jump to navigation

Medialens dittoheads on IBC July 4, 2009

Posted by dissident93 in Media Criticism.
trackback

Medialens once labelled IBC as an “Iraq Western Media Body Count”. But after a long list of IBC’s non-western media sources was posted to the Medialens message board, the Medialens editors tried a different approach instead: “And what percentage of reports in the IBC database originate from those sources you list?” (ML message board, 4/6/2006)

Fast-forward 3 years. Two of Medialens’s anon-supporters, Walter and Gunnar, are still repeating this line (even though Medialens seem, sensibly, to have dropped it). For example, Walter complains that IBC “emphasize” their “90” sources rather than the subset of “20” which picks up most incident/death reports. The problems with Walter’s account are that:

  1. IBC don’t list “90” sources, they list nearly 200.
  2. IBC don’t “emphasize” the full list above any particular subset. (They simply provide a full list and analyses of percentage coverage of subsets).
  3. The fact that a given subset of 20 (or 10, 12, 30, 5 – pick a number) media sources picks up a large number of incidents/deaths, proportionally, has little to do with IBC, whose database simply reflects this real-world fact.

Walter refers to the 90/20 thing about a dozen times, apparently without realising that it’s his own construct (possibly based on a mistaken interpretation of a statistical breakdown posted by Gunnar: “90% of citations come from 21 sources”), and with about as much bearing on IBC as the Daily Mail on immigration. Of course, it’s not a question of IBC’s “usual” (or “typical”) “20 sources” (as Walter incorrectly puts it), but of an unequal distribution of media coverage in the real world, which means that there will be a “top 20” (or 10 or 5 or whatever) sources in terms of proportional coverage of incidents/deaths. For example, Al Sharqiyah TV is in IBC’s top 10 sources much of the time, because it picks up a large percentage of total incidents/deaths relative to other sources. Walter misconstrues the issue in terms of “whether twenty sources is as good as ninety”, as if IBC “usually” (or “typically”) restricts itself to monitoring only 20 sources (which doesn’t follow at all from the above statistical beakdown).

Walter claims that “IBC give a misleading account of their comprehensive range of sources”. I asked him to point me to this alleged “misleading account”, but he hasn’t responded to this request. He also writes: “Given the 20 sources issue, IBC’s ‘comprehensive’ seems an exaggeration”.

This shows remarkable confusion on Walter’s part. IBC’s compilation of corroborated reported deaths is “comprehensive” to the extent that it misses none – not to the extent that it contains an artificial, unrepresentative (and impossible) flat distribution of all sources. In the absence of any evidence (over a period of more than six years) that IBC is missing a significant number of reported deaths, it would be pointlessly silly to attack the claim that it’s a “comprehensive” database of reported deaths.

Gunnar frames his criticisms of IBC in the same confused way as Walter, but to make matters worse he adds the following ignorant fallacy: “IBC gives a long list of media the[y] apparently cover, when it comes down to it they actually only quote a few different sources in their database”.

I asked Gunnar to go away and count the number of sources cited in IBC’s database. He hasn’t replied yet. Perhaps he’s found more than “a few”? (Hint: IBC’s long list is titled ‘Sources used by Iraq Body Count’ [my emphasis], every source listed is assigned an abbreviation for database citation, and they all seem to make an appearance in the form of db citations). I’ve attempted to patiently explain some of this stuff to Gunnar on previous occasions, but it doesn’t seem to register, so I end up repeating myself every few years.

……………………………………………………………………………………………………………….

Comments

I’ve invited Walter and Gunnar to send their responses to me for inclusion here. Walter has chosen to post a response on a third-party message board – link here

Advertisements