What's On Wikipedia and What's Not

Journal of Computer-Mediated Communication
Whatâ s on Wikipedia, and Whatâ s Notâ ¦?

Completeness of Information on the Online Collaborative Encyclopedia
Journal: Journal of Computer-Mediated Communication

Fo
Manuscript ID: JCMC-07-186
Manuscript Type: Full-length Research Article

r
Wikis, Social Network Analysis, Information Richness, Online

Keywords:
Communities
Pe
er
Re
vi
ew
International Communication Association

Page 1 of 33 Journal of Computer-Mediated Communication
1
2
3
4
5
6
7
8
9
10 What’s on Wikipedia and What’s Not…?
11
12 Completeness of Information on the Online Collaborative Encyclopedia
13
14
15 Abstract
16
17
18 The World Wide Web continues to grow closer to achieving the vision of becoming the
Fo
19
20 repository of all human knowledge. While improved search engines such as Google facilitate
21
22 access of knowledge across the Web, some sites have increased in popularity and have attracted
r
23
24
the attention of more Web users than others. Wikipedia is one such site that is becoming an
Pe
25
26
27 important resource for news and information. It is an online information source that is
28
er
29
increasingly used as the first, and sometimes only, stop for online encyclopedic information.
30
31
32 Much discussion has dealt with the accuracy of information on Wikipedia. While
Re
33
34 accuracy is important, that is not what this project is measuring. Using a method employed by
35
36
Tankard and Royal (2005) to judge completeness of Web content, completeness of information
vi
37
38
39 on Wikipedia is assessed. What we found was that some topics were covered more
ew
40
41 comprehensively than others and that predictors of these biases included recency, importance,
42
43
44 population, and financial wealth.
45
46
47
48 Keywords: Wikipedia, wiki, social network, completeness of information, open source, online
49
50
51 community
52
53
54
55
56
57
58
59
60
Journal of Computer-Mediated Communication Page 2 of 33
What’s on Wikipedia, and What’s Not…?

1
2
3
Introduction
4
5
6 The World Wide Web continues to grow closer to achieving the vision of becoming the
7
8 repository of all human knowledge (Heylighten, 1995). While improved search engines such as
9
10
11
Google facilitate access of knowledge across the Web, some sites have increased in popularity
12
13 and have attracted the attention of more Web users than others. Wikipedia is one such site. It is
14
15 an online information source that is increasingly used as the first, and sometimes only, stop for
16
17
18 online encyclopedic information.
Fo
19
20 Wikipedia (www.wikipedia.org), deemed “the free encyclopedia,” was launched on the
21
22 Web in 2001. (Wikipedia:About, 2007)) It was started by Jimmy Wales, formerly a futures
r
23
24
trader in Chicago, as an open information source, allowing anyone with access to the Internet to
Pe
25
26
27 post or edit content on the site. Wikipedia uses the wiki software format, which is a
28
er
29
collaborative development environment. Established as a non-profit organization, Wikipedia
30
31
32 currently receives over 38 million unique visitors per month and is ranked #13 on ComScore
Re
33
34 Media Metrix Top 50 Web Properties (Holiday Fever…, 2007). This open source project
35
36
operates under the assumption that more writers and editors are better than fewer, and that the
vi
37
38
39 community will develop and monitor content in a manner that is improved over that of
ew
40
41 traditional information publishing.
42
43
44 The open source concept has its roots in software development. One of its most notable
45
46 projects is the operating system Linux, which operates under the conditions of allowing and
47
48 encouraging multiple developers. Raymond (1997) compared this style of development using the
49
50
51 metaphor of bazaar and cathedral. “No quiet, reverent cathedral-building here - rather, the Linux
52
53 community seemed to resemble a great babbling bazaar of differing agendas and approaches
54
55
56
(aptly symbolized by the Linux archive sites, who'd take submissions from anyone) out of which
57
58
59
60
2

1
2
3
a coherent and stable system could seemingly emerge only by a succession of miracles.” Open
4
5
6 source is contrasted with propriety development environments in which only those with proper
7
8 license and authority can modify and implement source code. Benefits of this approach are the
9
10
11
inclusion of many and varied voices and agendas, the speed to which development can occur,
12
13 and policing of the environment by the community itself as opposed to regulatory or governing
14
15 bodies.
16
17
18 Wikipedia is now the Web's third most popular news and information source, with more
Fo
19
20 unique visitors than Yahoo News, MSNBC, AOL News, and CNN (Half of All U.S. Internet
21
22 Users…, 2006). Wikipedia's English-language version doubled in size last year and now has over
r
23
24
1 million articles. By this measure, it is almost 12 times larger than the print version of the
Pe
25
26
27 Encyclopaedia Britannica. It has over 100,000 contributors writing in 200 languages (The Wiki
28
er
29
Principle, 2006).
30
31
32 Wikipedia has become a popular site frequented by students, scholars, business people,
Re
33
34 family members, and government officials for finding information on a variety of topics. But,
35
36
due to the open nature of contributions, much attention has been given to the level of accuracy of
vi
37
38
39 information on Wikipedia. Many feel that Wikipedia’s policy of letting anyone create and edit
ew
40
41 content causes the information to be inaccurate, misleading, or generally incorrect, both
42
43
44 purposefully and accidentally. Instances have occurred in which rumors and falsities have been
45
46 planted on Wikipedia articles. For example, a Wikipedia entry was created that falsely
47
48 implicated John Siegenthaler, Sr. in the Robert Kennedy assassination (Giles, 2005; Udell, 2004;
49
50
51 Johnson, 2006). While the error was eventually corrected, it was not done so before being picked
52
53 up by other information resources and seen by untold numbers of users. Still, the philosophy of
54
55
56
the site is that with so many people looking at the content, in the long run, accuracy will prevail.
57
58
59
60
3

1
2
3
Biases in editing content have been revealed on Wikipedia. A system was developed by
4
5
6 a Cal Tech graduate student to trace the IP addresses of edits done by self-interested parties
7
8 (Borland, 2007). These edits often entailed removal of negative information and addition of
9
10
11
positive or public relations material, thus bringing into question the objectivity and democratic
12
13 potential of the Wikipedia model.
14
15 Wikipedia has sought to counter some of the criticisms by instituting measures designed
16
17
18 to reduce the number of attacks on the credibility of information on the site. Volunteer
Fo
19
20 administrators monitor content on the site, and can now block users from editing content on
21
22 specific articles. Some articles are temporarily protected from editing, until the climate for the
r
23
24
attack has died down. Others, like the article on George W. Bush, are semi-protected and open
Pe
25
26
27 to editing only by people who had been registered on the site for at least four days. (Hafner,
28
er
29
2006). But according to Wales, Wikipedia's founder, this type of protection affects a tiny
30
31
32 fraction of the 1.2 million entries on the English-language site. ''Protection is a tool for quality
Re
33
34 control, but it hardly defines Wikipedia,'' Mr. Wales said. ''What does define Wikipedia is the
35
36
volunteer community and the open participation.'' (Hafner, 2006)
vi
37
38
39 Some studies have actually refuted Wikipedia’s position as a reliable information source.
ew
40
41 In a recent study comparing the accuracy of science entries, Nature reported that Wikipedia’s
42
43
44 level of accuracy is close to that of Encyclopedia Britannica (Giles, 2005). The scientific journal
45
46 reported that, within 42 randomly selected general science articles, there were 162 mistakes in
47
48 Wikipedia versus 123 for Britannica, with the errors in Britannica being oriented towards
49
50
51 omissions rather than factual errors.
52
53 There is an indication that even librarians are finding value in the usage of Wikipedia
54
55
56
(Miller, et al., 2006). Attention to popular culture items and usage of links, objective presentation
57
58
59
60
4

