Understanding contribution of content to public document repositories
Mani Subramani
(msubramani@csom.umn.edu)
Naren Peddibhotla
(npeddibhotla@csom.umn.edu)
Information and Decision Sciences Department
Carlson School of Management
University of Minnesota
Minneapolis, MN 55455
Last revised: April, 2005
Abstract
There are many publicly accessible repositories on the Internet that are created by firms
and professional bodies. Such repositories are created by the uncompensated efforts of
individuals contributing content e.g. book reviews or review of attractions at tourist
destinations for the benefit of unknown others who may be considering reading the books
or planning visits to these destinations. While the potential value of such publicly
accessible online repositories is recognized, there has been little prior work examining
the motivations of individuals to make contributions to them. We draw on social
psychological theory and theory on prosocial behavior and volunteerism to examine this
phenomenon and highlight the factors motivating individuals to make contributions to
online repositories. We examine these questions using data on the profiles of
contributors of reviews to a publicly accessible review repository. We find that
contributors have a varied set of motivations for contributing – altruism, reciprocity,
social affiliation, utilitarian benefits and enjoyment - that have also been seen in other
online contexts. In addition, we observed that people also contributed to repositories
because it gave them an opportunity for self-development. We also saw that a potential
contributor will be more likely to contribute in an area with fewer prior contributions, and
that repositories include content both from experts and non-experts.
Keywords: Document repositories, qualitative data analysis, and prosocial behavior.
1
Understanding contribution of content to public document repositories
Introduction
A variety of public repositories of content in different subject areas are available on the
Internet. Such repositories hold knowledge in the form of documents contributed by
various participants. Some of these repositories involve contributors who do not have an
affiliation to the same organization – e.g., medical practitioners contributing reviews at
www.cochrane.org, individuals contributing knowledge on various topics at
www.wikipedia.org and researchers offering solutions to problems faced by underserved
communities at www.thinkcycle.org. There are also repositories built by business
organizations to facilitate and promote usage of their own products or services or those of
third parties – e.g., reviews of holiday cruises at www.cruisereviews.com, product
reviews at www.dooyoo.co.uk and www.amazon.com. The content in such repositories
are often considered central to the value delivered by online retailers.
In the IS literature, researchers have focused on participant behavior in online forums
such as an email network (Constant et al. 1996), an email listserv (Butler 2001) and
bulletin-boards (Wasko and Faraj 2000). Compared to such forums, the act of
contributing to a repository has both similar and some distinguishing characteristics.
Like email and bulletin boards there are few, if any, social cues (Sproull and Kiesler,
1986) as compared to face-to-face interactions. Moreover, extrinsic rewards for
contribution are often nonexistent and at best, minimal. However, unlike email and
bulletin boards, repository contributions are usually made without any appeals or
requests of help for information. In addition, contributors usually get very little feedback
on whether and how their contributions are helpful to others. While repositories usually
2
provide mechanisms for users viewing the contributed documents to provide feedback on
the quality of the material, providing such feedback is almost always voluntary and it is
often not provided.
Contributors play a critical role in the success of such repositories by contributing content
to them. A greater understanding of user motivations to make such contributions is thus a
very important requirement for their successful implementation. Therefore, our study
addresses the research question is: what are the factors motivating individuals to make
repository contributions? We conduct an exploratory study by suggesting some
propositions based on prior theory and analyze qualitative data in the form of the profiles
of reviewers at a large online public product review repository.
The rest of the paper is organized as follows. In the next section, we present a few
theoretical perspectives that can throw some light on the phenomenon and present our
propositions to address our research question. Following that, we describe our research
methodology including the research site, data collection and analysis methods. Then, we
present the results of our analysis along with our interpretations. We conclude with a
discussion of our findings, limitations of our study and suggestions for future research.
Theoretical perspectives
The content contributed to a repository can be considered a public good since once it is
given to a person, it is costless to provide it to everyone else (Bergstrom et al,. 1986). It is
also difficult to prevent someone who does not pay for the content from using it. An
individual may engage in free-riding by never contributing to the repository but at the
3
same time enjoying the contributions made by others. We therefore first explore how the
contribution of content might be examined using the public goods framework.
According to economic theory, a rational person will contribute to a repository only if he
is provided incentives to compensate for his effort. More importantly, such incentives
will work only when the person is recruited for the task, when he is involved in a stable
relationship with the repository and when his performance can be monitored (Olson
1967). In public repositories, however, there is no active recruiting. Moreover, because of
the large number of participants, it is not possible for a central authority to monitor each
individual and provide tailor-made incentives to induce contribution. As against this
individualistic view of the contributor acting alone, other researchers consider the
contributor as interacting with others within a social context. Public goods theory
suggests that a rational person will not contribute unless he has at least an expectation
that his efforts would be reciprocated by others (Thorn and Connolly 1987). The broader
literature has also identified conditions under which such exchange will occur. However,
the features of contexts studied in prior research are not commonly found in public
repositories. Prior research has found, for example, that social exchange will occur when
individuals know each other, at least in a limited way (Takahashi 2000). We next turn to
prior research in the IS literature using the lens of a social context with respect to
knowledge sharing in email networks and Usenet bulletin boards.
Examining the use of email for help seeking and giving within one organization, Constant
et al. (1996) suggest that citizenship behavior, the desire to benefit the organization was a
motive for helping in the online forums created by organizations. In a public repository,
4
however, the participants are not affiliated to any organization and so would not have the
commitment to an organizational purpose that employees would have. The study of
online helping in Internet newsgroups by Wasko and Faraj (2000) suggests that altruism,
generalized reciprocity and community interest created by ongoing interaction of the
members of these online groups are important motivations. However, unlike in the case
of repositories, participants of a bulletin-board share an understanding and common
interests in the topic that the group is devoted to. For the most part, this occurs because a
bulletin board is created with an intention toward discussion on a specific topic and there
are norms on what kinds of questions can be posted there. In a document repository, on
the other hand, contributors can come from different backgrounds and have different
subject interests.
Contributing to a document repository leads to benefits for others whether or not there are
any benefits for the contributing individual. A rich stream of research in social
psychology has studied such a class of activities called prosocial behavior. Prosocial
behavior covers the broad range of actions intended to benefit one or more people other
than oneself – behaviors such as helping, comforting, sharing, and cooperation (Batson,
1998). Using dispositional and situational factors, and drawing on established social
psychological theories, prior studies have examined different types of planned and
spontaneous prosocial behavior (see Batson 1998 for a complete review). See table 1
below for a comparison between different types of prosocial behavior and repository
contribution.
5
Versus
1
Spontaneous helping (e.g.,
helping an accident victim)
2
Planned helping involving
kinship or long-term friendship
(e.g., shopping for groceries)
Contributing time and money
(e.g., volunteering, donating
money to charity)
3
4
Contributing knowledge to a
person (e.g., teaching a person,
replying to an email query)
Contributing content
to a repository
Involves immediate
Involves planning of
reaction to a situation one’s time and effort.
requiring help.
Help is specifically
Help is not
asked for
specifically asked for
What is given away
is gone
Knowledge is given
only to the person
you choose.
(Table 1)
Content that is
contributed can be
used in more than one
place for more than
one purpose
Anybody can access
your contribution.
Much of the research on prosocial behavior has studied situations in which an individual
is unexpectedly called upon to help for a brief amount of time, often called spontaneous
helping (e.g., passengers helping an individual who falls in a subway car, Piliavin et al.
1975). The literature has also examined planned helping in the context of kinship
relationships, and other situations such as donating to one’s alma mater, donating blood
and volunteering. We draw upon results from some of this research on prosocial behavior
and prior IS research on helping in online contexts to formulate our understanding of
contribution to public repositories.
Altruism is one of the factors that have been highlighted as motivating helping in the
context of volunteering (Clary et al. 1998). Altruism is defined as helping without
expecting a direct reward. It has been seen as an important motivator for participants at
online forums such as email networks and bulletin boards (Constant et al. 1996; Wasko
and Faraj 2000). Altruism is valued by individuals and contributing to repositories might
6
provide a means to express such a value. People might contribute in order to simply help
others by giving them the knowledge they possess. Therefore, our first proposition is –
Proposition 1: Repository contributions are likely to be driven by altruistic motives.
