In our recent lecture we discussed the uses of mathematics in science. Data and statistics is a concept within mathematics that can be found in many other fields. For example, it is used in science or medicine. While mathematics in data and statistics is considered to be a very accurate measure, this is not always the case. The main issue that arose during the lecture was the impact of publication bias on the representation of data. I wanted to further look into this issue to consider how mathematics can be presented in a bias way and how this impacts society. I came across this very interesting TED talk by Ben Goldacre which inspired me to look into the issue further.

The term publication bias refers to when research that appears in published literature is unrepresentative of the population of completed studies (Rothstein, Sutton and Borenstein, 2005, p.1). This can be particularly misleading and even dangerous when considering the testing of new drugs. Much of the negative or inconclusive data remains unpublished which leaves doctors and researchers in the dark (Goldacre, 2012). Goldacre (2012) argues that positive data is twice as likely to get published than negative data. This can create huge misconceptions about a drug as part of the evidence is simply withheld by not being published.

One example of the dangers of publication bias occurred in the 1970s with the drug lidocaine. This anti-arrhythmic drug was said to reduce the risk of early death after a heart attack (Hampton, 2015). However, many trials to test the effects on mortality of this particular drug were not reported until the 1980s. Only at that time did it become apparent that the drugs actually increased mortality (Hampton, 2015). Publication bias meant that these drugs were promoted by doctors without any knowledge of the actual impacts of the drug. This is due to the fact that the negative data did not get published. This however is only one of many examples of publication bias in medicine. Goldacre (2012) emphasises the vast amount of negative data that remains unpublished.

Future Practice

I was intrigued to find out more about publishing bias, as this is something I had not considered previously when thinking about data and statistics. Implications of presenting bias data can be extremely dangerous to society and create misconceptions around certain drugs in medicine. In future this will help me analyse data more critically, not simply looking a the graph but questioning the validity of the data. This will also be useful for my future practice as a teacher, especially in doing science experiments. I will aim to convey that all results the students get should be acknowledged and need to be considered for it to be a fair experiment.

References

Hampton, J. (2015) Therapeutic fashion and publication bias: the case of anti-arrhythmic drugs in heart attack. Available at: http://www.jameslindlibrary.org/articles/therapeutic-fashion-and-publication-bias-the-case-of-anti-arrhythmic-drugs-in-heart-attack/ (Accessed: 2 December 2015).

Image. Photograph. Available at: http://www.medline.com/media/catalog/sku/abb/ABB427902H_HRE01.JPG (Accessed: 2 December 2015).

Rothstein, H., Sutton, A. and Borenstein, M. (2005) ‘Publication Bias in Meta-Analysis’, in Rothstein, H., Sutton, A. and Borenstein, M. (ed.) Prevention, Assessment and Adjustments. Place of publication: John Wiley & Sons, Ltd, pp.1-7.

TED (2012) Ben Goldacre: What doctors don’t know about the drugs they prescribe. Available at: https://www.ted.com/talks/ben_goldacre_what_doctors_don_t_know_about_the_drugs_they_prescribe?language=en#t-365862 (Accessed: 2 December 2015).