In previous news blogs, we have reported on statistical tests for fraudulent papers   and on linguistic comparison of the fraudulent and not-fraudulent papers written by a single low rectitude researcher, Diederik Stapel. We now report a further linguistic study  comparing 253 retracted papers with the same number of matched papers that were not retracted. And yes, there were systematic differences between the fraudulent and control papers with greater use of obfuscating language detected by computer-based linguistic techniques in the retracted papers. The results were not confounded by nationality of author. However, as in the first paper cited above, these statistical associations did not translate into sufficiently accurate labelling for the method to be used as a diagnostic test. Nevertheless, the results are consistent with studies in criminology and tell us something about the psychological stigma of a guilty conscience.
— Richard Lilford, CLAHRC WM Director
- Lilford R. Hyperbole. CLAHRC WM News Blog. 31 October 2014.
- Lilford R. …And While Talking About Culture and Misbehaviour. CLAHRC WM News Blog. 8 May 2015.
- Markowitz DM, Hancock JT. Linguistic Traces of a Scientific Fraud: The Case of Diederik Stapel. PLoS ONE. 2014; 9(8): e105937.
- Markowitz DM, Hancock JT. Linguistic Obfuscation in Fraudulent Science. J Language Soc Psychol. 2015; 1-11.