While helping driven by altruism reflects helping without direct benefits, another factor –
career interests – suggests that helpers provide help because of the recognition of the
broader link between helping and the advancement of their careers (Clary et al. 1998),
and increase in their own economic utility. The positive reputational benefits such as
recognition by peers could also play a role in motivating contributions as well (Shamir
1991). Prior research on online forums has found evidence of such motives driving
participation (Wasko and Faraj 2000; Lakhani and Wolf 2005). Thus we also expect that
people would contribute to public repositories because they see an opportunity to earn
direct rewards (monetary or otherwise), to earn a reputation or to leverage the activity for
benefits in their jobs. We club these driving forces under the term “utilitarian” motives
and propose that:
Proposition 2: Individuals are likely to contribute to repositories because of utilitarian
benefits they obtain.
Social factors such as the norm of reciprocity and gratitude towards the collective for
assistance received in the past also influence individuals to help, for example in a context
such as donating to one’s alma mater (Diamond and Kashyap, 1997). The need for social
affiliation and the need for professional self expression are the other social factors linked
to helping (Constant, Kiesler, Sproull 1994). Such motives have been observed in online
7
forums such as email networks (Constant et al. 1996), bulletin boards (Wasko and Faraj
2000) and open source software development (Lakhani and Wolf 2005). In public
repositories, people may similarly contribute in order to reciprocate other contributors or
socially bond with them by doing similar activities.
Proposition 3: Repository contributions of individuals will be driven by considerations of
generalized reciprocity and social affiliation.
Intrinsic motivation that provides enjoyment for an individual has also been identified as
a reason why people engage in a certain activity, such as helping (Deci, 1972). Intrinsic
motivation has been defined indirectly by excluding that which is not extrinsic, i.e.,
something that is within an individual. In the broad literature in this area, intrinsic
motivation is synonymous with the playful enjoyment that an individual feels when
performing an activity (Deci and Ryan 1985). In open source software development, for
example, people contribute bug fixes partly because of the thrill in writing code to solve a
specific problem (Lakhani and Wolf 2005). A similar set of intrinsic motives seems to
drive participation in email networks (Constant et al. 1996; Butler et al. 2002) and
bulletin boards (Wasko and Faraj 2000). In the case of public repositories too, people
might contribute because they enjoy submitting documents containing what they know.
The joy could potentially arise from constructing arguments and composing prose in the
text of the document. So we propose that:
Proposition 4: Individuals are likely to contribute to repositories if they enjoy the
activity.
8
An individual may not actually provide help in a given situation despite being motivated
to do so for any of the above reasons. Studies of factors that lead individuals to step
forward to provide assistance suggest that the nature of assistance provided by
individuals is linked to their perception of a sense of responsibility to act (e.g.,
Berkowitz, 1978). Experiments in situations, where onlookers witnessed incidents where
they could step forward to help, indicate that the sense of responsibility that any one
person feels for helping is inversely related to the number of other potential helpers (e.g.,
Piliavin et al. 1975). The overall conclusion of these studies is that the presence of
potential helpers diffuses the level of responsibility any individual feels to provide
assistance. This has been termed the bystander effect (Latane and Darley, 1970).
Similarly in public repositories, an individual might be less inclined to contribute on a
topic on which there are many existing contributions.
Proposition 5: The less the number of other contributions to a particular domain, a
potential contributor is more likely to write a contribution in that same domain.
Finally, prior research on prosocial behavior has also shown that people with greater
competence were the people who helped more (e.g., Clary and Orenstein, 1991). The
mere perception of competence in an individual had been found to increase the likelihood
of helping. Promoting the belief that one has the ability to help another individual
facilitated the expression of empathy for that individual (Barnett et al. 1985). Empathy
has been identified as a key antecedent of the altruistic motive (Batson et al 1981), which
in turn leads to helping. Discussions of knowledge repositories within organizations
consider it almost axiomatic that contributions would be made by experts (Davenport and
9
Prusak 1998). Similarly, we might also expect that for public repositories, only the
knowledgeable or competent individuals would step forward to contribute.
We therefore advance our final proposition:
Proposition 6: Individuals making contributions to repositories are likely to be content
experts.
Research methodology
Research site
We examined the above issues using qualitative archived data on reviewers at the product
review repository at Amazon.com. By focusing on contributions to one repository, we
expect to eliminate confounds from the characteristics of the user interface and features
provided by different technology underlying repositories.
The Amazon.com repository is visited by over 14 million users every month. Customer
reviews of products displayed on the screen are just one click away – drawn from the
online repository by a search when the user clicks for more information. Amazon
provides the facilities for any individual to sign up and contribute reviews of books,
music, videos and other products sold on the site. In writing a book or music review and
posting it on the Amazon.com site, reviewers share their opinion on these products with
other users of the repository – information that they believe may be useful in helping the
users decide if the book or CD is worth reading or listening to. The reviews are
moderated – Amazon.com employs a small group of editors who delete inappropriate or
offending content from reviews. Amazon.com allows a reviewer to develop and track a
referent group by creating a “Favorite People” list. When one of those people in this list
10
writes a review or recommends something, Amazon puts it in the individual’s customized
“Friends and Favorites” home page. This allows the reviewer to keep track of members in
his or her referent group and their opinions. Amazon also provides a bulletin board for
discussion among reviewers giving a reviewer and his referent group another forum to
communicate.
Contributing reviewing to the Amazon repository is not compensated – it is entirely
voluntary. However, Amazon ranks reviewers using a composite of the number of
reviews submitted and the average number of helpful votes received by reviews. The
possibility of recognition as a valued reviewer represents the only formal incentive
offered to contributors. Amazon.com also provides all reviewers the facility to create a
personal page on the site with a photograph and up to 4000 characters of information on
themselves.
Data collection and analysis
We collected detailed profiles of the set comprising the most prolific reviewers and the
most helpful reviewers (the top 1000). The profiles contained rich personal details that
contributors revealed about themselves. This often included personal details such as
where they lived, how old they were, details about their families and pets, descriptions of
their professional careers, information on their hobbies, their interests, their passions and
pet peeves, the factors motivating them to write reviews, personal life history, their
favorite books and the music that they liked 1 . Providing profile information is voluntary,
some contributors had opted to provide only their name and nothing else. Most
1
Amazon.com provides a standard template for the profile information but most contributors sampled had
chosen to use their own format.
11
contributors however provided a broad range of details about themselves and their
motivations. A representative profile is in Appendix –1.
We also conducted semi-structured telephonic interviews with two of the top 50
reviewers (see Appendix 3 for the interview questions). We obtained the email addresses
of these two reviewers from their posted profiles, and requested them for an interview
after explaining the purpose of our study. Each interview lasted about an hour. We
recorded one of the interviews, and transcribed both soon after the interview was
completed.
Next, we read each profile and coded them into categories. The names of the reviewers
were recorded going top to bottom and the categories were noted going left to right.
In analyzing the data, we followed the techniques of open coding and axial coding
advocated by Strauss and Corbin (1998). We used open coding to categorize the text in
the reviewer profiles into categories suggested by prior theory. We identified keywords
suggesting the different categories of motives such as, for example, reciprocity and
enriched this set with keywords we encountered in instances of behavior driven by such
motives (see appendix 4 for the rules we followed for coding responses into categories
derived from prior theory). We also used explanations suggested by the data in the
profiles. In creating categories derived from the data, we often backtracked to earlier
profiles to check if any of those profiles could be coded into a new category created
based on a new profile.
12
Second, after coding the data, we grouped the categories that reflected similar concepts
and themes, consistent with the notion of axial coding. These steps highlighted the core
phenomenon of the motivations underlying contribution of reviews and we used the
linkage between the categories to infer the theoretical explanation.
An alternate coder then coded a random set of profile texts drawn from the sample
independently using the categories drawn up by the first coder. This was done to verify
that those profile texts could indeed be classified under the categories chosen by the first
coder. Next, the alternate coder examined a random set of categories and examined the
profiles listed under each of the categories to confirm if the profiles warranted
classification under that category.
These procedures also ensured validity (ensuring accurate interpretations), and reliability
(consistency of the interpretations) of the conclusions drawn from the data (Yin, 1994).
