anonymous knowledge sharing in a virtual environment: a preliminary investigation

11
& Research Article Anonymous Knowledge Sharing in a Virtual Environment: A Preliminary Investigation Simon Burnett * , y and Lorraine Illingworth z The Centre for Knowledge Management, Aberdeen Business School, the Robert Gordon University, Aberdeen, UK This paper presents the results of a small pilot study examining how anonymity may potentially affect knowledge sharing in an organisation. Staff employed in a Scottish university were invited to post responses, anonymously or by including their e-mail address, in relation to a newly introduced car park scheme on campus. One hundred and forty-five staff participated in the study with just under half responding anonymously. The responses were analysed by content and length and the results show that very negative responses are less likely to be anonymous. While a similar proportion of identified and anonymous responses were negative in content, a higher proportion of identified responses supplied very negative responses. While suggested improvements and helpful comments are also less likely to be made anonymously, individuals are more likely to anonymously share or discuss complaints relating to past experiences and issues. In addition, anonymous responses tend to be shorter in length and more to the point than identified responses. Copyright # 2007 John Wiley & Sons, Ltd. INTRODUCTION AND RATIONALE This small preliminary study aimed to test the hypothesis that the ability to act anonymously in virtual environments can encourage individuals to share knowledge. The main objectives of the study were: To examine the literature surrounding knowledge sharing and communication in virtual environments. To examine the differences and/or similarities in the content, nature and quantity of anonymous and identified responses. To determine whether or not anonymous communi- cation in a virtual environment produces more valid responses. To identify the feasibility and potential research approaches for future research in this area. The rapid improvement in information and communication technologies (ICTs) as well as a reduction in their cost has resulted in companies worldwide becoming increasingly reliant on their use to store and share knowledge both internally and externally. Using ICTs (and in particular Internet-based technologies) customers can now secure the use of products and services from companies across the world. Inside organisations, employees communicate with each other to share ideas, solve problems, and answer questions, allowing them to learn on an ongoing basis. Employees can now communicate to individuals Knowledge and Process Management Volume 15 Number 1 pp 1–11 (2008) Published online 21 December 2007 in Wiley InterScience (www.interscience.wiley.com) DOI: 10.1002/kpm.294 *Correspondence to: Simon Burnett, The Centre for Knowledge Management, Aberdeen Business School, the Robert Gordon University, Garthdee Road, Aberdeen, AB10 7QE, UK. E-mail: [email protected] y Director z Research Assistant Copyright # 2007 John Wiley & Sons, Ltd.

Upload: simon-burnett

Post on 15-Jun-2016

216 views

Category:

Documents


2 download

TRANSCRIPT

Page 1: Anonymous knowledge sharing in a virtual environment: a preliminary investigation

Knowledge and Process Management

Volume 15 Number 1 pp 1–11 (2008)

Published online 21 December 2007 in Wiley InterScience

94

(www.interscience.wiley.com) DOI: 10.1002/kpm.2

& Research Article

Anonymous Knowledge Sharing in aVirtual Environment: A PreliminaryInvestigation

Simon Burnett*,y and Lorraine Illingworthz

The Centre for Knowledge Management, Aberdeen Business School, the Robert Gordon University,Aberdeen, UK

*CorManUnivE-mayDirezRese

Cop

This paper presents the results of a small pilot study examining how anonymity may potentiallyaffect knowledge sharing in an organisation. Staff employed in a Scottish universitywere invitedto post responses, anonymously or by including their e-mail address, in relation to a newlyintroduced car park scheme on campus. One hundred and forty-five staff participated in thestudy with just under half responding anonymously. The responses were analysed by contentand length and the results show that very negative responses are less likely to be anonymous.While a similar proportion of identified and anonymous responses were negative in content, ahigher proportion of identified responses supplied very negative responses. While suggestedimprovements and helpful comments are also less likely to be made anonymously, individualsare more likely to anonymously share or discuss complaints relating to past experiences andissues. In addition, anonymous responses tend to be shorter in length andmore to the point thanidentified responses. Copyright # 2007 John Wiley & Sons, Ltd.

INTRODUCTION AND RATIONALE

This small preliminary study aimed to test thehypothesis that the ability to act anonymously invirtual environments can encourage individuals toshare knowledge. The main objectives of the studywere:

� T

o examine the literature surrounding knowledge

sharing and communication in virtual environments.

