An interesting example of statistical methods and interpretation in a humanities thesis getting the author into hot-water...

http://www.timeshighereducation.co.uk/story.asp?storycode=207790

No, I certainly wouldn't, but I do agree that sometimes there is a very fine line with statistics between attempting to present your data in a useful and interesting way that will capture people's attention, and manipulating the figures to the level where they are no longer a true representative of the facts. You only have to look around you to see examples of people using stats in a way which might distort a person's perception of them, even if it is not technically 'wrong'. For example, students' exam marks being manipulated where a class has done particularly poorly so that they fit a normal distribution, dodgy statistics used in advertising or published by the government that don't tell 'the whole story'. And let's face it, I reckon a lot of us play around with different types of analysis etc to find the one that 'shows' something up the best, and select the interesting results to publish whilst ignoring the ones that didn't really get us anywhere or support the theory we wanted them to etc. No, it's not good, and it's not right either, but I think it would be naive to believe that every researcher writes up their results exactly as they found them, without doing a little tweaking here or there. Personally, I wouldn't because my work is health-related and it would be hugely irresponsible to deliberately mis-report things, but I expect when writing up I will be faced with the usual decision of which bits to write up and which bits to drop simply because they didn't go anywhere and I don't have enough words to cover every single little bit. I suppose with other people there are further pressures such as where the funding has come from, and what they have been 'told' to find, competition between research groups, the pressures of publishing etc, that perhaps we are not fully exposed to yet too. Complicated! Best, KB