> Guessing if they measured satisfaction on a scale, the majority were clustered near the "no major change" camp, as with any gaussian curve. They did measure

> Guessing if they measured satisfaction on a scale, the majority were clustered near the "no major change" camp, as with any gaussian curve. They did measure satisfaction on a scale. From page four of the report: > Students responded on a sliding scale of 0–100 with five labelled categories, ‘Much worse’, ‘Slightly worse’, ‘The same’, ‘Slightly better’, and ‘Much better’, anchored at scores 0, 25, 50, 75, and 100, respectively. We simplified the scale in this analysis by transforming it into groups of those doing ‘worse’ (scores 0–37.5 on the sliding scale), ‘the same’ (scores 37.6–62.5), or ‘better’ (scores 62.6–100) (N.B. categories are not evenly divided due to how the scale is presented in the survey; please see https://osf.io/fpbt3/ https://osf.io/fpbt3/ for an example of the survey scales). Fitting a normal distribution with 1/3 of the volume in the 37.6–62.5 range ends up with 1/5 of respondents above the "Slightly better" mark (i.e. above 75).