� T

o examine the differences and/or similarities in

the content, nature and quantity of anonymous and

identified responses.

respondence to: Simon Burnett, The Centre for Knowledgeagement, Aberdeen Business School, the Robert Gordonersity, Garthdee Road, Aberdeen, AB10 7QE, UK.il: [email protected] Assistant

yright # 2007 John Wiley & Sons, Ltd.

� T

o determine whether or not anonymous communi-

cation in a virtual environment produces more valid

responses.

� T

o identify the feasibility and potential research

approaches for future research in this area.

The rapid improvement in information andcommunication technologies (ICTs) as well as areduction in their cost has resulted in companiesworldwide becoming increasingly reliant on theiruse to store and share knowledge both internallyand externally. Using ICTs (and in particularInternet-based technologies) customers can nowsecure the use of products and services fromcompanies across the world. Inside organisations,employees communicate with each other to shareideas, solve problems, and answer questions,allowing them to learn on an ongoing basis.Employees can now communicate to individuals

Page 2: Anonymous knowledge sharing in a virtual environment: a preliminary investigation

RESEARCH ARTICLE Knowledge and Process Management

and/or teams either face-to-face or virtually eitherby e-mail or using the Internet. People can usetechnology ‘to exchange pleasantries and argue, engagein intellectual discourse, conduct commerce, exchangeknowledge, share emotional support, make plans, brain-storm, gossip, feud, fall in love, find friends, and losethem, play games, flirt, create a little high art and a lot ofidle talk’ (Rheingold, 1998). As such this ‘virtualcommunity’ enables people to communicate andshare knowledge with each other without movingfrom their workstation.

It has been suggested that these relationshipscombinedwith the ability tomanage knowledge canbe an important element in the generation ofcompetitive advantage for organisations of all types(DTI, 1998, p3). Because of the nature of the Internetand similar technologies they allow anonymouscommunication to take place within a variety ofcontexts which can be seen as unique and useful(Teich in Oakes, 1999). Marx (1999) believes that tobe fully anonymous a person cannot be recognisedby the following types of identity knowledge: legalname; localitability; pseudonyms linked to name orlocation and those that are not linked to name orlocation; pattern knowledge; social categorisation;and symbols of eligibility/non-eligibility. In rela-tion to anonymous communication in the work-place, recent research discovered that employeeswere more likely to communicate anonymouslywith regard to complaints and suggestions aboutthe organisation and management; complaints andsuggestions about co-workers; whistle-blowing;organisational surveys and performance feedback(Scott and Rains, 2005, p. 170).

The knowledge economy itself is predicated on arenewed understanding of the value of knowledge,most particularly within an organisational context.In order to create value and competitive advantage,this important asset must be shared and transferredthroughout the organisation. Although knowledgemay be shared using a variety of methods a numberof barriers can exist and people can be unwilling toshare knowledge for many reasons. Kinsey Gorman(2002) identifies a number of such instances whichinclude: unwillingness to admit to making amistake; wishing to avoid being seen to contradictor disagree with management; staff are insecureabout their value of their knowledge; staff can beway of negative consequences; lack of trust; knowl-edge is power; and colleagues withhold what theyknow. As a consequence, organisations may notbenefit from gathering valuable pieces of knowl-edge which can be used to improve performance.Furthermore, there are a variety of barriers tosharing knowledge at work. Pawar et al. (2001)identify a number of ‘human barriers’ to knowledge

2

transfer, which include: language; perceived inter-national differences; accuracy of knowledge; fear ofpenalty; fear of losing company stability; protectionof proprietary knowledge; maintenance of com-munication channels and time. Similarly, Daven-port and Prusak (1998) list a range of knowledgetransfer ‘frictions’. Included amongst these are: lackof trust; different cultures; lack of time; knowledgeowners receive status and intolerance of mistakes.

Research has found that the majority of organ-isations recognise the value of knowledge sharing(Dixon, 2000; Hall, 2000; Pawar et al., 2001).However, these studies uncovered certain attitudesand behaviours in relation to knowledge sharing.Indeed, from Dixon’s extensive research intoknowledge sharing in organisations, she discoveredthat this process does differ from organisation toorganisation. Dixon therefore suggests three factorswhich can determine successful knowledge trans-fer: the receiver, the nature of task and the typeof knowledge transferred (Dixon, 2000). Whilefocusing mainly on the technology side of knowl-edge sharing, Hall (2000) found that relationshipsbetween people affected the success of the knowl-edge exchange. This study also found that someemployees preferred to share knowledge throughtheir own networks rather than use the technicalsystems set up to enable knowledge sharingthroughout the organisation (Hall, 2000).