Results
Descriptive statistics
At the time of the data collection, we counted 814,948 people who had contributed at
least one review at Amazon. Of these we studied the profiles of the top 1009 reviewers 2 .
The descriptive statistics of the population as wells as our sample of Amazon reviewers
are as follows:
2
We had set out to sample the top 1000 reviewers. However, many individuals shared the same position at
rank number 986. We considered all these individuals and thus ended up having 1009 reviewers in our
sample.
13
Number of
reviewers
Number of
reviews
Number of
helpful votes
Population
814,948
Sample
1009
Total
2,082,765
152,065
Mean per
reviewer
Median per
reviewer
Mode per
reviewer
Highest per
reviewer
Lowest per
reviewer
Std. deviation
Total
2.56
150.71
1
107
1
64
3146
3146
1
18
13.34
7,898,974
186.29
892,383
9.69
884.42
0
612
0
563
21872
21872
0
192
299.70
1306.47
Mean per
reviewer
Median per
reviewer
Mode per
reviewer
Highest per
reviewer
Lowest per
reviewer
Std. deviation
(Table 2)
From the above table we see that the individuals in our sample are very prolific
contributors, with an average number of reviews per person that is about sixty times the
entire population average. The top five reviewers each contributed over 1000 reviews. On
the other hand, most of the individuals in the population are “one-review” contributors
with no new contributions after their initial effort. Unlike these casual contributors, the
individuals in our sample have contributed a minimum of eighteen reviews.
14
The reviewers in our sample are also contributors of the most useful reviews, with an
average number of helpful votes per person that is about ninety times the average per
person for the entire population. Among our sample of top 1009 contributors, reviews
received an average of about six helpful votes per review, with a minimum of about one
helpful vote per review on average per individual. In contrast, the reviews contributed by
most individuals in the population have not been rated as useful by others.
Even among the top 1009 reviewers, there is a large variability both in the number of
contributed reviews, and the number of helpful votes per reviewer. Amazon determines a
reviewer’s rank based on a combination of the number of contributed reviews, the
number of helpful votes and the number of “not-helpful” votes. So it is possible for an
individual who has contributed more reviews than another person to be ranked lower than
the latter because of fewer helpful votes. Whereas an individual can control the number
and quality of reviews he contributes, the number of helpful votes is determined by
others.
Out of the entire population of reviewers, 82,569 (10.13%) had written profiles about
themselves, as compared to 900 (89.20%) in our sample. Each profile in our sample
contained 115.46 words on average, with the maximum being 735 words.
The following table shows the background of some of the reviewers in our sample of top
1009 contributors, based on our reading of their profiles. From table 3, it is evident that
the contributors in our sample of top 1009 reviewers came from a wide variety of
backgrounds, and are not biased toward a narrow group in terms of what they do apart
from writing reviews.
15
Background / profession
Priest / evangelist / aspiring priest
Performing artist / performing arts related
job / aspiring to such a vocation
Moviemaker / aspiring moviemaker
Musician (both amateur and professional) /
aspiring to be a musician
Armed forces
Librarian / library aide
Psychiatrist / Psychotherapist / Counseling
Publishing industry / music industry /
literary business / aspire to be in one of
these
Owner / worker in a bookstore / video store
Academic job / home school teaching /
aspiring teacher
Planning for teaching career
Journalist / journalism related work /
aspiring to be a journalist
Editor
Lawyer / legal assistant / aspiring to be in
legal field
Tourist guide / travel agent
Consultant
Trainer / instructor
Mentoring role
Nurse
Commentator
Involved in book club / literary group /
music forum / writers’ workshops
Writer of books / plays / songs
Writer in newspapers / periodicals / other
publications; technical writer / screenplay
writer
Made music CDs
Web content writer
Actor / graphic artist / artist of any kind /
aspiring to be an artist
Employee of Amazon
Number (% of reviewer sample)
18 (1.78%)
22 (2.18%)
8 (0.79%)
36 (3.57%)
16 (1.59%)
15 (1.49%)
18 (1.78%)
28 (2.78%)
7 (0.69%)
90 (8.92)
12 (1.19%)
28 (2.78%)
34 (3.37%)
18 (1.78%)
2 (0.2%)
33 (3.27%)
16 (1.59%)
2 (0.2%)
3 (0.3%)
2 (0.2%)
23 (2.28%)
74 (7.33%)
96 (9.51%)
5 (0.50%)
74 (7.33%)
21 (2.08%)
1 (0.1%)
(Note: Some reviewers might fall in more than one category above)
(Table 3)
Motivations to contribute
16
Altruistic motive:
Our coding of the reviewer profiles in the sample revealed the following altruistic
motives:
S. No.
Type of response / observation
1
2
Number of
occurrences
1
5
Percentage of
occurrences (%)
0.1
0.50
“Become noble”
“Because it is my duty / because
of social compassion”
3
Share experience derived from the 58
5.75
book
4
To promote the reading habit
3
0.3
5
Enabling others to pick and
77
7.63
choose books, movies etc.
Total
136
13.48
(Note: The total figure above does not equal the sum of occurrences of individual
categories since some profiles contain more than one of the above types of responses.)
(Table 4)
By recommending or criticizing products sold on Amazon, a reviewer is saving the
readers of his / her review the cost involved in searching for good products. By sharing
one’s experience with the use of a product, a reviewer is making it easier for a reader of
his / her review to make a purchase decision. A person who is encouraging others to read
in general is showing potential readers the value they can gain from doing so. An
individual who is contributing reviews is thus providing a benefit to others. In their
profiles, such people who expressed the sentiment of helping did not indicate any
apparent benefit they obtained in return for contributing reviews. At an abstract level, a
reviewer sees this altruistic motive as being directed toward the entire community of
potential readers of his review. The following excerpt from a reviewer profile shows an
example of such an expression of altruism:
17
“Wanting to help is the primary reason I write book reviews on Amazon.com. Most
people don't have enough time to read so I want to be sure that people know which books
can help them, and which cannot, and for what purposes. I also like to point out the
bigger questions that sometimes the books and other reviewers fail to articulate.”
The data in the above table thus show that a small proportion of individuals (13.48%)
contribute reviews due to altruistic motives. This proportion is similar to the prevalence
of altruistic motives in a study by Wasko and Faraj (2000), who found 9.8% of
participants of online networks being driven by altruism. Similarly, participants in
another online network studied by Constant et al. (1996) gave a 16% weight to altruistic
motives among all motives for replying to email queries.
One of the reviewers whom we interviewed disclosed an altruistic motive when he said:
“When I write a review, I am saying: ‘here is my opinion’. I am trying to help others in a
purchase decision.” Knowing that his reviews were being found useful was important to
him: “I check my reviews everyday to see my vote count. It gives me an indication of
someone having read it.”
In addition, what’s interesting is that this reviewer indicated that the altruistic motive got
him started into contributing reviews, though it seems to have declined in importance as
time passed by: “Contributing reviews is like a social act. Is it altruism? Partly, I also
write reviews to motivate others to write reviews. In my initial days of writing reviews, it
was more altruistic. Later you get free stuff for writing reviews.” This suggests that
individuals perhaps need not have sustained altruistic motivation to be willing to make
18
sustained contributions to a repository. It might be enough to altruistic occasionally.
Other motivations can make it worthwhile for continuing to contribute.
The other reviewer whom we interviewed made a distinction between simply giving an
opinion in a review vs. helping others in their purchase decision: “Reviewing for a
popular book such as a Tom Clancy novel is an expression of opinion, a part of an
ongoing discussion. This is not of value in the same way as reviewing children’s books
where I am interested in telling teachers what you will get in a particular book.” In other
words, altruism is not just about doing something (giving opinions, criticizing, expressing
feeling about a product) that one would otherwise do in other contexts (e.g., among
friends). It is more about going out of one’s way and making an effort to address
someone’s need (explaining what’s in a product for a potential user, giving a balanced
view).
As a counterpoint to the above findings in computer-mediated communication (CMC)
contexts, we can examine the role of altruism in situations that are very similar to
contributing to a repository. For example, volunteering involves individuals actively
seeking out opportunities to help, committing time and effort to the activity, and
continuing with helping others over a long period of time. Clary et al. (1998) identify six
functions that volunteering activity can serve for an individual, one of which relates to
individuals expressing values related to altruistic and humanitarian concern for others.