1
2
3
of controversy, and up-to-date nature of information are all improvements over traditional
4
5
6 encyclopedias. Others characteristics of Wikipedia are considered strengths. Consistent
7
8 presentation of information and format of pages, organization of articles, and links to outside
9
10
11
sources provide users with a site that is robust and efficient. The basic search field allows users
12
13 to quickly find information on their desired topic or related subjects.
14
15 While accuracy of information is important, that is not what this project is measuring.
16
17
18 There are other criteria in which an information source can and should be judged. Using a
Fo
19
20 method employed by Tankard and Royal (What’s on the Web…, 2005) to judge completeness of
21
22 Web content, completeness of information on Wikipedia is assessed. With Wikipedia becoming
r
23
24
a popular online information destination, it might possess some of the biases inherent in the Web
Pe
25
26
27 at large. Are some topics covered more comprehensively than others and, if so, are there
28
er
29
systematic predictors, like those found on the Web in general, that determine the amount of
30
31
32 coverage?
Re
33
34 Tankard and Royal (What’s on the Web…, 2005) studied completeness of information on
35
36
the Web by creating systematic lists of topics to search in two popular search engines. Searching
vi
37
38
39 on several dimensions including recency, importance of information, country population, and
ew
40
41 company revenue, they also investigated these dimensions as predictors of completeness of
42
43
44 information on the Web. In each dimension, Tankard and Royal found that there were systematic
45
46 biases. Coverage for items that were current, deemed important, or held power measured by
47
48 country size or company wealth produced more hits, and thus were covered more frequently on
49
50
51 the Web. This study was unique in that it was the first to propose a method for assessing the
52
53 completeness of information on the Web (Finding Out What’s on the World Wide Web, 2005).
54
55
56
It seems appropriate to modify these methods and to apply them to specific Web resources,
57
58
59
60
5

1
2
3
particularly those that have become population information destinations. Wikipedia is a likely
4
5
6 candidate for analysis in that its goal is to provide information created and accessible by all with
7
8 an Internet connection, much like the Web itself.
9
10
11
12
13
14
15 Review of Literature
16
17
18 While there have been many articles questioning Wikipedia’s accuracy, few
Fo
19
20 communication studies have focused on Wikipedia. Lih (2004) studied news articles citing
21
22 Wikipedia and analyzed the trends in using Wikipedia as a source.
r
23
24
Denning, et al. (2005) listed several risks inherent in the Wikipedia model: accuracy,
Pe
25
26
27 motives, uncertain expertise, volatility, coverage, and sources. Of coverage, the authors said,
28
er
29
Voluntary contributions largely represent the interests and knowledge of a self-selected
30
31 set of contributors. They are not part of a careful plan to organize human knowledge.
32 Topics that interest the young and Internet-savvy are well covered, while events that
Re
33 happened “before the Web” may be covered inadequately or inaccurately, if at all. More
34 is written about current news than about historical knowledge.
35
36
vi
37
38 Other studies have looked at Wikipedia’s strength as a reference source. Bill Katz
39
ew
40 developed six fundamental evaluation criteria for reference work: purpose, authority, scope,
41
42
43
audience, cost, and format (Wallace and Van Fleet, 2005). Wikipedia did not perform well on
44
45 the brief analysis performed by Wallace and Van Fleet on these criteria. Value, however, was
46
47 identified in the democratic and timely circumstances under which articles are created and
48
49
50 revised. According to Bopp and Smith (2001), coverage in an encyclopedia reference source
51
52 “should be even across all subjects,” although “it is important to note that some subjects, by their
53
54 very nature, demand greater emphasis.” While Wikipedia boasts over 1 million articles, Wallace
55
56
57 and Van Fleet expressed that volume of articles alone is not a useful indicator of scope.
58
59
60
6

1
2
3
Like the Tankard and Royal study, this project challenges the notion that the Web may be
4
5
6 the repository of all human knowledge by assessing the coverage on one of its most popular
7
8 information destinations, Wikipedia. By making systematic measurements of the amount of
9
10
11
information on Wikipedia using the same dimensions, we attempt to identify factors that predict
12
13 Wikipedia’s completeness.
14
15 Borrowing methods from the Tankard and Royal study, this project measures the content
16
17
18 of Wikipedia against various indexes or standards of completeness to identify and uncover
Fo
19
20 potential inherent biases. Communication research provided direction in identifying predictor
21
22 variables. Journalism scholars have often included completeness as one of the basic concepts of
r
23
24
journalism. McQuail stated that completeness “is usually thought to be a precondition of proper
Pe
25
26
27 understanding of news, and the media generally promise completeness in the sense of a full
28
er
29
range of information about significant events of the day” (McQuail, 1992, p. 211).
30
31
32 In an early study of the completeness of newspaper coverage, Danielson and Adams
Re
33
34 (1961) examined coverage of the 1960 presidential election campaign. They developed a list of
35
36
1,033 campaign events and then drew a random sample of 42 events to be used as a checklist
vi
37
38
39 against which articles were judged.
ew
40
41 Tankard and Showalter (1977), in their study of coverage of the 1972 Surgeon General’s
42
43
44 report on television violence, constructed an index of completeness by checking for presence or
45
46 absence of “three elements that were judged necessary for full reader understanding.” While the
47
48 present study does not focus on individual news stories, it borrows the technique of using a list of
49
50
51 facts or concepts as an effective means of measuring completeness.
52
53 Research on news flow has identified a number of factors that influence the presence or
54
55
56
absence of information. A related research approach—theoretical influences on mass media
57
58
59
60
7