The value of informal processes for sharingknowledge, such as the coffee machine approach,is widely recognised, and is arguablymore beneficialthan more formal processes (Davenport and Prusak,1998). The informality in this context is used to meanad hoc, impromptu situations or events, which takeplace as result of a serendipitous meeting betweentwo or more individuals, for example discussing awork-related issue at a coffee break. Perhaps as aconsequence of this, organisations have sought toreplicate these arrangements within a digital orvirtual environment. However, organisations maybe missing out on valuable opportunities to allowindividuals to share knowledge by divorcing them-selves from specific knowledge objects.

One method of potentially addressing the unwill-ingness of individuals to share knowledge andopenly express their views is through the creation ofa virtual or digital environment whereby individ-uals may share knowledge anonymously withoutfear of retribution. Indeed, Marx suggests anonym-ity is important in order to ‘facilitate the flow ofinformation and communication on public issues’ and‘to encourage reporting, information seeking, commu-nicating, sharing’ (Marx, 2001).

Virtual working spaces are common in manyorganisations (Kimble et al., 2000), and increasingly

S. Burnett and L. IllingworthDOI: 10.1002/kpm

Page 3: Anonymous knowledge sharing in a virtual environment: a preliminary investigation

Knowledge and Process Management RESEARCH ARTICLE

common are virtual spaces for more informalpractices or discussions. From the perspective ofthe knowledge provider, in a virtual environmentthere can be several valid reasons why employeesshare knowledge anonymously. The majority ofrespondents of a study conducted by the AmericanAssociation for the Advancement of Science had‘very positive experiences in communicating anon-ymously’ with some believing it to be an essentialpart of their work (Kling et al., 1999, p. 82).Anonymity can offer a safe and secure method ofsharing knowledge without any accountability(Scott and Rains, 2005). Employees can assumedifferent roles and hide potentially discriminatingfactors such as race, gender and disability. As aresult employees may feel free to express opinionswithout fear of ridicule or consequences and areperhaps more likely to reach out for assistance.

Nissenbaum (1999) also stresses that ‘. . .the valueof anonymity lies not in the capacity to be unnamed, butin the possibility of acting or participating whileremaining out of reach, remaining unreachable’ (Nis-senbaum, 1999, p. 142). Nissenbaum suggests thenthat by remaining anonymous employees may stillcontribute and acquire valuable knowledge, andthat in fact this environment may in fact makethem more able to manage these processes as theorganisation itself is rendered unable to identifyspecific individuals who in the past they may havechallenged. This is true so far as individuals areable to remain truly anonymous. Nissenbaumalso suggests that due to the advances in ICT andits ability to track people through fragments ofinformation it is no longer sufficient to withhold aperson’s name in order to become anonymous(Nissenbaum, 1999) and as a result individuals mayin fact be more cautious of sharing knowledge thanthey would be in a physical environment.

Linked to the idea of unreachability and answer-ability is the suggestion that anonymity encourageshonesty amongst employees. As they may feel‘untouchable’ in a virtual environment, knowledgeproviders within an organisational setting may bemore likely to provide more candid, honest andtruthful views and opinions. Research into people’sevaluations regarding general anonymity foundthat those who were obliged to identify themselvesweremore likely to provide positive appraisals thanthose who were required to reply anonymously(Antonioni, 1994; Fuller, 1974) who also suggeststhat when respondents of surveys believe the studyis indeed anonymous this will result in honestresponses. This study seems to suggest thatanonymous expressions of opinion are more likelyto be honest than those expressions made throughidentified sources. Although this would appear to

Anonymous Knowledge SharingDOI: 10.1002/kpm

support the view of anonymous environmentssupporting the transfer of knowledge, there is noindication as to the content of the knowledge objectitself. Indeed Schultz (1953) suggests that anonym-ity can result in ‘a decrease in fullness of content’(Schultz, 1953, p. 18).

Lack of knowledge sharing within an organis-ation can be caused by reluctance to exhibit aperceived lack of knowledge on certain areas.Providing knowledge without revealing identitiescan improve knowledge sharing by overcomingemployees’ unwillingness to ask questions whichthey would not have challenged under otherconditions. This in turn allows employees to becomepart of a community as well as participating in anopen and honest exchange of knowledge. Marx(1999) identifies several rationales for anonymitysuch as the facilitation of the flow of information;to protect an individual’s time, space and person;to protect reputations; to avoid persecution and toencourage reporting, information seeking and self-help.