They find that the relative weight of altruistic motivation (measured by the mean score on
their volunteer function inventory scale) as compared to other motivations is 25.37% for
people actively involved in volunteering and 20.75% for those who are not. Though these
19
are approximations of the relative strength of the altruistic motive, we note that this
motive is more important in volunteering contexts as compared to repositories.
Contributing to a repository is also similar to alumni making donations to their alma
mater. It involves a deliberate and planned decision to help anonymous others by
providing funding to an institution over time. In their survey of alumni, Diamond and
Kashyap (1997) found that the perceived need of the alma mater has a relative weight of
25.81% as compared to other motives for donating to their alma mater. The perceived
need of the recipient as felt by an individual results in the altruistic motivation according
to the empathy-altruism hypothesis (Batson et al., 1981). We again see the higher
incidence of the altruistic motive with respect to other motives here as compared to a
repository context.
Utilitarian motive:
In contrast to altruistic motives, fewer reviewers stated utilitarian motives to explain why
they wrote reviews, as shown in table 5:
Type of response / observation
In order to obtain free copy of
book
To see one’s writing in print
For monetary gain
To be recognized as a reviewer by
authors by including blurb on
book
To promote one’s business
To be recognized a top reviewer /
expert in a particular field
Total
Number of
occurrences
3
Percentage of
occurrences
0.3
3
1
1
0.3
0.1
0.1
11
10
1.09
0.99
29
2.87
(Table 5)
20
Each of the above motives involves obtaining a material benefit in return for contributing
reviews. The benefit could be tangible (in the form of free products, or money), or
reputational wherein the reputation leads to tangible benefits in other ways to the
contributor. A socially undesirable quality such as selfishness would not tend to be
concealed as a motive when writing one’s profile because of response anonymity
(Aquilino 1998). So we can consider 2.87% of the individuals to be contributing in
exchange for personal gain. This proportion corresponds to the 3.1% of participants in the
study by Wasko and Faraj (2000), who indicated personal gain as a reason for
contributing to a Usenet group. It is also similar to the figure of 5.7%, which is the weight
given to selfishness among other reasons for answer queries by participants of another
online network studied by Constant et al. (1996). In their study on volunteering, Clary et
al. (1998) found that active volunteers rated the importance of the “career functional
motive” (that gives career-related benefits) as 10.80% among various motivations to
engage in volunteering. This figure was 17.54% for individuals who were not involved in
volunteering. The low number of occurrences of utilitarian motives in our study suggests
that altruism is more predominant compared to utilitarian motives in determining the
willingness of a reviewer to contribute.
Thus, we find that altruistic motives can be a potential reason why people contribute to a
repository though to a lesser extent as compared to traditional helping contexts, and that
they might be more important than utilitarian motives.
21
Reciprocity:
The reciprocity effect is also evident in the profiles of the top 1000 reviewers (see table 6
below). Individuals indicated that they wrote reviews in return for the benefit they had
received from other reviewers in the past. We coded reciprocity as a motivation by
mentions of norms of conduct, obligations and trust and from statements reflecting affect
towards the community.
Type of response / observation
Number of
occurrences
42
Percentage of
occurrences (%)
4.16
Benefited from reviews written
by others
Helping others in repayment for
13
1.29
help received in the past
Writing reviews in the
1
0.1
expectation that others will help
me
Total
49
4.86
(Note: The total figure above does not equal the sum of occurrences of individual
categories since some profiles contain more than one of the above types of responses.)
(Table 6)
The above responses indicate that an individual can contribute a review either because he
had benefited from reviews written by others in the past, or in order to be able to receive
the benefits of reviews that will be written by others in the future. It is also apparent that
an individual contributes not for the benefit of another person from whose review he had
benefited in the past, but for the potential benefit of all the participants in the review
repository.
As some reviewers admitted in their profiles:
“I got started reviewing because I enjoyed reading lots of other people's reviews to
choose a book.”
22
“I have consulted Amazon's public reviews for years before making a purchase and I
decided to start giving back to the Amazon community”
“I know I read these reviews prior to buying any book and they have been excellent help,
so if I can steer someone to one they will enjoy, well, then I've paid my dues.”
One of the reviewers whom we interviewed disclosed the reciprocity motive for
contributing reviews when he said: “I browse reviews written by others when I am buying
something. So, in a way, I write reviews because of the altruistic efforts of others
benefiting me.” Beyond merely reacting to the benefits he received from the reviews
contributed by others, he was also active in encouraging others to contribute: “I also
write reviews to motivate others to write reviews.” This suggests that an individual might
contribute to not only reciprocate others whose contributions he benefited from, but also
to potentially benefit others who might be expected to reciprocate him in the future.
However, only a small percentage of individuals indicate a reciprocity motive for their
contributions. There is thus little evidence of people contributing owing to a sense of
reciprocity to other reviewers. Further, this is to a lesser extent than found in past
research on helping in online communities. Wasko and Faraj (2000) found that 13.4% of
participants of Usenet groups participated out of a sense of reciprocity. Similarly,
Constant et al. (1996) had found that members of an email network gave a weight of
11.8% among reasons for responding to requests for help.
Reciprocity has also been studied as a motivation in non-CMC contexts such as alumni
making donations to their alma mater. Diamond and Kashyap (1997) found reciprocity
23
has a relative weight of 13.98% as compared to other motivations in determining an
individual’s obligation to support his alma mater in various ways. Again, we see a much
more important role for reciprocity in determining a willingness to contribute, as
compared to the context of a repository. This is perhaps indicative of the distinction
drawn by Yamagishi and Cook (1993), who compared “group-generalized” exchange vs.
“network-generalized” exchange and found greater sharing of resources in the latter than
in the former. In group-generalized exchange, group members contribute to a common
pool and each member receive benefits as a result of this pooling. In a community of
three individuals “A”, “B” and “C”, for example, A’s giving of resources to the pool
would make the resources available to both “B” and “C”. “A” has no direct interaction
with either “B” or “C”. Contributing to a repository has this feature as the knowledge that
is contributed once can be used repeatedly. In the network-generalized exchange,
participant “A” gives resources directly to another individual “B” who does not repay
“A” directly. Individual “A” instead receives resources from a third participant “C”, who
received resources from “B”. In this particular exchange cycle, A’s resources are given
only to “B” and not to the entire group. Similarly for B’s giving of resources to “C” and
C’s giving of resources to “A”. Donating to one’s alma mater, responding to queries on a
newsgroup or email network all have features similar to this form of exchange.
Finally, it is interesting to note that while reciprocity has been important in research
dealing with public goods (e.g., Thorn and Connolly, 1987), our findings show that it is
not a major factor in explaining why people contribute reviews.
Bonding with others (social affiliation):
24
The reviewers also considered it important to bond with a community of reviewers,
customers, publishers and authors as mentioned in the following reasons for reviewing
given in their profiles:
Type of response / observation
Number of
occurrences
49
Percentage of
occurrences
4.86
Bond with others with similar
interests in books / music /
movies etc.
Bond with others with similar
2
0.2
experiences NOT connected with
books / music / movies etc.
Reviewing to receive feedback
129
12.78
Share experience derived from a
58
5.75
product*
Influence the lives / thoughts of
9
0.89
others
Contribute to the literary cycle
1
0.1
To do service to authors with
4
0.40
praise and feedback
Total
208
20.61
(Note: a. The observation marked “*” has also been interpreted as an altruistic motive in
table 4 above, b. The total figure above does not equal the sum of occurrences of
individual categories since some profiles contain more than one of the above types of
responses.)
(Table 7)
By contributing reviews an individual can interact with other participants at the
repository, whether directly by one-one interactions or indirectly by responding to
reviews contributed by others. The responses in table 7 show that reviewers might
consider themselves as part of a group of like-minded individuals. A comment that is
indicative of this is shown in the following excerpt from a reviewer profile:
“This is so cool that Amazon permits us book lovers the space to share our thoughts
about what we're reading. I love to peruse other peoples' thoughts on the books I'm about
to buy and enjoy exchanging comments and ideas with other readers.”
