Widespread overestimation by young people of peer substance misuse and permissive attitudes is real. These misperceptions are also real in their consequences (reign of ‘error’), contributing to the significant actual problems of substance use in youthful populations [1–5]. Hundreds of studies have documented this phenomenon since it was brought to light some 25 years ago [6]. It is not uncommon, however, for researchers new to the topic to suggest, as do young people themselves, that the error might be in the measure of the actual norm, rather than in perceptions of norms. When first confronted with the gap between aggregate personal reports and perceived norms, a common tendency is to assume that personal reports are incorrect as a way to dismiss the dissonance. Thus, the suggestion that the perception gap might be an artefact of faulty methods is nothing new.

For those familiar with the extensive range of studies on misperceptions, however, there are many counterarguments, not the least of which is the large number (now hundreds) of studies in peer-reviewed journals and other readily available academic sources documenting the gap with multiple measures. Curiously, Pape's paper[7], presented as an empirical literature search, cites only about 25 original studies on misperceived norms. A thorough rebuttal of Pape's claims would require citing scores of omitted relevant studies but, alas, this commentary must be short with few references.

Pape suggests that the difficulty some respondents might encounter in performing mental calculations posed by frequency and quantity questions about personal use might lead to under-reporting. Given that concern, it is surprising that research examining injunctive (attitudinal) norms is excluded from the review without explanation. The measures of injunctive norms do not require complex calculations of past history and are not subject to personal recall error, yet they similarly document large overestimates of permissive peer attitudes compared to actual attitudes.

Pape suggests that young people might be biasing downwards their responses about personal behaviour in fear of reporting socially undesirable or illegal behaviours, but large actual-perceived norm gaps have been found in anonymous surveys in times and places where young people were of legal drinking age, and found with regard to legal tobacco use and in previous decades when there was little stigma about smoking. Also, when respondents believe the majority of peers are engaging in and support heavy drinking and the use of many other drugs, this can hardly be a situation where young people would be concerned about anonymously revealing their own drinking, tobacco use or other drug use levels while they think so many others approve and practise it. Furthermore, studies using anonymous breathalyzer assessments to determine actual drinking norms among students also support the finding that students typically perceive the norms for the amount of drinking among peers to be substantially greater than is actually the case, and that they do not, on average, under-report their own consumption [8,9].

Pape's paper also proposes that the actual-perception gap may be accounted for partially by reliance on convenience samples, studies with low response rates and use of arithmetic means to measure norms rather than median categories or measures of prevalence. In so doing, the author inexplicably ignores numerous studies published with extremely large nation-wide samples, as well as local population studies with representative samples and high response rates and studies that examine prevalence rates and medians, all still documenting a large gap (e.g. [1,2,6,10–16]).

Finally, only two published empirical studies actually questioning the misperception gap are reported [17,18]. One [17] is fundamentally flawed by its comparison of mixed measures [19]. Prevalence of personally consuming four/five or more (female/male) drinks at a sitting was about equal to the prevalence of perceived ‘binge’ drinking among peers, but only the word ‘binge’ was used for the latter measure (rather than a specifically defined amount) and most respondents interpreted ‘binge’ drinking as meaning six or more drinks (more than 20% thought it meant 10+). The other study [18] concluded unjustifiably that their finding of less misperception when the question about personal consumption was omitted meant an erroneously inflated peer norm perception when the question of personal behaviour was included. Inexplicably, that study does not consider the opposite—that without a separate personal question, one is likely to include and weight heavily one's own behaviour in the subjective estimation of the peer norm. Perceived peer norms, irrespective of one's own behaviour, and the inaccuracy of these perceptions among youth, are precisely what studies should and do document by including both measures.