From the perspective of the user, how does ano-nymity affect the knowledge sharing process andthe perception of shared knowledge? Again, as inthe case of the provider, there are both positive andnegative effects. If anonymity encourages contri-bution on the part of the provider by reducing oreliminating inhibiting factors, then this can be seento benefit the user by providing him with a widerrange of sources than he might otherwise haveaccess to. The wider the range of sources, the richerand more diverse they are likely to be; anonymitymay encourage contribution from individuals whoare not recognised experts on a subject, but whonevertheless may have a worthwhile contributionto make. This contribution will offer a differentperspective from that of the recognised subjectexpert. Such contributions may only be preferredanonymously, as individuals may feel inhibitedabout communicating on subjects they feel lessqualified in, particularly in the presence of recog-nised ‘experts’. This effect has been demonstrated inthe context of electronic brainstorming (Kay, 1995).Similarly, input that may be regarded as contro-versial will only be forthcoming through guaran-teed anonymity. This wider range of contributionsmeans that the knowledge contributed is morerepresentative of what is available throughout theorganisation as a whole, rather than representingthe availability from the usual sources. It is oftenthis sort of unstructured and impulsive contributionof knowledge from a previously unused source thatleads to innovation. All contributors are viewed asequals in the context of an organisation; this feelingof equality might make the user feel more part of

3

Page 4: Anonymous knowledge sharing in a virtual environment: a preliminary investigation

RESEARCH ARTICLE Knowledge and Process Management

the organisation and in addition equal to all otheremployees within the organisation.

Some people argue that anonymity may result inthe contribution of a wider and more diverse rangeof perceived knowledge which may in turn lead toan overload situation. Linked to this is another maindisadvantage of anonymous communication: lackof quality control. With known sources, informationabout the source of the knowledge and thecontributor’s expertise can serve as an indicatorof quality. Without this, knowledge can only bejudged on its intrinsic value. Some categories ofknowledge, such as technical knowledge, do requiresome guarantee of quality before the knowledge canbe safely applied. Lack of accountability can makethe contributor less careful about what he expressesand so rather than being more reliable, in somecases the knowledge contributed is likely to be lessreliable. Furthermore, with no referencing, it isdifficult to track down the originator of a pieceof knowledge, in order to clarify ambiguity or ifthe knowledge proves useful, to request furthercontributions from the same source. However, anonline environment, providing anonymity andmedia leanness, is particularly conducive to elicit-ing honest communication by reducing factors,which lead to ‘Communication Apprehension’(Montoya-Weiss et al., 1998).

The literature review aimed to determine issuesrelating to knowledge sharing in organisations invirtual environments. It is evident from the reviewthat there is a lack of research specifically looking atthe content and nature of the knowledge sharedanonymously. It is therefore anticipated that thissmall preliminary study may be the initial stage ofdeveloping better understanding of this subject.

METHODOLOGY

The Robert Gordon University had introduced carpark charges for both staff and students. Thesecharges had been met with opposition and resist-ance from a large portion of the Universitycommunity. As it was a very controversial move,it was anticipated that this topic would encourageparticipation and it was therefore deemed as anappropriate topic for discussion for the purposes ofthis research.

In order to obtain anonymous responses fromstaff members an online discussion area wasdeveloped with the assistance of the Transportand Travel Co-ordinator which was available forone month to allow all staff the opportunity toparticipate. It was developed not only to gather thecomments of RGU transport policy from all RGU

4

staff but also to gain an understanding of the typesof responses individuals communicate anon-ymously. On receiving an e-mail from the TransportCo-ordinator, staff was asked to access a link whichwould take them to page on the University’sTransport and Travel website. The sample consistedof members of staff within the University’sAberdeen Business School, approximately 466,covering five schools over four sites within theUniversity and included academic staff, researchstaff, and support and administrative staff.

The interface for the online anonymous environ-ment, as shown in Figure 1, is web based whichprovides easy access to the survey via the Internet orthe Intranet. On following the link included in thee-mail message, participants are met with a screenwhich requires them to express their opinion to thefollowing question: in what way has the new carparking system affected your work? Their response isadded to the comments box and they are thenprovided with two options: to either submit theirresponse anonymously or by entering their e-mailaddress and therefore revealing their identity. Oncethey select their chosen option, they either submittheir response if it is anonymous by selecting the‘submit response’ button, or enter their e-mailaddress prior to submitting their response if theywish to be identified. On submitting the response,the comments are sent via e-mail to the inbox of amember of the research team for collation andanalysis.