25
What individuals seem to value most is the interactions with others who are not part of
their regular circle of acquaintances, as expressed by one of the reviewers whom we
interviewed: “Most pals, buddies know about my writing reviews. They do not look at my
reviews. Feedback from readers of my magazine reviews is usually from people whom I
know. What is noteworthy is the feedback from customers at Amazon: people who do not
know you. I get mail from people all over the world.” This suggests that a repository that
spans or is accessible by a large cross-section of individuals will be more attractive to
potential participants.
This incidence of the motivation to socialize with others is similar to another CMC
context. In their study of Usenet groups, Wasko and Faraj (2000) found that 18.7% of
individuals participate because they want to be part of a community of practice. In those
newsgroups, participants get to exchange multiple viewpoints, draw upon the experience
of peers, and also maintain and advance the community as a whole.
Among non-CMC contexts, Clary et al. (1998) found that individuals who were active in
volunteering gave a 10.21% weight to social motivations as compared to other
motivations for engaging in volunteering activity. For individuals who were not involved
in volunteering, the figure was 11.4% in determining an intention to engage in
volunteering. Diamond and Kashyap (1997) found that alumni gave a weight of 34.41%
to “individual attachment” (indicating connection and unity with other alumni) among
other motivations in determining the obligation they felt for supporting their alma mater.
Contributing reviews to a repository is different from interactions on a newsgroup or
other communication media, or the above face-to-face contexts. As noted earlier,
26
contributions are made without specific requests for help from particular individuals. Yet
we find 20.71% of individuals indicating that they contribute reviews to socialize with
other participants of the repository. These findings suggest that the act of contributing a
review to a repository is more than just an interaction with a database. The high incidence
of the social motivation in the contexts of both repository contributions and alumni
support for their alma mater is interesting. It seems to reinforce our above inference that
the activity of contributing reviews is a way for an individual to connect with others like
oneself. This is apparently less of a case in volunteering if the activity allows contact
mostly with the recipients of the volunteer’s help and less with other volunteers like him.
Intrinsic motive
We also found that some reviewers were writing reviews for the enjoyable experience
that the activity provided, as described in the following reasons given in their profiles:
Type of response / observation
Number of
Percentage of
occurrences
occurrences
For a creative release
3
0.30
For fun / relaxation
20
1.98
Total
23
2.28
(Note: The total figure above does not equal the sum of occurrences of individual
categories since some profiles contain more than one of the above types of responses.)
(Table 8)
For a task-oriented situation such as writing a review, these individuals feel a sense of
“flow” (Csikzentmihalyi 1975) or enjoyment from an activity that is equivalent to
playing. For example, some of the reviewers said in their profiles:
“… and reviewing books is remarkably cathartic”.
27
“I have taken great pleasure and profound joy in writing reviews about the works of
artists in the entertainment field.”
“Since I write reviews for my own enjoyment, I tend mostly to review books I enjoyed…”
“I am doing this for fun and imagine that, besides myself and I, no one else will ever read
this.”
One of the reviewers whom we interviewed said “Reviewing is fun. I do it for my own
enjoyment. If others enjoy it, that is fine.”
In prior studies of online forums, 6.5% of people indicated such enjoyment as a reason
for participating on Usenet groups (Wasko and Faraj 2000), and 9.5% for an email help
network (Constant et al. 1996).
There is thus a relatively infrequent occurrence of the enjoyment motivation among
contributors to repositories as compared to the other types of online forums. Selfdetermination theory (Deci and Ryan, 1985), which posits that an individual needs to feel
competent and in control, may suggest a reason for this observation. According to this
framework, the intrinsic enjoyment motivation arises when a person succeeds on a task
that requires some skill. This is the case with solving the problem posed by someone on
an email network or a bulletin board, or using a new computer technology (Venkatesh
2000). Writing a product review, on the other hand, does not address a specific problem
posed by others. The level of detail and the sophistication of the argument in a review is
entirely defined by the writer, thus making a non-task-oriented situation. As a result,
there would be a lesser scope for deriving joy from the activity.
28
Choice of content contributed:
We also found that people tend to choose areas to contribute in based on the contributions
made by other reviewers. We coded profiles into the categories in table 9 based on
statements saying that the reviewer chose to write about products that nobody (or a few
people) had reviewed, or had provided inadequate reviews in their opinion.
Type of response / observation
Number of
occurrences
46
Percentage of
occurrences
4.56
Bring awareness about the books,
CDs / genre of books, CDs that
customers may not otherwise
learn about
No one else has reviewed those
16
1.59
CDs / books etc.
Add information that other
11
1.09
reviewers have not covered
Provide a complete assessment
4
0.40
(as compared to professional
reviewers)
Total
64
6.34
(Note: The total figure above does not equal the sum of occurrences of individual
categories since some profiles contain more than one of the above types of responses.)
(Table 9)
In the above responses, contributors are stating that they tend to contribute reviews about
a particular book, CD, toy etc. (or in a particular area such as, e.g., art in the Middle East)
for which few (or no) other individuals have made contributions. Even if others have
contributed reviews, an individual tends to contribute his / her own review only if he / she
feels that the existing reviews in that area are inadequate in some way.
An example of this can be seen in the following excerpt from a profile:
“I try to review items that have no reviews, especially if there is no publisher description.
I want to encourage readers to try some of the more obscure titles that I enjoy. You may
29
note that most of my reviews have no feedback ... that is, at least in part, because few
people wander into these wonderful bycorners of literature/theology/folklore.
Do I read more conventional books? Yes, but if there are already reviews that
approximate what I think needs to be known, I don't bother to review them.”
One of the reviewers whom we interviewed also stayed away from contributing to areas
that others had contributed in, when he noted: “For new music CDs, I review everything.
But for catalog material (stuff that is on my shelves at home), I do not write reviews if
others have written reviews.”
There can be many individuals who can potentially contribute a review for a particular
product for the benefit of customers. Whether a particular individual steps forward to do
so can depend on the number of other individuals who have already taken upon
themselves to contribute a review. If an individual sees that others have already
contributed adequately in terms of covering all issues about the product in question, he
might feel less inclined to make his own contribution. The final word might not have
been written in all the existing reviews about the product. There might be even be many
gaps or flaws in those reviews. Yet a potential individual might not be keen about
contributing his own review simply because there are reviews contributed by others as
noted in the above responses in some profiles. This diffusion of responsibility is
consistent with prior theory that the greater the number of bystanders who witness an
emergency, the less likely is one of them to help the victim (Latane and Darley, 1970).
30
This effect might not be operative for individuals who contribute reviews for the benefit
of other reviewers (instead of potential buyers of the product). A reviewer whom we
interviewed said: “Reviewing is both like writing for a database and for other people.
People who look at reviews have already made up their minds. I think most votes that
people give reviews come from other reviewers. I myself vote for other reviewers’ reviews
before I write my own review. Most of the stuff I review has already been there for a
while (i.e., has not been recently published). I think half the voters are buyers and the
other half are reviewers.” So if an individual assumes that his reviews are being read and
appreciated by other reviewers, then his review can be part of an ongoing discussion, an
ongoing joint effort to extend the knowledge about a subject. Thus, the intended target for
one’s contribution can determine whether an individual is dissuaded or encouraged to
make a contribution in an area with many existing contributions.
Who contributes?
Tapping helpers versus tapping experts:
In the context of product reviews, a contributor can be said to be an expert if he / she has
educational qualifications or experience (as part of a professional career or in terms of
usage of current product and / or products similar to the one being reviewed) in the
subject that he / she is reviewing in. As expected, we did find instances of reviewers
writing about subjects in which they were experts.
For example one reviewer stated this in his profile:
“I also have a long time interest in history, and developed some expertise in medieval
history while assisting my brother in preparing the family genealogy.”
31
In the course of our detailed reading of the profiles of the top 1009 reviewers, we found
an individual with a PhD in statistics who reviews only books on statistics and a house
construction / repair contractor who writes only about tools and hardware. Many of the
reviewers also stated that they were writers of books/plays/songs, musicians, performing
artists, moviemakers, thereby indicating some level expertise in their respective areas in
which they wrote reviews. For example, among the reviewers we had editors (3.35%),
consultants (3.35%), and authors / composers (7.3%). One of the reviewers whom we
interviewed, who reviews music CDs, had been involved with evaluation of music
records in a 20-year career as a disc jockey and a 10-year experience as a music reviewer
in print magazines.