1
2
3
content—has identified five major categories of influence on news content: the individual
4
5
6 journalist, media routines, the journalistic organization, extramedia sources, and ideology
7
8 (Shoemaker & Reese, 1995).
9
10
11
Current information is the bedrock of journalistic reporting (Berkowitz, 1990; McMillin,
12
13 1996; Curtin & Rhodenbaugh, 1999). With regard to the Web, currency comes into play in
14
15 another sense. Shoemaker & Reese (1995) identified the individual as a news influencer. Web
16
17
18 users and content creators tend to be young, with strong ties to current popular culture. The
Fo
19
20 contributors to Wikipedia are likely to mirror the demographics of the Web at large. This factor
21
22 would tend to weight the content of the Web, and ostensibly Wikipedia, toward material that
r
23
24
these individuals would be interested in—material of greater currency or recency.
Pe
25
26
27 Galtung & Ruge (1965), identified signal strength, or amplitude as another significant
28
er
29
factor influencing the flow of news. This factor might also be thought of as the importance of
30
31
32 information. When considering the probability of information being on Wikipedia, importance of
Re
33
34 the information is likely to be a useful predictor, with the more important items having the most
35
36
attention paid to them.
vi
37
38
39 Kariel and Rosenvall (1984) identified country population as an important predictor of
ew
40
41 international news flow. Countries with larger populations have more individuals to become the
42
43
44 focus of news coverage, hold greater political influence, and have more people who could
45
46 potentially create and contribute to online content.
47
48 Shoemaker and Reese (1995, p. 190) suggested that capitalist-owned media content tends
49
50
51 to favor those with economic power. In addition, corporations that are larger have more market
52
53 impact, have larger budgets for advertising and public relations, and have influence on more
54
55
56
people.
57
58
59
60
8

1
2
3
4
5
6 Research Questions
7
8 Borrowing from the Tankard and Royal study of completeness of information on the
9
10
11
Web, the following research questions were developed as they related to Wikipedia:
12
13 1. Are there some systematic gaps or biases in the overall presentation of information
14 made available on Wikipedia?
15
16 2. Is recency (or currency) a predictor of amount of information on Wikipedia?
17
18
3. Is importance of information a predictor of amount of information on Wikipedia?
Fo
19 4. Is population a predictor of amount of information about particular countries on
20 Wikipedia?
21
22 5. Is economic power a predictor of amount of information about individual corporations
r
23 on Wikipedia?
24
Pe
25
26
27
28
er
29
30 Method
31
32 Using the same predictors as Tankard and Royal, recency, importance, country
Re
33
34 population, and economic power, several systematic searches on Wikipedia were conducted.
35
36
Lists were developed within each of the dimensions, the contents of which are described in the
vi
37
38
39 results section. Each term on the lists was searched using the Wikipedia search feature. A
ew
40
41
42
determination was made as to the main page of content for that term. In some cases, such as the
43
44 countries of the United Nations, the list of countries on the United Nations page was used to find
45
46 the main article on a particular country. Each page was visited and the relevant content was
47
48
49 highlighted. Wikipedia navigation and other superfluous links that were not related to the actual
50
51 term being searched were not included in the selection. To capture the word count of items
52
53 selected on a page, an extension of the Firefox Web browser, Word Count, was downloaded
54
55
56 (http://roachfiend.com/archives/2005/03/03/word-count/). This extension counted the number of
57
58
59
60
9

1
2
3
words in the selection by simply using the Ctrl key on the computer’s keyboard. Word counts
4
5
6 were captured in a spreadsheet for each dimension. Items were plotted on charts, first in
7
8 ascending order, then by predictor variable. Items within dimensions were then compared and
9
10
11
correlated with predictor variables. When possible, the same search terms that were used in the
12
13 Tankard & Royal study were employed here.
14
15 All statistical analyses were conducted with Spearman (rank order) correlation
16
17
18 coefficients because parametric statistics (such as the Pearson correlation coefficient) are
Fo
19
20 inappropriate for L-shaped distributions (Bradley, 1982), which occurred with most of our data.
21
22 The correlations represent relative, as opposed to absolute, relationships.
r
23
24
Pe
25
26
27 Results
28
er
29
Several variables were used to test the currency dimension. First, using the same method
30
31
32 as Tankard and Royal, years were assessed. Wikipedia conveniently provided an article
Re
33
34 depicting the highlights of each year. Figure 1a depicts the word count of each article in
35
36
ascending order, disregarding year. A backward L-shaped curve is evident. Figure 1b depicts
vi
37
38
39 the word count by year in chronological order, starting with 1900 and going through 2010.
ew
40
41 There is a clear progression of the length of each article with a dramatic increase occurring
42
43
44 starting in 2001. Years in the future, understandably, were shorter, given that there was not yet
45
46 much to write about them. The average word count for the years since 2001 was 90% greater
47
48 than the average for the entire preceding 100 years (4566 vs. 8692).
49
50
51 The chart in Figure 8 depicts correlations of dimensions variable with predictor variables.
52
53 The Spearman correlation for Years was .79, indicating a very strong relationship of article word
54
55
56
count to the recency of information.
57
58
59
60
10

1
2
3
Figure 1a
4
5 Years - Ascending Order
6
7
8 12,000
9
10
11
10,000
12
13
14
15 8,000
16
17
18 6,000
Fo
19
20
21 4,000
22
r
23
24 2,000
Pe
25
26
27 0
28
1 9 17 25 33 41 49 57 65 73 81 89 97 105
er
29
30
31 Figure 1b
32 Years – Chronological Order
Re
33
34
12,000
35
36
vi
37
10,000
38
39
ew
40
8,000
41
42
43
6,000
44
45
46
4,000
47
48
49
2,000
50
51
52
0
53
54
00
06
12
18
24
30
36
42
48
54
60
66
72
78
84
90
96
02
08
19
19
19
19
19
19
19
19
19
19
19
19
19
19
19
19
19
20
20
55
56
57
58
59
60
11

1
2
3
Figure 2a shows the word count for articles on Wikipedia for the Academy Award
4
5
6 winning films in ascending order (Appendix A lists the films by year). This list was not searched
7
8 in the Tankard & Royal study, as it was difficult for them to identify only Web sites associated
9
10
11
with films with common names, such as Wings or Rebecca. However, this was made easier on
12
13 Wikipedia, with each film having a specific article associated with it. Another backward L-
14
15 shaped distribution is displayed. With few exceptions, such as Gone with the Wind (1939) and
16
17
18 Casablanca (1943) the analysis in Figure 2b plotted by year (1928-2005) shows a progression
Fo
19
20 favoring more current films. This demonstrates that while recency is an important predictor,
21
22 some films transcend time and are deemed important for other reasons, and thus have a strong
r
23
24
share of coverage on Wikipedia. The average word count for the films since 2001 was 80%
Pe
25
26
27 higher than the average word count for the time prior to 2000 (3190 vs. 1771). These last five
28
er
29
years accounted for 11% of the total word count for the 78 years of the award. The Spearman
30
31
32 correlation for films over years was .49 (see Figure 8), but that increased to .62 simply by
Re
33
34 removing the two outliers mentioned above. This indicates a strong relationship between word
35
36
count and time for films.
vi
37
38
39
ew
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
12

