Statistics and Emotions

Tim Harford has convincingly gathered some information together about how bias infects our responses to statistics (https://www.theguardian.com/science/2020/sep/10/facts-v-feelings-how-to-stop-emotions-misleading-us). Clearly, people have always been 'picky' in their responses to data. They feel good about evidence that supports their view or find fault with or simply ignore data that refutes their already established viewpoint. This is something that is clearly exaccerbated by the internet, as we rapidly skim through masses of material. People are essentially emotional in their responses to statistics. The world falls into groups believing or not believing in climate change, the advantages of vaccination, the benefits of Brexit or that health is more important than the economy as a basis for a prolonged Covid-19 lockdown. Harford notes that we generally approve (and share) the attitudes of members of our group but tend to regard people who take the opposite stance as being cynical (or biased) in their alledged beliefs (we used to call this 'negative apperception'). They are not well-informed and 'open' -like us. People all (and this has been clear for some time, as we are social beings) want to be accepted and respected by their 'team' and even modify their language, so as not to be labelled a 'heretic'. Counter-intuitively (to me, at least), he suggests that intelligent, educated people are not actually more able to view statistics accurately, honestly and dispassionately. They are simply better (or more creative) at finding fault with 'inconvenient' evidence, that does not support their beliefs (i.e. they can delude themselves more effectively than the less bright). It is actually very difficult to convert people by presenting them with the 'facts'. Harford, not unreasonably, suggests that we take a moment to ask ourselves what we feel when exposed to a new statistical finding? If it makes us feel good, ask ourselves why that is (and vice versa)? This might stop people sharing material online that they might, with a few moments reflection, realise is suspect or partizan. He feels we shouldn't try to take emotion out of our responses to statistical 'findings' but we should try to assess whether emotion is colouring our reaction.