Stuart Rennie, University of North Carolina-Chapel Hill
3 October 2005
A contribution I made to a leading bioethics blog (http://blog.bioethics.net/2005/02/gilead-saga.html)was cited in the article by Mills et. al. as an example of inaccurate media reporting. I disagree with this judgment, for the reasons stated below. The reasons are not merely personal; they also expose weaknesses in the study as a whole.
First, there is a methodological problem with the study's inclusion criteria. The author's write "We included articles from any media source, but excluded non-media articles posted in non-governmental organization (NGO) activist websites as these contain mostly blogs or position papers and are not intended to be objective." If this criterion was applied consistently, it would also have excluded contributions (like my own) to the American Journal of Bioethics Editor's Blog. But apart from that question, Amnesty International, Doctors without Borders, or the International Crisis Group will not agree that their websites 'do not intend to be objective.' That exclusion criterion needs to be justified and applied consistently.
Second, the methods of the study are apparently not sensitive to the distinction between direct and indirect reporting. The piece I wrote was devoted to what people were saying about the tenofovir trials, not what was true of those trials. If the content analysis cannot pick out this common distinction, it makes one wonder about the accuracy of the other data cited in the study. This point is significant, because the research seeks to promote accuracy in reporting.
Misrepresentation in article data
3 October 2005
A contribution I made to a leading bioethics blog (http://blog.bioethics.net/2005/02/gilead-saga.html)was cited in the article by Mills et. al. as an example of inaccurate media reporting. I disagree with this judgment, for the reasons stated below. The reasons are not merely personal; they also expose weaknesses in the study as a whole.
First, there is a methodological problem with the study's inclusion criteria. The author's write "We included articles from any media source, but excluded non-media articles posted in non-governmental organization (NGO) activist websites as these contain mostly blogs or position papers and are not intended to be objective." If this criterion was applied consistently, it would also have excluded contributions (like my own) to the American Journal of Bioethics Editor's Blog. But apart from that question, Amnesty International, Doctors without Borders, or the International Crisis Group will not agree that their websites 'do not intend to be objective.' That exclusion criterion needs to be justified and applied consistently.
Second, the methods of the study are apparently not sensitive to the distinction between direct and indirect reporting. The piece I wrote was devoted to what people were saying about the tenofovir trials, not what was true of those trials. If the content analysis cannot pick out this common distinction, it makes one wonder about the accuracy of the other data cited in the study. This point is significant, because the research seeks to promote accuracy in reporting.
Competing interests
none