1
2
3
Figure 2a
4
5 Films - Ascending Order
6
7 9,000
8
9 8,000
10
11
7,000
12
13
14 6,000
15
16 5,000
17
18 4,000
Fo
19
20 3,000
21
22
2,000
r
23
24
1,000
Pe
25
26
27 0
28
1 5 9 13 17 21 25 29 33 37 41 45 49 53 57 61 65 69 73 77
er
29
30
31
32
Re
33 Figure 2b
34 Films – By Year
35
36
9,000
vi
37
38 8,000
39
ew
40 7,000
41
42 6,000
43 5,000
44
45 4,000
46
3,000
47
48 2,000
49
50 1,000
51 0
52
1928
1932
1936
1940
1944
1948
1952
1956
1960
1964
1968
1972
1976
1980
1984
1988
1992
1996
2000
2004
53
54
55
56
57
58
59
60
13

1
2
3
4
5 Figures 3a and 3b show another test of recency not performed in the Tankard & Royal
6
7 study by looking at Time Magazine’s Person of the Year (Appendix B lists the people by year).
8
9 Some years that did not include an individual were discarded (for example, in 2002, “The
10
11
12 Whistleblowers”). Figure 3a shows a backward L-shaped distribution when disregarding time,
13
14 although not as steep as some of the others experienced in this analysis. The progression appears
15
16
evenly distributed, only slightly skewed to the upper half of the distribution (the median was
17
18
93% of the average). But, Figure 3b shows a more random pattern than those experienced with
Fo
19
20
21 Year and Film. The Spearman correlation (see Figure 8) for recency was close to 0, thus
22
r
23
24
indicating no relationship with time. This indicates that while a bias is evidenced in the
Pe
25
26 consistently upward progression of Figure 3a, the bias is not due to recency in regard to Person
27
28 of the Year, but perhaps to some other measure of importance.
er
29
30
31
32
Re
33
34
35
36
vi
37
38
39
ew
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
14

1
2
3
Figure 3a
4
5 Time’s Person of the Year - Ascending Order
6
7 18,000
8
9 16,000
10
11
14,000
12
13
14 12,000
15
16 10,000
17
18 8,000
Fo
19
20 6,000
21
22
4,000
r
23
24
2,000
Pe
25
26
27 0
28
1 5 9 13 17 21 25 29 33 37 41 45 49 53 57 61
er
29
30
31
32 Figure 3b
Re
33 Time’s Person of the Year - By Year

34
35
36 18,000
vi
37 16,000
38
39 14,000
ew
40
41 12,000
42 10,000
43
44 8,000
45
6,000
46
47 4,000
48
49 2,000
50
0
51
52
27
31
35
39
43
47
52
57
62
67
74
79
85
91
96
01
53
19
19
19
19
19
19
19
19
19
19
19
19
19
19
19
20
54
55
56
57
58
59
60
15

1
2
3
Another search that was added that was not performed in the Tankard & Royal study was
4
5
6 to consider musical artists over time. An artist holding the #1 song on the Billboard Top 100 for
7
8 the first week in February of each year since 1940 was selected (Appendix C shows artists by
9
10
11
year). Figure 4a depicts the word count of the main Wikipedia article associated with that artist
12
13 in ascending order for each of the selected artists, again depicting the backward L-shaped
14
15 distribution. Figure 4b shows each artist by year. While the pattern in the graph appears to
16
17
18 indicate a random distribution, the Spearman correlation with time was .30 (See Figure 8). By
Fo
19
20 eliminating just two outliers (Bing Crosby – 1945 and the Beatles – 1964), the correlation
21
22 increases to .40. The average word count for the artists since 1990 was 32% higher than for the
r
23
24
years from 1940-1989 (3332 vs. 2511). Similar to the trends found in film, it shows that while
Pe
25
26
27 the recency relationship is strong, some artists transcend time and receive more coverage on
28
er
29
Wikipedia than would be indicated by their currency.
30
31
32
Re
33
34
35
36
vi
37
38
39
ew
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
16

1
2
3
Figure 4a
4
5 Artists with #1 Songs on Billboard – Ascending Order
6
7 14,000
8
9
10 12,000
11
12
13 10,000
14
15
8,000
16
17
18 6,000
Fo
19
20
21 4,000
22
r
23
24 2,000
Pe
25
26
27 0
28 1 5 9 13 17 21 25 29 33 37 41 45 49 53 57 61 65
er
29
30
31
32
Figure 4b
Re
33
34 Artists with #1 Songs on Billboard – By Year
35
36 14,000
vi
37
38
12,000
39
ew
40
41 10,000
42
43 8,000
44
45
6,000
46
47
48 4,000
49
50 2,000
51
52 0
53
1940
1943
1946
1949
1952
1955
1958
1961
1964
1967
1970
1973
1976
1979
1982
1985
1988
1991
1994
1997
2000
2003
2006
54
55
56
57
58
59
60
17

1
2
3
To measure comprehensiveness of information, we used the same random sample
4
5
6 employed in the Tankard & Royal study of 100 topics from the Micropaedia of the
7
8 Encyclopaedia Britannica (See Appendix D). Figure 6 shows the word count of each term’s main
9
10
11
page on Wikipedia. Once again, a backward L-shaped distribution emerged. Of the 100 items,
12
13 14 did not have a Wikipedia entry (included phrases such as “Russian Association of Proletariat
14
15 Writers,” “League for the Independence of Vietnam,” and “urethane”). Fifteen of the terms had
16
17
18 articles with a word count of 2000 or more. The average word count for those 15 terms was 5
Fo
19
20 times that of the average word count for the other items on the list with Wikipedia articles.
21
22 A Spearman correlation was used to compare inches of content in the Micropaedia of the
r
23
24
Encyclopaedia Britannica with word count on Wikipedia. This correlation was calculated at .26,
Pe
25
26
27 indicating some relationship with the importance placed on information in the traditional
28
er
29
encyclopedia with that in Wikipedia (See Figure 8). In some cases, the articles on Wikipedia
30
31
32 indicated that the content had been derived from a print encyclopedia source. There was no time
Re
33
34 dimension or other predictor variable with which to compare for encyclopedia terms.
35
36
vi
37
38
39
ew
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
18

1
2
3
Figure 5
4
5 Encyclopedia Terms – Ascending Order
6
7
8 14,000
9
10 12,000
11
12
13 10,000
14
15 8,000
16
17 6,000
18
Fo
19 4,000
20
21
22 2,000
r
23
24 0
Pe
25 1 8 15 22 29 36 43 50 57 64 71 78 85 92 99
26
27
28
er
29
30
31
32
Re
33
34
35
36
vi
37
38
39
ew
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
19