The discussion/questionnaire area was dissemi-nated by e-mail to a total of 466 eligible Universitystaff based in four sites of the University and 145valid responses were received. The 31.1% responseachieved for this study was indicative of a goodresponse rate and well within the parameters for arelevant analysis to be achieved, and meaningfulfindings to be obtained.

Once the online environment was taken offlineand the questionnaires returned, the data werecollated and analysed by:

� C

ontent (either by sharing experiences, asking

questions or offering suggestions and categorised

from very positive through to very negative com-

ments).

� Q

uantity (measured in the number of lines of text of

the response).

FINDINGS

The main aims of this small study were: to identifythe differences and/or similarities in the contentand nature between anonymous and identified

S. Burnett and L. IllingworthDOI: 10.1002/kpm

Page 5: Anonymous knowledge sharing in a virtual environment: a preliminary investigation

Figure 1 Online anonymous questionnaire

Knowledge and Process Management RESEARCH ARTICLE

responses; to determine whether or not anonymouscommunication in a virtual environment producesmore valuable responses; and to identify thefeasibility and potential research approaches forfuture research in this area. The results are outlinebelow by content and quantity.

Content

Sixty-five of 145 or 44.8% were anonymousresponses. Three different types of responsesrelating to the car park scheme were identifiedand these comprised: sharing experiences/com-plaints; asking questions; and/or offering sugges-tions. Of the 145 responses, 139 participants took theopportunity to share experiences of the car parkarrangement while an additional 13 respondentsasked questions relating to the introduction of thescheme and a further 31 response offered sugges-tions to improve the situation. Responses receivedfrom certain individuals contained a combination ofthese three responses and as a result there are moretypes of responses than there are respondents.

Anonymous Knowledge SharingDOI: 10.1002/kpm

Of the anonymous responses the majority wereshared experiences, issues, and difficulties (95.3% or62 of 65). Equally, a similar proportion of identifiedresponses related to individuals’ experiences, with96.35% or 77 of 80 sharing issues, problems andexperiences relating to the car park scheme. Incontrast, only a small proportion of both anon-ymous and identified responses asked questionswith 10% of the total identified responses and 7.6%of anonymous responses.

While the majority of anonymous and identifiedresponses used the survey to share their experi-ences, respondents who either asked questionsrelating to the parking or offered suggestions toimprove the situation were more likely to includetheir e-mail addresses in their responses. Indeed, agreater number of identified responses (25% or 20 of80) offered suggestions on how to improve thecar-parking situation, the figure for anonymousresponses being lower at 16.9% (11 of 65). It cantherefore be suggested that helpful and positivecomments/suggestions are more likely to be namedand perhaps issues and problems of a difficult

5

Page 6: Anonymous knowledge sharing in a virtual environment: a preliminary investigation

Figure 2 Number of anonymous and identified responses

RESEARCH ARTICLE Knowledge and Process Management

nature are more likely to be shared or discussedanonymously.

For the response type known as shared experi-ences, marginally less of these were received with-out the individuals’ names and e-mail addresses;44.6% (62 of 139) were without identification whilethe remaining 55.4% (77 of 139) were identified. Interms of the remaining types of responses themargin is slightly greater with just over a third ofthe offered suggestions (11 of 31 or 35.5%) and theasked questions (5 of 13 or 38.5%) responsesincluding an identifier (see Figure 2 for a break-down).

Figure 3 Content

6

In order to identify differences in nature eachresponse was arranged into the following sixcategories:

� V

of

ery positive (very pleased with new car park

arrangement);

� P

ositive (pleased with new car park arrangement);

� P

art positive part negative (respondents who are

pleased personally, but displeased on the effect the

new car park arrangement has on colleagues);

� I

ndifferent (not affected by new car park arrange-

ment);

� N

egative (displeased with new car park arrangement);

responses

S. Burnett and L. IllingworthDOI: 10.1002/kpm

Page 7: Anonymous knowledge sharing in a virtual environment: a preliminary investigation

Table 1 Type of questionnaire responses by content

COMMENT Anonymous responses Identified responses

Description No. ofresponses

% of total % ofanonymous

No. ofresponses

% of total % ofidentified

Very positive 1 100 1.5 0 0 0Positive 1 50 1.5 1 50 1.3Part positive part negative 6 46.1 9.2 7 53.8 8.8Indifferent 6 40 9.2 9 60 11.2Negative 44 49.4 67.7 45 50.6 56.3Very negative 7 28 10.8 18 72 22.5

Knowledge and Process Management RESEARCH ARTICLE

� V

ery Negative (very displeased with new car park

arrangement, to such an extent that in some cases

respondents are engaged in seeking alternative

employment).