On the other hand, we also observed that many reviewers wrote on topics that were not
focused on one area of specialty. Amazon classifies the various products it sells into
broad categories such as books, music, videos, toys, and electronics. Each of these
categories has subcategories as shown in the table below:
Sub-categories in books
Arts & Photography
Biographies & Memoirs
Business & Investing
Children's Books
Computers & Internet
Cooking, Food & Wine
Entertainment
Gay & Lesbian
Health, Mind & Body
History
Home & Garden
Horror
Literature & Fiction
Mystery & Thrillers
Sub-categories in music
Alternative Rock
Broadway & Vocalists
Children's music
Classic Rock
Country
Dance & DJ
Folk
Hard Rock & Metal
International
Jazz
Latin Music
Miscellaneous
New Age
Pop
Sub-categories in video
Action & Adventure
Animation
Anime & Manga
Art House & International
Classics
Comedy
Cult Movies
Documentary
Drama
Horror
Kids & Family
Military & War
Music Video & Concerts
Musicals & Performing
Arts
32
Nonfiction
Outdoors & Nature
Parenting & Families
Professional & Technical
Reference
Religion & Spirituality
Romance
Science
Science Fiction & Fantasy
Sports
Teens
Travel
R&B
Rap & Hip-hop
Rock
Sub-categories in toys
Birth-12 months
12-24 Months
2 Years
3-4 Years
5-7 Years
8-11 Years
12-14 Years
Action Figures
Activities & Learning
Age Ranges
Arts & Crafts
Brands
Dolls
Electronics
Favorite Characters
Games
Grownups
Interests
Sports & Outdoor play
Stuffed animals
Vehicles & Die-cast
Mystery & Suspense
Science Fiction & Fantasy
Special Interests
Television
Westerns
Sub-categories in
electronics
Accessories & Supplies
Computer Add-ons
Handhelds & PDAs
Office Electronics
(Table 10)
To obtain data on the number of product categories in which individuals wrote reviews,
we sampled 50 reviewers ranked 1-10, 101-110, 251-260, 501-510, and 1000-1009. For
each individual, we went to his / her profile page, and noted the number of product subcategories for which he had written reviews using the counts provided by Amazon. We
tabulated these counts and calculated the number of sub-categories that each individual
had written reviews for. Table 11 presents a summary of our findings:
33
Statistic
All subcategori
es 3
Subcategories
in Books
Subcategories
in Music
Subcategories
in Video
Subcategories
in Toys
Maximum
Minimum
Average
Std
Deviation
Median
Mode
Count
52
1
13.72
10.21
21
0
7.72
6.18
15
0
2.4
3.98
17
0
2.4
3.90
15
0
1.1
2.79
Subcategories
in
Electronics
2
0
0.1
0.36
12
7
50
6.5
7
46
0
0
0
0
20
23
(Table 11)
0
0
10
0
0
4
Most individuals had written reviews for books, whereas only four out of the fifty had
written reviews for electronics items. Within each category, we see a large variation (in
terms of range and standard deviation) in the number of product sub-categories in which
reviews were written.
Expertise is usually built after continued working in one or a few related domains
(Ericsson, 1996). As a result, the fewer the number of categories in which a reviewer
writes reviews, the greater the expertise that he or she can be said to have in those areas.
The large variation in the number of categories reviewed by reviewers indicates that there
is a mix of both experts and non-experts. One of the reviewers whom we interviewed has
a PhD in rhetoric but writes reviews in 52 categories covering books, toys, music, video
and electronics.
In addition, a significant proportion of the reviewers cannot be deemed to be experts in
one or more areas according to their own admission. Many people indicated that they
3
The column “All categories” has data for all categories of products – books, music, video, toys and
electronics, considered as one group. Data in this column cannot be derived from data in the columns for
individual categories. For example, there is an individual who has contributed in a maximum of 52
different categories all considered together. But he is not the same as the individuals who contributed the
maximum in the individual books, music, video, toys and electronics categories.
34
were striving to attain expertise in areas such as performing arts, movie-making, and
music. For example, the table below shows the number of individuals who aspired to be
experts in writing skills and did not see themselves as experts currently.
Type of response / observation
Aspiring to be a published writer
Hope to be recognized as a top
reviewer / expert in a particular
field
Improve one’s writing
Total
Number of
occurrences
76
10
Percentage of
occurrences in sample
7.53
0.99
10
91
0.99
9.02
(Table 12)
The above reviewers are writing reviews in order to give advice as opposed to
demonstrating their deep knowledge about the subject. This is illustrated by the following
comments made by some reviewers in their respective profiles:
“I don't profess to be an expert at anything. In fact, I've always been something of a
professional diletante (sic). I write the reviews to help others make some kind of
judgement about whether they want to spend the money I did on a book.”
“I am not an expert in any of those subject areas, but I don't feel I'm boasting when I say
that my reviews of non-fiction books in these areas are more than just rantings”
“I don't claim to be an expert, but I am familiar with a great deal of music, and hope I
can identify a cd's strong points and identify the type of listener who will enjoy it.”
Thus the repository of reviews is being built by both experts and non-experts and not just
by experts alone.
35
Other findings:
We also coded reviewer profiles into the following category that we did not anticipate
from prior theory:
Contribution for self-development:
Some individuals indicated motives directed toward developing oneself to explain why
they contributed reviews as shown in the table below:
Type of response / observation
Number of
occurrences
6
10
9
Percentage of
occurrences
0.60
0.99
0.89
Sharpen thoughts by articulation
Improve one’s writing
Chronicle what I have seen, heard
or read
Developed expertise on a subject 16
1.59
and seeking to build on it
Total
39
3.86
(Note: The total figure above does not equal the sum of occurrences of individual
categories since some profiles contain more than one of the above types of responses.)
(Table 13)
The above responses indicate that the activity of contributing a review an individual can
improve one’s writing skills, organize and clarify the content of one’s thoughts about the
subject, and provide an avenue for enjoyment. Unlike the other-oriented motives
discussed earlier, the writing of reviews is itself a reward to the contributor. As one of the
reviewers said in his profile:
“I write reviews on Amazon.com's website for two reasons: first, to clarify and organize
my own thoughts and second, to assist others in selecting worthwhile reading material”.
One of the reviewers whom we interviewed elaborated on these sentiments: “I write
reviews to keep myself active and thinking (especially when something pops up related to
36
my teaching class)… My reviewing helps me with the critical thinking required for
school…Reviewing kept me active. It got me back into teaching. It wasn’t a problem
getting into teaching again. The mind was not rusty because of the reviewing I have been
doing. Taking things apart, being critical is important…” So there is more to enjoying the
activity of contributing reviews than just the playful fun involved.
Prior research has viewed self-oriented motives in terms of “intrinsic” motivation, in the
form of the joy one experiences when doing an activity. Wasko and Faraj (2000) found
that 6.5% of participants in the Usenet groups they studied participate because they enjoy
the experience. Similarly, in the email network that Constant et al. (1996) studied,
participants gave a weight of 9.5% to the joy one experiences when solving problems
faced by others, among reasons why they answer queries from others.
However, our findings indicate a more subtle reason for contributing reviews. Much like
the volunteers studied by Clary et al. (1998), individuals might contribute reviews to
“exercise knowledge, skills, and abilities that might otherwise go unpracticed”. In their
study, active volunteers indicated a weight of 19.35% to this “understanding function”
among various motivations for their volunteering activity, and non-volunteers gave a
weight of 19.82% among motivations that would determine their intention to engage in
volunteering. Each review can be an opportunity to develop one’s expertise in a particular
genre of products (corresponding to the sub-categories we discussed earlier), by
collecting one’s thoughts and organizing them in the form of written content. This ties in
with our earlier finding that contributors of reviews have varied levels of expertise.
Summary of findings:
37
We found evidence of well-established drivers that explain motivations for reviewer
contribution: altruism, reciprocity, social affiliation, utilitarian benefits and enjoyment.
We also found evidence of a situational factor that affects the content of reviewer
contribution: the bystander effect. But perhaps the most interesting findings were that
non-experts too contribute to the review repository and that self-improvement motives
can also drive individuals to participate in the repository.