1
2
3
Figure 6a shows the word count for the main Wikipedia article by country in the United
4
5
6 Nations in ascending order. Articles were analyzed for all 192 countries of the United Nations.
7
8 Once again, a backward L-shaped distribution emerged. The distribution is fairly even, with a
9
10
11
sharp increase experienced for the top 22 countries. Figure 6b shows a gradual upward
12
13 distribution when charted in order by population (higher number indicates higher population).
14
15 Spearman correlation for countries with population was .55, indicating that the larger countries
16
17
18 were more represented on Wikipedia in terms of word count per article (see Figure 8). The top
Fo
19
20 10% of countries by population accounted for 15% of the total word count for country articles
21
22 and the average word count for the top 10% of countries was 63% higher than those on the rest
r
23
24
of the list.
Pe
25
26
27
28
er
29
30
31
32
Re
33
34
35
36
vi
37
38
39
ew
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
20

1
2
3
Figure 6a
4
5 Countries in UN – Ascending Order
6
7 14,000
8
9
10 12,000
11
12
13 10,000
14
15
8,000
16
17
18 6,000
Fo
19
20
21 4,000
22
r
23
24 2,000
Pe
25
26
27 0
28 1 13 25 37 49 61 73 85 97 109 121 133 145 157 169 181 193
er
29
30
31
32 Figure 6b
Countries of the UN - Ordered By Population
Re
33
34
35
36 14,000
vi
37
38 12,000
39
ew
40
41 10,000
42
43 8,000
44
45
46 6,000
47
48 4,000
49
50
51 2,000
52
53 0
54
1 13 25 37 49 61 73 85 97 109 121 133 145 157 169
55
56
57
58
59
60
21

1
2
3
Figure 7a shows the word count for a random selection of 86 Fortune 1000 companies in
4
5
6 ascending order (Appendix E lists the selection of companies). This chart shows the backward
7
8 L-shaped distribution with a sharp increase for 10% of the companies. Another 10% of the
9
10
11
companies did not have Wikipedia entries. Figure 7b shows the companies ranked by revenue
12
13 (higher number indicates higher revenue). The chart shows a distribution trending toward
14
15 increased word count for companies with the highest revenue. The Spearman correlation for
16
17
18 word count of these articles with company revenue was .49. The top 10% of the companies by
Fo
19
20 revenue accounted for 30% of the total word count for articles about companies.
21
22
r
23
24
Pe
25
26
27
28
er
29
30
31
32
Re
33
34
35
36
vi
37
38
39
ew
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
22

1
2
3
Figure 7a
4
5 Fortune 1000 Companies – Ascending Order
6
7 6,000
8
9
10 5,000
11
12
13
14 4,000
15
16
17 3,000
18
Fo
19
20 2,000
21
22
r
23
1,000
24
Pe
25
26
27 0
28 1 5 9 13 17 21 25 29 33 37 41 45 49 53 57 61 65 69 73 77 81 85
er
29
30
31
32 Figure 7b
Fortune 1000 Companies - By Revenue
Re
33
34
35 6,000
36
vi
37
38 5,000
39
ew
40
41 4,000
42
43
44 3,000
45
46
47 2,000
48
49
50 1,000
51
52
53 0
1
5
9
13
17
21
25
29
33
37
41
45
49
53
57
61
65
69
73
77
81
85
54
55
56
57
58
59
60
23

1
2
3
4
5 Figure 8
6 Spearman Correlation (Rank Order) with Predictor Variables
7
8 Dimension Variable Predictor Variable Correlation
9
Year Time .79
10
11 Academy Award Winning Films Time .47
12 Time’s Person of the Year Time .00
13 Artist w/ #1 Song Time .30
14 Encyclopedia Term Column Inches of Encyclopedia .26
15
16 Country 2005 Projected Population .55
17 Company 2005 Revenue .49
18
Fo
19
20 Discussion
21
22
In each of the searches performed for the dimensions, a bias was evident. And, when
r
23
24
Pe
25 considering the predictor variables, except when measuring Time’s Person of the Year, a strong
26
27
28
correlation was experienced. Within the currency or recency dimension, by looking at Year,
er
29
30 Films, and Musical Artist, the more current topics were the most covered. While the Person of
31
32 the Year category showed some bias, it was not closely correlated with time. This is perhaps due
Re
33
34
35 to the nature of the decision as being more editorial and reflecting opinions of the staff of Time.
36
While they often select people who are historically important, over time, according to our study,
vi
37
38
39 they choose people with staying power or continued importance only about half the time.
ew
40
41
42 When looking at a random selection of encyclopedia terms, bias was also inherent. Most
43
44 of the items that we searched had some information about them on Wikipedia, indicating broad
45
46 coverage. But, it was clear that the more common or popular terms had the most detailed
47
48
49 coverage. Coverage in Wikipedia was loosely correlated with the inches dedicated to the topics
50
51 in one traditional encyclopedia, indicating the strength of the agenda established by these
52
53
54
publications.
55
56
57
58
59
60
24

1
2
3
In terms of country population, biases toward larger countries were found and were
4
5
6 positively correlated with country size. This indicates that the democratic nature of Wikipedia
7
8 on its own cannot counteract the effects of the magnitude of people that are available to
9
10
11
participate.
12
13 And, in regard to Fortune 1000 companies, those with larger revenue streams and
14
15 resources are more likely to have greater coverage on Wikipedia. This points to the strength of
16
17
18 financial power in circumventing any type of democratizing feature of an online space.
Fo
19
20
21
22 Conclusion
r
23
24
In some ways, this was a more straightforward study than the one performed by Tankard
Pe
25
26
27 & Royal. In their study, they had difficulty in determining whether certain searches were
28
er
29
capturing all the information on a topic while not including irrelevant information. For example,
30
31
32 search for years in a search engine can provide references to the numbers rather than the years.
Re
33
34 They attempted to alleviate this problem by searching for the word “year” before the numerical
35
36
year and putting quotation marks around that text string. This did not capture hits regarding
vi
37
38
39 years that were not preceded by the word “year”. Some searches were difficult to perform if the
ew
40
41 topics were not presented consistently, as in the encyclopedia terms.
42
43
44 Shariatmadari (2006) identified characteristics of Wikipedia that make this case as well.
45
46 Wikipedia is specifically intended as a work of reference while using a search engine is not. A
47
48 search engine’s purpose is to identify various sites as opposed to finding immediate context.
49
50
51 Shariatmadari also indicated some coverage issues with Wikipedia, finding content more on
52
53 popular culture and science fiction than history.
54
55
56
57
58
59
60
25