Due to the subject matter it may not be surprisingto note that the majority of the 145 responses wereeither very negative or negativewhich indicates thatmany respondents were either displeased or verydispleased with the new car park arrangements.Indeed, 61.3% of all responses (89 of 145) werenegative while 17.2% (25 of 145) were very negative.Of the remaining responses only 2.06% were eitherpositive or very positive while 8.9% responses werepart positive/negative and 10.3% were viewed asindifferent responses (see Figure 3 for more details).

Half of these negative responses were anon-ymous (44 of 89) while in contrast only 28% (7 of 25)of the very negative responses were not sent with anidentifying e-mail address. Similarly, just under halfof the part positive/negative responses (6 of 13) andtwo thirds of the indifferent responses (6 of 9)were received anonymously. Interestingly, the verypositive response was also received anonymously.In addition, the two positive responses werereceived from motorcycle users while the indiffer-

Figure 4 Type of responses

Anonymous Knowledge SharingDOI: 10.1002/kpm

ent responses were mainly provided by non-drivers, public transport users and/or priority passholders. Of the total 65 anonymous responses, 78%were negative or very negative (51 of 65) while asimilar percentage of the identified responses werealso negative (78.5% or 63 of 80). However, therewas a higher percentage of identified participantswho supplied very negative comments 22.5% (18 of80) as opposed to only 10.8% of the anonymousresponses (7 of 65), see Table 1.

The majority of all three types of responses werenegative or very negative comments. For example,80.6% or 25 of 31 of those who had offeredsuggestions and 80.5% or 112 of 139 who hadshared experiences responded either negatively orvery negatively. In contrast, however, a smallerproportion of those who had asked questionsresponded negatively or very negatively (69.2%or 9 of 13) as indicated in Figure 4.

Of the 62 anonymous shared experienceresponses, 80.6% (50 of 62) were negative or verynegative while 80.5% (62 of 77) of the identifiedshared experience responses were also negative.Similar percentages were also evident in terms ofthe anonymous (81.1%) and identified (80%) offeredsuggestions responses as well as anonymous asked

by nature of responses

7

Page 8: Anonymous knowledge sharing in a virtual environment: a preliminary investigation

Table 2 Length of responses

Length Anonymous responses Identified responses

Number oflines of text

No. of responses % of total % of anonymous No. of responses % of total % of identified

1–2 17 48.5 26.1 18 51.5 22.53–5 24 54.5 36.9 20 45.5 256–10 15 35.7 23.1 27 64.3 33.711–15 5 41.6 7.7 7 58.4 8.7516–31 4 33.3 6.2 8 66.6 10

RESEARCH ARTICLE Knowledge and Process Management

questions responses (80%). However, in terms of theidentified asked questions responses, 75% werevery negative as opposed 40% of the anonymousasked questions responses. In addition, 40% of verynegative identified offered suggestions and none ofthe anonymous responses offered very negativesuggestions.

Quantity

As the questionnaire contained only one openended question, the quantity of response wasdetermined by the number of lines of text fromthe response. The majority of responses were shortin length. Indeed, only 16.6% of responses werebetween eleven and thirty-one lines in length (24 of145) while 30.3% were three to five lines in length(44 of 145), 29% between six to ten lines (42 of 145)and finally 24.1% (35 of 145) between one to twolines of text in length.

It seems that identified responses are on averagelonger in length while the anonymous responseswere on average less than five lines of text. 36.9% ofanonymous responses (24 of 65) were six or more

Figure 5 Types of respon

8

lines of text, whilst this figure rose to 52.5% foridentified responses (42 of 80). In contrast, 63.1%(41 of 65) of all anonymous responses were betweenone and five lines in length as opposed to only 47.5%(38 of 80) of identified responses. Of the responseswhich were one to two lines of text in length 48.5%were anonymous as were 54.5% of three to five lineresponses; 35.7% of six to ten lines in length; 41.6%of eleven to fifteen line responses and 33.3% ofsixteen to thirty-one lines of text (see Table 2 for abreakdown).

It became clear that those individuals who sharedtheir experiences of the car park scheme were morelikely than those asking questions or offeringsuggestions to use fewer words (as illustrated inFigure 5). Indeed, just over half of all sharingexperiences responses (53.3%) were between oneand five lines in length while in contrast, only23.07% of asked question responses and 29.1% ofoffering suggestions were between one and fivelines of text.