The above findings can be depicted in the following theoretical framework (see figure 1
below):
Self-oriented factors:
•
•
Self-development
Enjoyment
Willingness to
contribute
Other-oriented factors
• Altruism
• Reciprocity
• Social affiliation
• Utilitarian benefits
Discussions and Conclusions
(Figure 1)
In trying to understand the reasons why people might contribute to document
repositories, we study a large, public, and well-known repository of product reviews. In
this first-time field exploration of the phenomenon, we use the profiles that the
contributors of the reviews have written about themselves as a source of data to glean
their motivations. Our findings are discussed below:
38
First, our above findings suggest that a variety of motivations can drive individuals to
contribute to a repository. This is consistent with the functionalist perspective in
psychology, which holds that people perform a certain activity because it serves one or
more functions (Snyder and Cantor, 1998):
a) Value-expressive: an activity may allow an individual to express his / her personal
values and to his / her concept of self. Contributing to a repository can be a way of
expressing one’s values about altruistic concern for others;
b) Utilitarian: a certain behavior may result in rewards from the person’s external
environment. So a person might contribute to a repository in order to receive monetary or
other rewards;
c) Social adjustive: doing a certain thing may lead an individual to better fit in with his /
her peer group. Contributing to a repository, for example, can offer a way for an
individual to be and socialize with other participants there, and reciprocate the benefits he
receives from them;
d) Knowledge: by engaging in particular task, an individual might have a new learning
experience, and be able to exercise one’s knowledge, skills and other abilities. Writing a
contribution for a knowledge repository can be just one such example of an activity that
allows a person to put down, organize and improve one’s thoughts about a specific topic.
Unlike the perspective of economic theory (in particular of public goods), we see that it is
not just reciprocity or utilitarian motives that drive contribution to repositories. The
results also clarify the nature of the intrinsic motivation that has long been studied in IS
39
literature. In the context of repositories, it springs out of the knowledge motive described
above.
It is likely that a particular feature of a knowledge repository system will be critical in
motivating particular individuals to contribute because that feature serves a function that
is salient for them. Having a range of such features increases the likelihood that at least
one feature will serve a motivational function of different individuals and therefore
encourage them to contribute.
Second, a potential contributor will be more likely to contribute in an area that has little
or no other contributions from others. This observation qualifies the finding of Thorn and
Connolly (1987) whose model of an information repository leads to reduced contribution
as the overall number of participants increases. Our study, instead, finds that contribution
decreases only with increase in number of contributors in a given domain. This also
throws some light on the observation by Butler (2001), who found that the number of
participants can be both positively and negatively related to the amount of
communication activity going on in an online community.
By providing a wide variety of topics in which individuals can contribute to the
repository, and highlighting those that have little or no contributions, a repository system
might be able to induce more contributions. The system might even leverage the
bystander effect by having a “call for contributions” for a period of time in which
potential contributors are not informed how many contributions have already been made
in a particular topic. So long as a potential contributor does not know if many other
individuals have contributed, he / she is more likely to contribute.
40
Third, the presence of a large number of non-experts among the reviewers on Amazon
suggests it does not take too much expertise to sustain a thriving repository. It is likely
that the non-experts are using the forum to organize and archive their own thoughts, and
thus improve their own knowledge. This is attested by the statements listed under the
“contribution for self-development” section above.
In addition, each individual has different requirements for knowledge on a particular
topic. Individuals vary in their absorptive capacity to assimilate knowledge of different
kinds on the same subject (Cohen and Levinthal, 1990; Szulanski, 1996). For example,
for a novel by Stephen King, an individual who is new to the genre would look for
different kinds of information as opposed to someone who is familiar with that line of
books. Such a variation in need for information calls for reviewers with varying levels of
expertise.
This observation among reviewers also goes against the finding from the model
developed by Thorn and Connolly (1987). Their model suggests that information
asymmetry leads to reduced contribution. Sender asymmetry exists when “some
participants have access to better information than others” (p. 518), i.e., when some are
more expert than others. User asymmetry exists when “some stand to benefit more from
information of a given quality than do others” (p. 518), i.e., when participants have
different needs for knowledge. Assuming that only the “hope for reciprocity” is a
consideration for participants, Thorn and Connolly argue that an individual will
contribute only when he expects to receive information of identical quality from others.
At Amazon, differences in expertise among reviewers would lead to sender symmetry.
41
Different shoppers are looking for different information in reviews leading to user
asymmetry. However, our study shows that since participants contribute because of
considerations apart from reciprocity, asymmetry in knowledge levels can actually induce
more contributions to the repository.
To encourage individuals with varying levels of expertise to contribute to the repository,
the repository system might be designed to help them organize their contributions in
terms of topics and keywords. It might also aid potential contributors by providing a
template for making contributions.
Limitations:
First, we conducted this exploratory study at only one site, the Amazon.com repository of
product reviews.
Secondly, we did not ask the reviewers specific questions about why they contributed.
Rather, we culled out their motivations from whosoever had mentioned their motivations
in their profiles. As such, we might have missed out some factors that might also be
relevant to the contribution activity.
Future research:
The findings in our study call for further theoretical development of the factors causing
individuals to contribute to repositories. They can then be tested on participants of
knowledge repositories in organizational settings. A similar theoretical model might also
be used to study knowledge sharing in other forms of online networks, such as bulletin
boards, and knowledge directories.
42
43
References
Aquilino, W. S. (1998) Effects of interview mode on measuring depression in younger
adults. Journal of Official Statistics, 14, 1, 15-29.
Bandura, A. 1977. Social learning theory. Englewood Cliffs, NJ: Prentice-Hall.
Barnett, M. A., Thompson, M. A., and Pfeiffer, J. R. (1985). Perceived competence to
help and the arousal of empathy. Journal of Social Psychology, 125, 5, 679-680.
Batson, C. D. (1998). Altruism and prosocial behavior. In D. T. Gilbert, S. Fiske, and G.
Lindzey (Eds.), The Handbook of Social Psychology (4th ed.), pp. 282-316. New York:
McGraw-Hill.
Batson, C. D., Duncan, B.D., Ackerman, P., Buckley, T., and Birch, K. (1981). Is
empathic emotion a source of altruistic motivation? Journal of Personality and Social
Psychology, 40, 290-302.
Bereiter, C. and Scardamalia, M. (1993). Surpassing ourselves: An enquiry into the
nature and implications of expertise. Open Court Publishing: Peru, IL.
Bergstrom, T., Blume, L., and Varian, H. (1986). On the private provision of public
goods. Journal of Public Economics, 29, 25-49.
Berkowitz, L. (1978). Decreased helpfulness with increased group size through lessening
the effects of the needy individual’s dependency. Journal of Personality, 46, 299-310.
Butler, B. S. (2001) Membership size, communication activity, and sustainability: A
resource-based model of online social structures. Information Systems Research, 12, 4,
346-362.
44
Butler, B., Sproull, L., Kiesler, S., and Kraut, R. (2002). Community effort in online
groups: Who does the work and why? In S. Weisband and L. Atwater (Eds.) Leadership
at a distance. Mahwah, NJ: Lawrence Erlbaum.
Charles, S.K. "Knowledge management lessons from the document trenches," Online,
(26:1), 2002, pp. 22-28.
Clary, E. G. and Orenstein, L. (1991). The amount and effectiveness of help: The
relationship of motives and abilities to helping behavior. Personality and Social
Psychology Bulletin, 17, 58-64.
Clary, E. G., Snyder, M., Ridge, R. D., Copeland, J., Stukas, A. A., Haugen, J., and
Miene, P. (1998). Understanding and assessing the motivations of volunteers: A
functional approach. Journal of Personality and Social Psychology, 74(6), 1516-1530.
Constant, D., Kiesler, S., and Sproull, L. (1994) “What’s Mine Is Ours, or Is It? A Study
of Attitudes About Information Sharing,” Information Systems Research, (5:4), 1, 400421.
Cohen, W. M., and Levinthal, D. A. (1990). Absorptive capacity: A new perspective on
learning and innovation. Administrative Science Quarterly, 35, 128-152.
Constant, D., Sproull, L., and Kiesler, S. (1996) The kindness of strangers: The
usefulness of electronic weak ties for technical advice,” Organization Science, 7, 19-135
Davenport, T. H. and Prusak, L. Working Knowledge: How Organizations Manage What
They Know, Harvard Business School Press, Boston, MA, 1998.