1
2
3
In general, the searches on Wikipedia revealed individual articles on each topic, making
4
5
6 it easier to identify the relevancy of it to the search item. And, Wikipedia conveniently provided
7
8 an article stub for each year. A stub is an empty article that is ready to receive content. While this
9
10
11
approach does not measure all the content on Wikipedia related to a particular year, it does
12
13 provide one indicator of the amount of coverage and attention given to a year. Additional
14
15 searches that were not done in the Tankard & Royal study were performed on the recency or
16
17
18 currency dimension to help improve this area, including Time Magazine Person of the Year,
Fo
19
20 Academy Award Winning Films by Year, and Artists having #1 Songs by Year.
21
22 Length of the individual article was all that was included in the Word Count for each
r
23
24
topic. One feature of the Web that is also a feature of Wikipedia is the usage of links. Most
Pe
25
26
27 articles included links to other articles that enhanced or augmented the content of a particular
28
er
29
stub. Often these links are tangents, describing other people or events mentioned in the article.
30
31
32 Trying to capture the word count of associated links would have made for an unwieldy study.
Re
33
34 Information on Wikipedia is extremely volatile and dynamic. Articles can change
35
36
dramatically over time. This study was performed during November 2006 and each search
vi
37
38
39 within a variable was performed on the same day during the same time period, to improve the
ew
40
41 comparison of that information. This project merely captures the presence of information in the
42
43
44 timeframe under analysis. Some of the biases uncovered may subside or change over time. So,
45
46 while this study uncovered important biases in information being presented on Wikipedia, it will
47
48 be important to continue research in the area of measuring both accuracy as well as completeness
49
50
51 of information on online sites that are becoming important information resources, particularly
52
53 those taking advantage of the democratic and open source features of the technology.
54
55
56
57
58
59
60
26

1
2
3
Bibliography
4
5
6 Berkowitz, D. (1990). Refining the gatekeeping metaphor for local television news. Journal of
7 Broadcasting & Electronic Media, 34, p. 55-69.
8
9 Bopp, R.E. & Smith, L.C. (2001), Reference and Information Services: An Introduction, 3rd ed.,
10 Englewood, Colorado, Libraries Unlimited, p. 436.
11
12
13 Borland, John (August 14, 2007), See Who's Editing Wikipedia - Diebold, the CIA, a Campaign,
14 Wired (http://www.wired.com/politics/onlinerights/news/2007/08/wiki_tracker).
15
16 Bradley, J. V. (1982). The insidious L-shaped distribution. Bulletin of the Psychonomic Society,
17 20, p. 85-88.
18
Fo
19 Curtin, P. A., & Rhodenbaugh, E. (1999). It's not easy being green: Building the news media
20
21 agenda on the environment. Association for Education in Journalism and Mass
22 Communication. New Orleans, LA.
r
23
24 Danielson, W. A., & Adams, J. B. (1961). Completeness of coverage of the 1960 campaign.
Pe
25 Journalism Quarterly, 38, p. 441-452.

26
27 Denning, P., Horning, J., Parnas, D., & Weinstein, L. (2005, December). Wikipedia Risks.
28
Communications of the ACM.
er
29
30
31 Giles, J. (2005, December). Internet Encyclopaedias Go Head to Head. Nature.
32
Re
33 Hafner, K. (2006, June 17). Growing Wikipedia Revises Its 'Anyone Can Edit' Policy.” New
34 York Times.
35
36 Half of All U.S. Internet Users Visited News Sites in June 2006 (2006, August 7), ComScore
vi
37
38
Media Metrix Press Release, http://www.comscore.com/press/release.asp?press=971.
39
Heylighten, F. (1995). From World-Wide Web to Super-Brain. Principia Cybernetica Web.
ew
40
41 Retrieved April 22, 2002, from http://pespmc1.vub.ac.be/SUPBRAIN.html
42
43 Holiday Fever Drives Traffic to Shopping Sites in December (2007, January 16,, ComScore
44 Media Metrix Press Release, http://www.comscore.com/press/release.asp?id=1177.
45
46
47
Johnson, G. (2006, January 3). The Nitpicking of the Masses vs. the Authority of the Experts.
48 New York Times.
49
50 Kariel, H. G., & Rosenvall, L. A. (1984). Factors influencing international news flow.
51 Journalism Quarterly, 61, 509-516.
52
53 Lih, A. (2004, April 16-17). Wikipedia as Participatory Journalism: Reliable Sources? Metrics
54
55
for evaluating collaborative media as a news resource. 5th International Symposium on
56 Online Journalism. University of Texas at Austin.
57
58
59
60
27

1
2
3
McMillin, D .C. (1996). Roles journalists play: An examination of journalists' roles as
4
5 manifested in samples of their best work. Association for Education in Journalism and
6 Mass Communication. Anaheim, CA.
7
8 McQuail, D. (1992). Media performance: Mass communication and the public interest. London:
9 Sage.
10
11 Miller, B.X., Helicher, K., & Berry, T. (2006, April 1). I Want My Wikipedia. Library Journal.
12
13
14 Raymond, Eric (1997), The Cathedral and the Bazaar, First Monday,
15 http://www.firstmonday.org/issues/issue3_3/raymond/.
16
17 Shariatmadari, D. (2006, Jul/Aug). Is A Million Articles Proof of Authentic Information?
18 Intermedia, 34, 3, p. 17.
Fo
19
20 Shoemaker, P., & Reese, S. (1995). Mediating the message: Theories of influences on mass
21
22 media content. Reading, MA: Addison-Wesley.
r
23
24 Tankard, J. W. & Royal, C. L. (2005), Finding Out What's On the World Wide Web,
Communication Impact: Designing Research that Matters, Susanna Hornig Priest, Ph.D.,
Pe
25
26 Editor, p. 253-264.
27
28 Tankard, J. W. & Royal, C. L. (2005, Fall), What’s on the Web and What’s Not, Social Science
er
29
30 Computer Review.
31
32 Tankard, J. W., & Showalter, S. S. (1977). Press coverage of the 1972 report on television and
Re
33 social behavior. Journalism Quarterly, 54, p. 293-298.

34
35 The Wiki Principle. (2006, April 22). The Economist, 379, p. 14-15.
36
vi
37 Udell, J. (2004, January 9). Wikipedia’s Future. Retrieved from http://www.infoworld.com

38
39
Voss, J. (2005). Measuring Wikipedia. In Proceedings International Conference of the
ew
40
41 International Society for Scientometrics and Informetrics. Stockholm (Sweden).
42
43 Wallace, D. & Van Fleet, C. (2005). The Democratization of Information? Wikipedia as a
44 Reference Resource. Reference and User Services Quarterly, 45.
45
46 Wikipedia:About (2007), http://en.wikipedia.org/wiki/Wikipedia:About, accessed 1/23/2007.
47
48
49
50
51
52
53
54
55
56
57
58
59
60
28