The majority of asked questions responses(38.46%) and offering suggestions (35.4%) wereactually six to ten lines in length. In fact, these types

ses by length of text

S. Burnett and L. IllingworthDOI: 10.1002/kpm

Page 9: Anonymous knowledge sharing in a virtual environment: a preliminary investigation

Figure 6 Number of responses by length of text

Knowledge and Process Management RESEARCH ARTICLE

of responses had a higher proportion of responseswith more than eleven lines of text than the sharedexperiences responses. 38.3% of asked questionsresponses and 35.4% of offering suggestions asopposed to only 16.5% of shared experiences werebetween eleven and thirty-one lines of text.

The majority of the indifferent responses wereunder five lines of text (93.3% or 14 of 15), while60.6% (54 of 89) of all negative responses, 53.8% (7 of13) of all part positive/negative responses and only12% of very negative responses (3 of 25) were of asimilar length. Indeed, very negative responseswere more likely to be longer in length. Forexample, 64% (16 of 25) of these particular responseswere between six and fifteen lines in length and 24%(6 of 25) of all very negative responses were betweensixteen and thirty-one lines of text. Please seeFigure 6 for a breakdown of indifferent, negativeand very negative responses by the length ofresponses.

On average, identified responses are more likelythan anonymous responses to be longer in length,whereas the anonymous responses tend to beshorter and indeed more direct. In this context,negative responses are also likely to be shorter thanthe very negative responses. Indeed, the verynegative responses which mainly consisted ofcomplaints, tended to be between eleven andthirty-one lines in length. It also seems thatrespondents require fewer words to share theirexperiences and problems as opposed to thoseresponses which offer suggestions or ask questions.

CONCLUSIONS AND DISCUSSION

A total of 145 responses were received withidentified responses being slightly more numerous

Anonymous Knowledge SharingDOI: 10.1002/kpm

than anonymous responses. As a result of theanalysis of the responses, a number of conclusionscan be drawn.

Somewhat surprisingly, in this context verynegative responses are less likely to be anonymous.While a similar proportion of identified andanonymous responses were negative, a higherproportion of identified responses supplied verynegative responses which mainly consisted ofcomplaints. In spite of this, due to the fact thatthe majority of both the anonymous and identifiedresponses were complaints, it is difficult to suggestwhether or not these findings reiterate those of astudy conducted by Scott and Rains (2005).

However, in contrast to suggestions made byFuller (1974) and Antonioni (1994) the results of thisstudy do seem to imply that the honest responses,the majority of which in this context are eithernegative or very negative, are more likely to beexpressed in a way that the source can be identified.Very negative experiences and responsesweremorelikely to be received by those who had chosen tosubmit their name and e-mail address. Therefore, inconflict with Antonioni’s (1994) findings, this studyindicates that those who identified themselves bye-mail address provided less ‘positive appraisals’regarding the car park scheme. Despite this,suggested improvements to the car park schemewere more likely to be received by namedindividuals.

In light of the contentious topic of the study, itis interesting to note a large proportion of verynegative comments from identified respondents,who it seems, felt the need to respond in thisway despite the fact that being identified couldpossibly result in recriminations. This might bepartially explained by the fact that this topic fallsslightly outside the scope of the being directly work

9

Page 10: Anonymous knowledge sharing in a virtual environment: a preliminary investigation

RESEARCH ARTICLE Knowledge and Process Management

related, and as such people feel there would be noretribution or fear of ridicule from supervisors, andso are less inhibited about stating their feelingspublicly (Marx, 1999).

Due to the large proportion of negative and verynegative responses from identified sources, thestudy also seems in some ways to contradictNissenbaum’s belief that employees are morecautious in the virtual environment than theywould be in a physical environment. Despite this,the small margin between the number of anon-ymous and identified responses may be explainedby the idea that people are aware of the abilityto track individuals through computer systems(Nissenbaum, 1999) and can therefore be guardedand suspicious of the guarantee of anonymity.

While the study suggests that an anonymousenvironment can perhaps encourage knowledgesharing particularly relating to past experiences,and complaints, the study was inconclusive as tosuggesting a potential value for the anonymousknowledge sharing environments. It is difficultto suggest on the basis of this study that ananonymous virtual environment can encouragepeople whomay not normally participate in sharingwith others to be involved and to share theirknowledge, thoughts and experiences. This isinconclusive but had there been no anonymousoption in this study, it is quite possible that someof the anonymous respondents would not haveparticipated.

