The fourth involves n recond-mund of inttmew~and a ques/ionnmmre.T h e sccond-round of inten-iews are required to clmi& the reseaxcher's field obsen-ations and add contcst to the Web episodes recorded in the usage logs. Participants will discuss these events in sufficicnt detaii so that the researcher can understand the motivaelig factors why the Web infornation systems were used and the degree to which the participants were successhil in using the Web to solve their problems. The questionnaire wdl be used to generate descriptive statisacs o n user profle charactelistics of the rarious groups being

studied. Participation in this study is rolunmry. Participants may withdraw from the study at any tirne without penalty and without an? affect to employment

Confïdentiality and Anonymity Ali data will be kept confidentid and anonymous. The data will be coilected in such a way as to prevent an!- association between coiiected data and participants names. Interiiew data, tracking logs, field notes, and questio~airesw-iü be coded with ID numbers to group collected data on individual subjects. To prevent linkage to participant names, consent forms w i U not include I D numbers. Any reference that im-olves particular participants or groups WU be reported in a way that protects individual identiw.

Benefits The proposed study d l aiiow for the capture and understanding of a rich and comprehensive set of data. The goal is to obtaui insights Gom this data on how Web information systems are cunently being used, the ways in which they support knowledge work activity, and the differences in the use of these systems between various groups of users which may be ataibuted to factors of the organizational information entironment. Recommendations will be made on how to improve the design of such systems or d e changes to the organizational context in which they are used.

APPENDIX G: PARTICIPANT CONSENT FORM

W E B INFORMATION SYSTEMS USABUITY STUDY Investigator. B& Detlor, Ph.D. student, dcdor(ii fis.uroronro.sa, Superoisor: Dr. Chun Wei Choo, ch0062 fis.iiroronro.ca Faculty of Information Studies, University of Toronto

Consent Please s g n below if you agree with al1 of the foiloMng statements: 1. 1 have freely volunteered to participate in t h snidy-.

lbe, înd 2. 1 have been infornicd in advance about the nature of the s e , what my tasks d what procedures d be foilowed. 3. 1 have been gken the opportunïty to ask questions and have had rny questions answered to my satisfaction. -1.

I understand that the information 1 provide w d be treated confidentially and with anon\mitx. hIy identity udl not be re\-ealed in the reporüng of the study's results.

5 . 1 am aw-are that 1 have the right to withdraw consent and discontinue participation at an!tirne.

Signature:

Phone: Date:

APPENDIX H:LETTER O F PERMISSION FROM PUBLISHER

Kluwer academic publishers Spuiboulevard 50 P.O. Box 17 3300 AA Dordrecht The Netherlands

Bankers: ABN*WRO. Dordrecht Account Nurrbef 51.YS2.206 Postal Cheque Account Numbef 4447384

Cham&erofCommcr#Oordrecht

1

Dr. Brian Detior Faculty of uiforrnation Studies University of Toronto 140 St. George Street Toronto, Ontario, M5S 3G6 CANADA

C o r n m m Regder Number 23051C97 VAT nurnber: NUX1J3.61.337.801 Tekphone (direct):(0178-

Date €-mail:

63922 10 22 May 2000 Bsrcndinzb-anStra;ilcn~wkap.nI

Ref: Choo, Weh Work,pp. 87 figure 3.3 and pp. 118 figure 4.2,2000. Dear Dr. Detlor: With reference to your request (copy herewith) to reprint material in which Kluwer Academic Publishers control the top-vright, our permission is granted, Free of charge, and at the following conditions: It contains original material which does not carry references to other sources (if material in question appears with credit to another source, authorization from that source is required as well); O Permission is granted for non-exclusive English world rights only. For other languages, please reapply separately for each one required. Permission includes use in an electronic form (CD-ROM). O Full credit (Kluwer Academic Publishers book,journal title, volume, year of publication, page, chapterhrticle title, name(s) of author(s), figure nurnber(s), original copyright notice) is given to the publication in which the matenal was originally published, by adding: with kind permission fiom Kluwer Academic Publishers.

pp mrs. Bèrendina Schermers-van Straalen Rights and Permissions Manager

*Please be certain to include Our reference in al1 correspondence.

APPENDIX 1: SIGNIFICANT EPISODES BY KNOWLEDGE W O U ACTMTY T h e following chart shows the breakdown of W I S usage by knowledge work activity. The data is sorted by WIS type (the portai, departmental inmanet, Web-based groupware, and the \WV,V), Problem Type, Problem Reason, and Feature. The kliQ (Knowledge Work) Indes is a diree character boolean consa-uct that represents the presence of knowledge creation,

distribution, and use activitv respectively. The data is funhered sub-divided bv functional group:

-L-1 (admuiis trative assistants); BA (business analyst s ) ; MM (rniddle managers); PM (project managers); and SD (spstem developers). Sumrnary statistics are provided at the WIS type level.

PORTAL: PROBLEM TYPE

PROBLEM REASON

FEANRE

Browse for Information

Company PeoplelEvent lnfo Execuîke Speeches News IndustryBusiness

Company Faces Dialogues News

-

M BA MM PM SD TOTAL

O 1 1 1 3

1 0 5 0 6

O 0 0 0 0

1 0 1 0 2

O 0 1 0 1

1 12

2 1

8

KW INDEX TFF TFF TFF lTF

ExplorelChecl

Suggest Documents