St. Petersburg for This kind of finding could have implications for a range of groups; for example, hospitalised patients could be encouraged to form IIs about their home-based rehabilitation in order to improve exercise adherence and recovery post-discharge. While some people spontaneously form IIs ISSUES How the wording and ordering of questions may influence the data obtained It has been suggested that studies reporting evidence of unrealistic optimism UO in a sample may actually be witnessing a measurement artefact; i.e. UO may be appearing as a result of the manner in which the questions are asked Harris and Middleton Weinstein originally asked one simple question in an attempt to establish the presence of UO compared with others of my sex/age my chances of developing ? disease x? are great/average/ low’. In other studies, however, two questions have been asked the first generating a rating for personal risk, the second generating a rating for the risk of similar others see, for example, Perloff and Fetzer van der Velde et al. , and UO is considered to be present when the second rating is higher than the first. Hoorens and Buunk manipulated the ordering of these two questions and also the comparison group that their sample of adolescents were required to think of when making their risk judgements. St. Petersburg 2016.