FURTHER RESEARCH

It is evident from the review of literature that thereis a lack of research specifically looking at knowl-edge sharing anonymously in virtual environmentsand as such several issues and research ideas haveemerged from the literature review and from theresults of this small study.

Little is known to what extent people inorganisations fear retribution or ridicule duringthe knowledge exchange process or whether or notemployees can feel equal in communicating at workanonymously. As it was outside the scope of studyto identify participants’ motivation in completingthe form and submitting the responses anon-ymously, further research is required to identifythe motives behind anonymous knowledge sharingbehaviour and the reasons for choosing to shareknowledge in a virtual and anonymous way. Inorder to fully determine the value of anonymity onknowledge sharing within a virtual environment itis anticipated that further follow-up research isrequired to establish why and when individuals

10

communicate and share knowledge anonymouslyand indeedwhat they gain from this type of sharing.It is anticipated that a further in-depth study wouldalso help to determine whether or not employeeshave positive experiences of sharing knowledgeanonymously as suggested by Kling et al. (1999),whether anonymous sharing mainly consists ofcomplaints and criticisms as indicated by Scott andRains (2005) and whether or not there is a need forthis type of knowledge exchange.

REFERENCES

Antonioni D. 1994. The effect of feedback accountabilityon upward appraisal ratings. Personnel Psychology 47:349–357.

Davenport T, Prusak L. 1998. Working Knowledge: HowOrganisations Manage What They Know. HarvardBusiness School Press: Boston.

Dixon N. 2000. Common Knowledge: How Companies Thriveby Sharing What They Know. Harvard Business SchoolPress: Boston.

Fuller C. 1974. Effect of anonymity on return rate andresponse bias in a mail survey. Journal of Applied Psy-chology 59: 292–296.

Hall H. 2000. Exploring knowledge sharing in distributedorganizations: report on research in progress. http://www.dcs.napier.ac.uk/�hazelh/esis/hazel3.pdf [08June 2007].

Kay G. 1995. Effective meetings through electronicbrainstorming. Journal of Management Development 14:4–25.

Kimble C, Li F, Barlow A. 2000. Effective Virtual Teamsthrough Communities of Practice (Department ofManagement Science Research Paper Series, 00/9),University of Strathclyde, Strathclyde, UK, 2000.

Kinsey Gorman C. 2002. Five reasons why peopledon’t tell what they know. Destination KM. http://www.providersedge.com/docs/km_articles/Five_Reasons_People_Dont_Tell_What_They_Know.pdf [08June 2007].

Kling R, Lee Y, Teich A, Frankel A. 1999. Assessinganonymous communication on the Internet: policydeliberations. The Information Society 15: 79–90.

Marx G. 1999. What’s in a name? Some reflections on thesociology of anonymity. The Information Society 15:99–112.

Marx G. 2001. Identity and anonymity: some conceptualdistinctions and issues for research. http://web. mit.edu/gtmarx/www/identity.html [08 June 2007].

Montoya-Weiss M, Massey A, Clapper D. 1998. Onlinefocus groups: conceptual issues and a research tool. TheEuropean Journal of Marketing 32: 713–723.

Nissenbaum H. 1999. The meaning of anonymity in aninformation age. The Information Society 15: 141–144.

Pawar K, Horton A, Gupta A, Wunram M, Barson R,Weber F. 2001. Inter-organisational knowledge man-agement: focus on human barriers in the telecommu-nications industry. http://www.ppc.biba.uni-bremen.de/projects/corma/download/CE2001.pdf [08 June2007].

Rheingold H. 1998. The virtual community http://www.rheingold.com/vc/book [08 June 2007].

S. Burnett and L. IllingworthDOI: 10.1002/kpm

Page 11: Anonymous knowledge sharing in a virtual environment: a preliminary investigation

Knowledge and Process Management RESEARCH ARTICLE

Schultz. 1953. Common-sense and scientific interpretationin human action. In Philosophy and PhenomenologicalResearch 14: 1–38.

Scott C, Rains S. 2005. Anonymous communication inorganizations. Assessing use and appropriateness.Management Communication Quarterly 19: 157–197.

Anonymous Knowledge SharingDOI: 10.1002/kpm

Teich A, in Oakes C. 1999. Study: online anonymitycritical. http://www.wired.com/politics/law/news/1999/06/20480 [08 June 2007].

The department of trade and industry. Our competitivefuture: building the knowledge driven economy1998. http://www.consumer.gov.uk/comp/main.htm[08 June 2007].

11