LINGUIST List 17.3131

Wed Oct 25 2006

Qs: Software for Contrasting Two Corpora of Papers

Editor for this issue: Kevin Burrows <kevinlinguistlist.org>

---

Directory         1.    Erick Turner, Software for Contrasting Two Corpora of Papers

---

Message 1: Software for Contrasting Two Corpora of Papers
Date: 23-Oct-2006
From: Erick Turner <turnereohsu.edu>
Subject: Software for Contrasting Two Corpora of Papers


I am interested in the topic of publication bias in the medical literature.I have two groups of journal articles, one in which the results arereported accurately and another in which the results have been 'spun'.

(By that I mean that they started out with one hypothesis, found that to benot statistically significant, so they reported on a post hoc endpoint thathappened to be to be significant, and wrote it up as if that was theoriginal hypothesis. (This has been dubbed HARKing, for hypothesizing afterthe results are known.)

I am interested in contrasting the spun with the 'unspun' articles to lookfor differences between them. I am thinking that some words or phrasesmight turn out to be hidden signals for spin.

To that end, I am wondering what software folks would recommend to handlethis best.

It could be on either the Mac or PC platform, with some preference for theformer, and a not-too-steep learning curve would also be preferred, but anysuggestions would be welcome.

Thanks.

Erick

Linguistic Field(s): Computational Linguistics                             Discourse Analysis                             Text/Corpus Linguistics