Forced transparency: corporate image on Wikipedia and what it means for public relations
|Forced transparency: corporate image on Wikipedia and what it means for public relations|
|Authors:||Marcia W. DiStaso, Marcus Messner|
|Citation:||Public Relations Journal 4 (2): missing pages. 2010 Spring|
|Database(s):||Google Scholar cites|
|Web:||Bing Google Yahoo! — Google PDF|
|Article:||BASE Google Scholar PubMed|
|Restricted:||DTU Digital Library|
|Extract:||Talairach coordinates from linked PDF: CSV-formated wiki-formated|
Forced transparency: corporate image on Wikipedia and what it means for public relations is a study on articles on Wikipedia on 10 selected Fortune 500 companies with description of the reach, length, topics and "tonality" (sentiment), etc. in the articles. This interesting study is probably one of the largest and most thorough content analysis studies of Wikipedia. The researchers have performed a large multidimensional manual labeling of sentence in Wikipedia.
In the longitudinal study the generally found that negativity increased through the 3 years examined.
Marcus Messner blogged about the study in Watch your wiki!.
They were downloaded in February 2006, 2008 and 2010. Furthermore the revision history for yearly periods was also examined for each company.
The authors extract sentences from corporate Wikipedia articles.
- The sentences was manually labeled with respect to topic (topic mining). Topics was fixed to the categories company, historical, performance, employees, financial, competition, corporate social responsibility, legal concerns/scandals, international and "other".
- Manually labeling of "tonality" (manual text sentiment analysis) into positive, neutral and negative.
- Examined the search engine ranking of the Wikipedia articles
- Number of articles view.
- The number of edits ("rigor")
- The number of different contributors ("diversity")
- Article length
The analysis was longitudinal examining articles in 2006, 2008 and 2010.
- Wikipedia articles on corporation were on average around rank 8 in search engines in 2006 rising to around 3 in 2010 (Table 1).
- Wikipedia articles has hundreds of thousands of views per year. 1.4 million for Wal-Mart and 124 thousands for ConocoPhillips (Table 2).
- Of yearly statistics the 2008 (March 2007 to February 2008) showed typically the highest number of edits and number of different users (Table 3).
- The length of the articles increase over time (Chart 1).
- 3'800 sentences extracted.
- The tone/sentiment of corporate Wikipedia articles was more negative in 2010 than in other years.
- IBM has had the highest positive content, while AIG the lowest.
- The topics have changed over time, e.g., historical information has decreased with "legal concerns/scandals has increased.
- CSR is typically the most positive topic scoring highest in 2006 and 2010. "legal concerns/scandals" are the topic that is most negative. For most companies in 2010 "legal concerns/scandals" were the top topic with most negativity, while for Wal-Mart it was "employees". "Performance" and "CSR" was the most positive topics in 2010.
 Related papers
(All/some of) these are taken from the bibliography section of the articles:
- Collaboration, consensus, and conflict: negotiating news the wiki way
- Journalistic objectivity redefined: Wikinews and the neutral point of view
- Legitimizing Wikipedia: how U.S. national newspapers frame and use the online encyclopedia in their converage
- User-generated content on the Internet: an examination of gratifications, civic engagement and psychological empowerment
- Wikiganda: identifying propaganda through text analysis
- Wikipedia - free and reliable
- Wikipedia in the newsroom
- Wikipedia leeches? The promotion of traffic through a collaborative Web format
- Wikipedia's role in reputation management: an analysis of the best and worst companies in the USA
- The article states "A critic of a company, a former employee, or even someone completely unrelated to the company has the same editorial authority on Wikipedia as a corporate executive." But actually the corporate executive has a conflict of interest giving him lower editorial authority on Wikipedia.
- The article states "Thousands of popular articles (none of them corporate articles), such as the one of the U.S. president, are now locked for editing by only experienced administrators (Waters, 2009)." However, The George W. Bush article is actually semi-locked which means that unregistered, not autoconfirmed or confirmed editors cannot edit, while ordinary confirmed users can edit the article.
- The article states "The simple concept that any fact about an organization can be placed on a corporate article and not be removed if it is factual and has a citation". This is incorrect. Any kind of information can be removed from an article.
- It is unclear with the search engine ranking is done in a reproducible way. Search engines may personalize the ranking. It may be very difficult to get around this problem and most researchers just takes the Internet search engines "at face value"
- There is no statistical significance testing on the numbers.
- The article states "It is no surprise that AIG had the highest percentage of negative content for 2010 out of all the companies analyzed." It is unclear why it is no surprise.
- The article states "Although topics that many companies would rather not have highlighted (such as legal concerns or scandals) often have prominent placement, there is nothing a company can do (or should do) about it. That is, unless it is not factual information. Then, and only then, can a company edit the content in its Wikipedia article." This is questionable.
- The article states "This means that public relations practitioners can add content to an article, but it must be backed up by a citation." The PR company is in a conflict of interest and is "strongly discouraged" in Wikipedia.
- The article states "The rules of Wikipedia make it almost impossible to erase negative and controversial content from these articles and make it a very tedious task to implement changes." This is somewhat speculative, but might be right
- The exact dates/revisions when the articles were downloaded is not stated.