Ulrich Lichtenthaler, who has already retracted five papers for statistical irregularities, has retracted a sixth, “The Impact of Accumulating and Reactivating Technological Experience on R&D Alliance Performance,” in the Journal of Management Studies. The notice, which was first reported by the Open Innovation Blog, reads:
The following article from the Journal of Management Studies, The Impact of Accumulating and Reactivating Technological Experience on R&D Alliance Performance by Holger Ernst, Ulrich Lichtenthaler and Carsten Vogt, published online 17 March 2011 on Wiley Online Library (http://www.wileyonlinelibrary.com), has been retracted by agreement between the authors, the journal’s General Editors, Andrew Corbett, Andrew Delios and Bill Harley and Blackwell Publishing Ltd. The article is retracted due to errors in the reported empirical results, which form part of the basis for the conclusions drawn by the authors in the study. While the second author did not collect the data, he takes the responsibility for these technical errors.
The study has been cited twice, both times by papers in the Journal of Management Studies, according to Thomson Scientific’s Web of Knowledge.
“While the second author did not collect the data, he takes the responsibility for these technical errors.”
Very noble but I do not think you can do it. I mean, Ulrich Lichtenthaler bears responsibility for the paper but the person who cooked the data cannot be absolved by transferring his responsibility to UL. Voluntary scapegoating does not fly.
If the entire Journal of Management Studies was filled with cooked data from cover to cover :
A. Would anybody be able to tell the difference?
B Would anybody care?
interesting observation, would you care to develop it a little further?
Well I hope I am not being overly cynical here, but these types of journals are essentially vanity publications – places for academic non-managers to justify their academic position and perks.
Real medical specialists read journals in their specialty – I doubt if many managers read the journal of Management Studies
(A) It depends.
(B) Yes.
Did someone else notice, that Ernst withdraws being co-author from papers already being published as online-first and thereby leaves Lichtenthaler alone:
http://onlinelibrary.wiley.com/doi/10.1111/j.1540-5885.2012.00957.x/abstract
http://www.whu.edu/forschung/forschung-an-der-whu/publikationen/
Also mysterious:
The WHU publication page lists two more articels within Journal of Product Innovation Management as forthcoming. But they do not apear online anymore. It seams that two more papers got retracted…
Lichtenthaler, U., Ernst, H., Hoyer, W.: Determinants of Absorptive Capacity: The Value of Technology and Market Orientation for Open Innovation, Journal of Product Innovation Management (forthcoming)
Lichtenthaler, U., Ernst, H., Lichtenthaler, E. Desorptive Capacity: A Capability-Based Perspective on Commercializing Knowledge Assets, Journal of Product Innovation Management (forthcoming)
The WHU page no longer lists these. I guess this is part of Ernst cutting his ties to Licthenthaler
But this should mean, that these two papers got retracted two and the journal did not announce that, since those articles were online first?
Out of curiosity, I just perused the paper The Impact of Aligning Product Development and Technology Licensing: A Contingency Perspective in The Journal of Product Innovation Management. When you look at the regression Output they report, you can fairly easily see that SE’s and coefficients don’t add up to the reported significance levels. I ran these through a p-value calculator, an found the following:
1. The t-statistic for the interaction term in model 3 is reported as p<.05 when it in fact lies between .051 (one-tailed) and .102 (two-tailed).
2. The t-statistic for the interaction term in model 4 is reported as p<.05 when it in fact lies between .038 (one-tailed) and .076 (two-tailed).
3. The t-statistic for the interaction term in model 5 is reported as p<.1 when it in fact lies between .128 (one-tailed) and .257 (two-tailed).
4. The t-statistic for the interaction term in model 6 is reported as p<.05 when it in fact lies between .069 (one-tailed) and .139 (two-tailed).
The cheating is so obvious and easily detected so we probably don't even think that anyone will attempt it. That is probably why it worked — for a while at least! Poor guy — he obviously has a talent for writing. A shame that his science is corrupt!
(p-values courtesy of Daniel Soper, http://danielsoper.com/)
The editor informs me that they are conducting a full scale investigation of all the works published by Lichtenthaler.
“The editor informs me that they are conducting a full scale investigation of all the works published by Lichtenthaler.”
Maybe thats why, we haven’t seen a retraction in a while?
So who checks Prof. Ernst???
Retraction number 7 I believe: http://orgsci.journal.informs.org/content/23/6/1781; see also: http://orgsci.journal.informs.org/content/21/5/1054
To be continued…