1
2
3
Appendix A: Academy Award Winning Films
4
5
6 1928 Wings 1968 Oliver!
7 1929 Broadway Melody 1969 Midnight Cowboy
8 1930 All Quiet on the Western Front 1970 Patton
9 1931 Cimarron 1971 The French Connection
10
11
1932 Grand Hotel 1972 The Godfather
12 1933 Cavalcade 1973 The Sting
13 1934 It Happened One Night 1974 The Godfather Part II
14 1935 Mutiny on the Bounty 1975 One Flew Over the Cuckoo's Nest
15 1936 The Great Ziegfeld 1976 Rocky
16
1937 The Life of Emile Zola 1977 Annie Hall
17
18 1938 You Can't Take it With You 1978 The Deer Hunter
1939 Gone with the Wind 1979 Kramer vs. Kramer
Fo
19
20 1940 Rebecca 1980 Ordinary People
21 1941 How Green Was My Valley 1981 Chariots of Fire
22 1942 Mrs. Miniver 1982 Gandhi
r
23
24
1943 Casablanca 1983 Terms of Endearment
1944 Going My Way 1984 Amadeus
Pe
25
26 1945 The Lost Weekend 1985 Out of Africa
27 1946 The Best Years of Our Lives 1986 Platoon
28 1947 Gentleman's Agreement 1987 The Last Emperor
er
29
1948 Hamlet 1988 Rain Man
30
31 1949 All the King's Men 1989 Driving Miss Daisy
32 1950 All about Eve 1990 Dances With Wolves
Re
33 1951 An American in Paris 1991 The Silence of the Lambs

34 1952 The Greatest Show on Earth 1992 Unforgiven
35 1953 From Here to Eternity 1993 Schindler's List
36
1954 On the Waterfront 1994 Forrest Gump
vi
37
38 1955 Marty 1995 Braveheart
39 1956 Around the World in 80 Days 1996 The English Patient
ew
40 1957 The Bridge on the River Kwai 1997 Titanic

41 1958 Gigi 1998 Shakespeare in Love
42
43
1959 Ben-Hur 1999 American Beauty
44 1960 The Apartment 2000 Gladiator
45 1961 West Side Story 2001 A Beautiful Mind
46 1962 Lawrence of Arabia 2002 Chicago
47 1963 Tom Jones 2003 Lord of the Rings: Return of the
48 1964 My Fair Lady King
49
50 1965 The Sound of Music 2004 Million Dollar Baby
51 1966 A Man for All Seasons 2005 Crash
52 1967 In the Heat of the Night
53
54
55
56
57
58
59
60
29

1
2
3
Appendix B: Time Person of the Year
4
5
6 1927 Charles Augustus Lindbergh 1962 Pope John XXIII
7 1928 Walter P. Chrysler 1963 Martin Luther King Jr.
8 1929 Owen D. Young 1964 Lyndon B. Johnson
9 1930 Mohandas Karamchand Gandhi 1965 General William Childs
10
11
1931 Pierre Laval Westmoreland
12 1932 Franklin Delano Roosevelt 1967 Lyndon B. Johnson
13 1933 Hugh Samuel Johnson 1970 Willy Brandt
14 1934 Franklin Delano Roosevelt 1971 Richard Milhous Nixon
15 1935 Haile Selassie 1973 John J. Sirica
16
1936 Mrs. Wallis Warfield Simpson 1974 King Faisal
17
18 1937 Chiang Kai-Shek 1976 Jimmy Carter
1938 Adolf Hitler 1977 Anwar Sadat
Fo
19
20 1939 Joseph Stalin 1978 Teng Hsiao-P'ing
21 1940 Winston Leonard Spencer Churchill 1979 Ayatullah Khomeini
22 1941 Franklin Delano Roosevelt 1980 Ronald Reagan
r
23
24
1942 Joseph Stalin 1981 Lech Walesa
1943 George Catlett Marshall 1984 Peter Ueberroth
Pe
25
26 1944 Dwight David Eisenhower 1985 Deng Xiaoping
27 1945 Harry Truman 1986 Corazon Aquino
28 1946 James F. Byrnes 1987 Mikhail Sergeyevich Gorbachev
er
29
1947 George Catlett Marshall 1989 Mikhail Sergeyevich Gorbachev
30
31 1948 Harry Truman 1991 Ted Turner
32 1949 Winston Leonard Spencer Churchill 1992 Bill Clinton
Re
33 1951 Mohammed Mossadegh 1994 Pope John Paul II

34 1952 Elizabeth II 1995 Newt Gingrich
35 1953 Konrad Adenauer 1996 Dr. David Ho
36
1954 John Foster Dulles 1997 Andy Grove
vi
37
38 1955 Harlow Herbert Curtice 1999 Jeff Bezos
39 1957 Nikita Krushchev 2000 George W. Bush
ew
40 1958 Charles De Gaulle 2001 Rudy Giuliani

41 1959 Dwight David Eisenhower 2004 George W. Bush
42
43
1961 John Fitzgerald Kennedy
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
30

1
2
3
Appendix C: Artist with Billboard #1 Song First Week of February
4
5
6 1940 Tommy Dorsey 1974 Ringo Starr
7 1941 Artie Shaw 1975 Neil Sedaka
8 1942 Glenn Miller 1976 Ohio Players
9 1943 Harry James 1977 Rose Royce
10
11
1944 Glen Gray 1978 Player
12 1945 Bing Crosby 1979 Chic
13 1946 Vaughn Monroe 1980 Michael Jackson
14 1947 Sammy Kaye 1981 Blondie
15 1948 Vaughn Monroe 1982 Daryl Hall & John Oates
16
1949 Evelyn Knight 1983 Men at Work
17
18 1950 Andrews Sisters 1984 Yes
1951 Patti Page 1985 Madonna
Fo
19
20 1952 Johnny Ray 1986 Dionne Warwick
21 1953 Perry Como 1987 Billy Vera and The Beaters
22 1954 Eddie Fisher 1988 INXS
r
23
24
1955 Joan Weber 1989 Phil Collins
1956 Dean Martin 1990 Michael Bolton
Pe
25
26 1957 Guy Mitchell 1991 Surface
27 1958 Danny and the Juniors 1992 George Michael
28 1959 The Platters 1993 Whitney Houston
er
29
1960 Johnny Preston 1994 Mariah Carey
30
31 1961 The Shirelles 1995 TLC
32 1962 Joey Dee & the Starliters 1996 Boyz II Men
Re
33 1963 The Rooftop Singers 1997 Toni Braxton

34 1964 The Beatles 1998 Janet Jackson
35 1965 Petula Clark 1999 Britney Spears
36
1966 Simon & Garfunkel 2000 Savage Garden
vi
37
38 1967 The Monkees 2001 Destiny's Child
39 1968 John Fred & His Playboy Band 2002 Nickelback
ew
40 1969 Tommy James & the Shondells 2003 B2K

41 1970 The Jackson 5 2004 OutKast
42
43
1971 Dawn 2005 Mario
44 1972 Don McLean 2006 Nelly
45 1973 Stevie Wonder
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
31

