Just Stuff

Just another WordPress.com weblog

Archive for February 2010

Not Lying Until Pressed?

leave a comment »

UK civil servants like most of us prefer not to lie when asked a direct question. This applies even when they cannot tell the truth for various reasons. When faced with such a question one tactic often employed is deflection. Deflection entails either answering a different question which appears to the casual oberver to be that asked, or to reply with a question.

The reason for this post is the recent allegations about Gordon Brown and bullying of Downing Street junior staff. A very senior civil servant has replied to questions which are equivalent to “have you had a word with the PM on the QT about his bullying junior staff”. The first two or three times the person in question replied each time with different deflections. Eventually the person in question had to give a direct answer and they denied that they had done so.

If that straight answer were in fact the truth there was no reason for deflection so we may conclude that it was probably a lie. They could not tell the truth if the answer were yes so eventually must lie, and we all know that. Now the questioners knew what deflection meant from the start. So instead of pursuing this question they could just have observed that the answer given was a deflection and we would draw the appropriate conclusions from that. There was and is no need to drive the subject to lying.

I suppose that just observing that the interviewee was deflecting etc would be considered bad television and so the Jeremy Paxmans would never adopt this approach.

Written by CaptainBlack

February 24, 2010 at 07:15

Posted in Blogroll

Tagged with

Peer Rejected

leave a comment »

It seems that some problems with abuse of peer review have been highlighted by the stem cell community [1].

My own experience with this was a paper that I wrote for SIAM Review [2] essentially criticising something they had published (on the speed of the domino effect). The criticism was that [3] presented a mathematical model of the domino effect the results of which were counter-intuitive and no attempt had been made to compare the model predictions with any experimental data despite there being a number of data sets in the literature. In fact the principle references in [3] did contain experimental data (slightly flawed in my opinion but that is not relevant here) which if used would have invalidated the model.

The problem with my paper might in the eyes of the editors (and possibly the referees, if the same referees were used as for the original paper) be that not only was I critical of the original paper but of the reviewers in not picking up that no attempt at validation had been made. This might also be construed as a criticism of the who refereeing and editorial policy at SIAM Review.

It took SIAM Review a year to decide that they would not publish this paper. IIRC the reason for rejection was that they did not accept papers that were critical of stuff they had already published. I am not able to recover the email to quote the exact wording as my work email system seems to have deleted all emails older than a few months. Why it took them a year to tell me that they were not going to publish  I don’t know. They could have determined that it did not satisfy their policy in about 5 minutes.

Peer review is not some panacea that guarantees that what is published is correct, methodologically sound and/or important. It is influenced by the the foibles of the editorial staff, policy and any number of other things. The need of researchers to publish guarantees that a large proportion of papers published are trivial (assuming they have no other defects).

It is also the case that a paper or report has not been published in a peer reviewed journal is not necessarily worthless. There has been a piece in the news recently about the IPCC using a report from mountaineers in the Alps about the state of glaciers. This is not a controlled experimental result nor has it been published in a peer reviewed journal (IIRC it was published in some mountaineering journal) but it is expert testimony and has some value in the absence of other data.

Of course I might have broken an academic taboo in sending [2] to SIAM Review without first informing the authors of [3], but in this case I consider that an irrelevance, they had published and there is nothing they can say that would retrospectively introduce validation into their paper also the criticism was not just of what they wrote but of the refereeing process itself. I’m afraid that I am not a very courteous person and I an not going to tell you that I think what you write is a heap of crap before doing so, which is why the internet is such a wonderful (virtual) place I can write what I want without asking anyone’s permission.

References

1. “Journal stem cell work ‘blocked'”, BBC News, http://news.bbc.co.uk/1/hi/sci/tech/8490291.stm

2. “Validation of a Model of the Domino Effect?”, Ron Larham arXiv.org paper: arXiv:0803.2898, http://arxiv.org/abs/0803.2898

3. “Domino Waves”, C.J Efthimiou, M.D. Johnson,  SIAM Review 49 (2007) 111

Written by CaptainBlack

February 3, 2010 at 08:30

Posted in Maths and Stuff

Tagged with