Deci, E. L. (1972) Intrinsic motivation, extrinsic reinforcement, and inequity. Journal of
Personality and Social Psychology, 18, 105-115.
Deci, E. L., and Ryan, R. M. (1985). Intrinsic motivation and self-determination in
human behavior. New York: Plenum.
Diamond, W. D., and Kashyap, R. K. (1997). Extending models of prosocial behavior to
explain university alumni contributions. Journal of Applied Social Psychology, 27(10),
915-928.
45
Ericsson, K. A. (1996) The acquisition of expert performance: An introduction to some of
the issues. In K. Anders Ericsson (Ed.) The road to excellence: The acquisition of expert
performance in the arts and sciences, sports and games, Lawrence Erlbaum: Mahwah, NJ.
Festinger, L. A. (1954). A theory of social comparison processes. Human Relations, 7,
117-140.
Harris, M. B. and Huang, L. C. (1973). Competence and helping. Journal of Social
Psychology, 89, 2, 203-210.
Lakhani, K. R., and Wolf, R. G. (2005) Why hackers do what they do: Understanding
motivation and effort in free/open source software projects. In J. Feller, B. Fitzgerald, S.
Hissam, and K. R. Lakhani (Eds.) Perspectives on Free and Open Source Software, MIT
Press (forthcoming).
Latane, B. and Darley, J. (1970)The Unresponsive Bystander: Why Doesn’t He Help?
Prentice Hall, Englewood Cliffs, NJ, 1970.
Nonaka, I. (1994). A dynamic theory of organizational knowledge creation. Organization
Science, 5, 1, 14-35.
Olson, M. (1967). The logic of collective action. Public goods and the theory of groups.
Cambridge, MA: Harvard University Press.
Piliavin, Irving M; Piliavin, Jane A; Rodin, Judith. (1975). Costs, diffusion, and the
stigmatized victim. Journal of Personality and Social Psychology, 32(3), 429-438.
Shamir, B. (1991) “Meaning, self and motivation in organizations,” Organization
Studies, 12, 1991, pp. 405-424.
46
Snyder, M. and Cantor, N. 1998. Understanding personality and social behavior: A
functionalist strategy. In D. T. Gilbert, S. Fiske, and G. Lindzey (Eds.), The Handbook of
Social Psychology (4th ed.), 635-679. New York: McGraw-Hill.
Sproull, L. and Kiesler, S. (1986). Reducing social context cues: Electronic mail in
organizational communication. Management Science, 32, 11, 1492-1512.
Strauss, A. L. and Corbin, J. M. Basics of Qualitative Research: Techniques and
Procedures for Developing Grounded Theory, 2nd ed., Sage Publications, Thousand
Oaks, CA, 1998.
Szulanski, G. (1996). Exploring internal stickiness: Impediments to the transfer of best
practice within the firm. Strategic Management Journal, 17, 27-44.
Takahashi, N. (2000). The emergence of generalized exchange. American Journal of
Sociology, 105, 4, 1105-1134.
Thorn, B. K. and Connolly, T. (1987). Discretionary data bases: A theory and some
experimental findings. Communications Research, 14(5, 512-528.)
Venkatesh, V. (2000) Determinants of perceived ease of use: Integrating control, intrinsic
motivation, and emotion into the Technology Acceptance Model. Information Systems
Research, 11, 4, 342-365.
Wasko, M. M. and Faraj, S. (2000) “’It is What One Does’: Why People Participate and
Help Others in Electronic Communities of Practice,” Journal of Strategic Information
Systems, 9, 2000, pp.155-173
Yamagishi, T., and Cook, K. S. (1993). Generalized exchange and social dilemmas.
Social Psychology Quarterly, 56(4), 235-248.
Yin, R. K. (1994) Case Study Research: Design and Methods, Sage Publications,
Thousand Oaks, CA.
47
Appendix 1:
List of questions forming a template for reviewers to respond to when writing their
personal profile:
•
•
•
•
•
•
Where do you live?
What you do for work? For fun?
What's your all-time favorite movie or book?
What are your passions? Your pet peeves?
If someone handed you an Amazon.com gift certificate, what would you buy?
If you had a month off--no work, no responsibilities--what would you do?
Appendix 2
Mr. D
Reviewer Rank: 23
A very good reason to write reviews for Amazon.com is to share ideas with others. As a
writer I find I have a much better understanding of what I have read when I write about it.
Additionally, when I write about a book I enter into a dialogue, however distant, with the
author. Most books I read are excellent, and even the flawed ones have much that is
valuable. Writing a book is a long and sometimes lonely process full of uncertainty and
doubt. When the book is finally published after several years of work there is a sense of
accomplishment immediately followed by a let down. Most books are published and
quickly go out of print. Their authors sometimes wonder, was it worth all the time and
energy, especially when they get superficial reviews from professionals who often do not
have time to actually read the books they review. 'It fell stillborn from the press' is an old
book business cliché and lament, and it is also sadly often the truth.
48
Appendix 3:
Interview Questions
Thank you for agreeing to this interview.
If a question is not clear to you, please let us know and we will clarify it for you. You are
free to decline an answer to any of these questions.
1. Introductory questions (5 minutes):
a. Can you tell us something about yourself:
i. your personal background, educational background etc.
ii. your interests
iii. your profession
b. How long have you been writing reviews?
c. Do you participate in the discussion boards at Amazon?
d. How much time, on average do you spend writing reviews?
2. Questions to understand details about writing reviews: (20 minutes)
a. To what extent do you feel posting a review is like interacting with a
database as opposed to being like an interaction with other people. When
you write your reviews, do you have a mental picture of the other people
at the Amazon website (customers, other reviewers)?
b. Have you received any feedback on your reviews – from others online?
From friends, colleagues and other persons offline?
c. What do you think you get in return for or as a result of writing the
reviews? What do you think are the benefits to others from the reviews?
e. Do you ever compare the opinions of other reviewers on the same book
you have reviewed? Do you do this before or after you write a review?
f. Do you like the competition that goes with the ranking system? Have you
experienced any censorship of reviews by Amazon?
g. Why write reviews for Amazon, and not for other websites such as, say,
Barnes and Noble, Books-A-Million, or other book review sites?
h. When writing your reviews, do you expect that the reader will know about
some of the concepts you use (e.g., genre of the type of novel being
reviewed etc.)?
3. Closing questions: (5 minutes)
a. Can you describe one or two things that you like very much about
Amazon (not necessarily connected with writing reviews)?
b. Can you describe one or two things that you dislike very much about
Amazon?
c. May we contact you again by email in connection with this project if we
have any questions?
Thank you so much for your time and sharing your thoughts on the reviews you write.
Have a great day!
49
Appendix 4:
Coding rules
Altruism
•
•
•
Explicit statement of wanting to help others without expecting any benefit for
oneself when doing the activity.
Expressing sentiments of doing one’s duty.
Indications of wanting to attain a goal larger than writing a review for others to
read.
Utilitarian motives:
•
•
Mention of actual or expected receipt of any tangible benefit (money, products
etc.) from others in exchange for contributing reviews.
Mention of actual or expected receipt of any non-tangible benefit (e.g., reputation)
from others in exchange for contributing reviews.
Reciprocity
•
•
Stating that one is writing reviews as some form of repayment to others for the
benefit one has gained by reading the reviews contributed by others.
Mention of having benefited by reading reviews written by others.
Social affiliation
•
Saying how the writing of reviews enabled the person to connect to the
community of reviewers or content producers.
Enjoyment motive
•
•
Mention of how the writing of reviews is a joyful or playful experience.
Statement saying that review writing provides some form of pastime or relaxation.
Choice of content contributed
•
•
Statement indicating why the individual chose to contribute reviews for certain
products, areas or categories.
Statement indicating why the individual chose NOT to contribute reviews for
certain products, areas or categories.
50
Expertise of contributor
•
•
Mention of commonly acknowledged indicators of expertise such as educational
qualifications, experience, certifications, etc., with respect to the domain in which
he contributes reviews.
Self-declaration by individual indicating a degree of expertise in any area
concerning the topic(s) on which he contributes reviews.
Self-development
•
Statement indicating that the writing of reviews organizes, develops or improves
one’s knowledge about a subject.
51