1
2
3
Appendix D: Random Selection of Encyclopedia Terms
4
5
6 actin Jefferson Davis Sporting Record
7 Ain River John Davis Sterling Price
8 Albert Nile Kobenhavn Suceava
9 alkaline phosphatose League for the sun rose
10
11
An Srath Ban Independence of Vietnam temenggong
12 analytic geometry Lennart Torstenson The Spectator
13 Antigonish Leroy Randle Grumman Time magazine
14 Augsburg Confession Louis-Armand de Lom Tirso de Molina
15 August Baron Lambermont d'Arce Universal Declaration of
16
Big Sandy River Lydd Human Rights
17
18 Bogdah Mab urethane
botanical garden Maes videodisc
Fo
19
20 calamine brass Maravi Confederacy Wesselenyi Conspiracy
21 cecum Marc A. Mitscher William Pole
22 Central African Federation Marcus Eremita William Wrigley, Jr.
r
23
24
cesura marlin Zenshin
Copernicus Max Weber Ziya Gokalp
Pe
25
26 Cote-Saint-Luc Mistinquett
27 Darcy's Law Nicomachus
28 domestication Normandy Invasion
er
29
dragon Olympias
30
31 dysmenorrhea ostinato
32 Earl Carroll paper
Re
33 electromagnetic induction Paris Basin

34 equine Pastoral Epistles
35 Fort Portal Paul Signac
36
frame design Philip Schaff
vi
37
38 George Edward Stanhope Pierre Nicole
39 Molyneux Herbert Ponca
ew
40 Giacomo Meyerbeer PTA

41 gyroscope Quintus Fabius Pictor
42
43
Hans Geiger rampion
44 Haratin Republican River
45 Hei-ho Robert Lansing
46 Henry Jackson rose moss
47 Herman Busenbaum Rudolph Jacob Camerarius
48 horntail Russian Association of
49
50 humour Proletarian Writers
51 Ijma Saint George
52 Ismail Gasprinski Saint Irenaeus
53 Itaipu Dam Salem
54 Jack Miner Salon
55
56
Jacobus van Looy silica
57 James H. Doolittle Socialist Realism
58
59
60
32

1
2
3
4
5 Appendix E: Random Selection of Fortune 1000 companies
6 Air Products & Chemicals Ikon Office Solutions
7 Alberto-Culver ITT Industries
8 Allegheny Energy J.C. Penney
9 America West Holdings KB Home
10
11
American Greetings Kellogg Company
12 Armstrong Holdings Kellwood
13 AT&T Knight-Ridder
14 Auto-Owners Insurance Legg Mason
15 AutoZone Lehman Brothers
16
Avnet Lennar
17
18 Avon Products Lennox International
BellSouth Lockheed Martin
Fo
19
20 Benchmark Electronics LSI Logic
21 Beverly Enterprises Mandalay Resort Group
22 Boeing Manor Care
r
23
24
Briggs & Stratton Manpower Inc.
Cablevision Systems Marsh & McLennan
Pe
25
26 Chubb McKesson
27 Citizens Communications MDC Holdings
28 Colgate-Palmolive Mutual of Omaha
er
29
ConAgra National Fuel Gas
30
31 Countrywide Credit Northwest Airlines
32 CUNA Mutual Group Omnicare
Re
33 Eastman Kodak Phelps Dodge

34 Echostar Communications PNC Financial
35 Ecolab Primedia
36
El Paso Safeco
vi
37
38 Eli Lilly Schering-Plough
39 Energy East Scientific-Atlanta
ew
40 Equity Office Properties Sentry Insurance Group

41 Expeditors International of Washington Snap-On
42
43
Gap Inc. Sonoco Products
44 General Motors SPX
45 Genesis Health Ventures Stanley Works
46 Gold Kist Starbucks
47 Goodrich Sun Microsystems
48 Great Plains Energy Swift Transportation
49
50 H.B. Fuller Tenneco Automotive
51 Hershey Foods Thermo Electron
52 Hewlett-Packard Viacom
53 Hilton Hotels Walt Disney
54 Home Depot Western Digital
55
56
Hovnanian Enterprises
57 Humana
58
59
60
33

What's On Wikipedia and What's Not

Transféré par

Informations du document

Description originale:

Titre original

Copyright

Formats disponibles

Partager ce document

Partager ou intégrer le document

Options de partage

Avez-vous trouvé ce document utile ?

Ce contenu est-il inapproprié ?

Droits d'auteur :

Formats disponibles

What's On Wikipedia and What's Not

Transféré par

Droits d'auteur :

Formats disponibles

Journal of Computer-Mediated Communication

Whatâ s on Wikipedia, and Whatâ s Notâ ¦?

Journal: Journal of Computer-Mediated Communication

Manuscript ID: JCMC-07-186

Manuscript Type: Full-length Research Article

Wikis, Social Network Analysis, Information Richness, Online

International Communication Association

What’s on Wikipedia, and What’s Not…?

What’s on Wikipedia, and What’s Not…?

What’s on Wikipedia, and What’s Not…?

What’s on Wikipedia, and What’s Not…?

What’s on Wikipedia, and What’s Not…?

What’s on Wikipedia, and What’s Not…?

What’s on Wikipedia, and What’s Not…?

What’s on Wikipedia, and What’s Not…?

What’s on Wikipedia, and What’s Not…?

What’s on Wikipedia, and What’s Not…?

What’s on Wikipedia, and What’s Not…?

What’s on Wikipedia, and What’s Not…?

What’s on Wikipedia, and What’s Not…?

What’s on Wikipedia, and What’s Not…?

33 Time’s Person of the Year - By Year

What’s on Wikipedia, and What’s Not…?

What’s on Wikipedia, and What’s Not…?

What’s on Wikipedia, and What’s Not…?

What’s on Wikipedia, and What’s Not…?

What’s on Wikipedia, and What’s Not…?

What’s on Wikipedia, and What’s Not…?

What’s on Wikipedia, and What’s Not…?

What’s on Wikipedia, and What’s Not…?

What’s on Wikipedia, and What’s Not…?

What’s on Wikipedia, and What’s Not…?

What’s on Wikipedia, and What’s Not…?

What’s on Wikipedia, and What’s Not…?

25 Journalism Quarterly, 38, p. 441-452.

What’s on Wikipedia, and What’s Not…?

33 social behavior. Journalism Quarterly, 54, p. 293-298.

37 Udell, J. (2004, January 9). Wikipedia’s Future. Retrieved from http://www.infoworld.com

What’s on Wikipedia, and What’s Not…?

33 1951 An American in Paris 1991 The Silence of the Lambs

40 1957 The Bridge on the River Kwai 1997 Titanic

What’s on Wikipedia, and What’s Not…?

33 1951 Mohammed Mossadegh 1994 Pope John Paul II

40 1958 Charles De Gaulle 2001 Rudy Giuliani

What’s on Wikipedia, and What’s Not…?

33 1963 The Rooftop Singers 1997 Toni Braxton

40 1969 Tommy James & the Shondells 2003 B2K

What’s on Wikipedia, and What’s Not…?

33 electromagnetic induction Paris Basin

40 Giacomo Meyerbeer PTA

What’s on Wikipedia, and What’s Not…?

33 Eastman Kodak Phelps Dodge

40 Equity Office Properties Sentry Insurance Group

Vous aimerez peut-être aussi