The study by Abraham et al ^{Reference Abraham, Easow, Ravichandren, Mushtaq, Butterworth and Luty1} suggests that a single exposure to selected Time to Change campaign material (those including the ‘1 in 4’ message) delivered via post was not effective at improving attitudes towards people with mental illness. Findings were based on a sample of 250 adults recruited through various adverts. The study showed that attitudes were not significantly better than in a group of the UK general public previously recruited for scale validation.

We are undertaking the overall evaluation of the campaign. Our evaluation design is based on a conceptual framework which describes stigma as problems of knowledge (ignorance/misinformation), attitudes (prejudice), and behaviour (discrimination). Therefore, in addition to measuring prompted campaign awareness, our evaluation included measures of mental health-related knowledge (measured by the Mental Health Knowledge Schedule), attitudes (measured by the Community Attitudes towards Mental Illness scale) and behaviour (measured by the Reported and Intended Behaviour Scale). ^{Reference Evans-Lacko, London, Little, Henderson and Thornicroft2} To address the multifaceted nature of the campaign, we use several levels of evaluation, including assessments of: the overall programme at a national level, specific target groups (e.g. medical students, trainee teachers) and regional and local interventions. ^{Reference Henderson and Thornicroft3}

Our initial evaluation of the campaign in Cambridge used a pre/post-evaluation design among the campaign target population. These findings suggested modest but significant changes in this group. An important finding was that although campaign awareness was not sustained following the first phase of activity, significant and sustained shifts occurred for knowledge items 2 weeks following the campaign. There was a 24% (P<0.001) increase in the number of persons agreeing with the statement ‘If a friend had a mental health problem, I know what advice to give them to get professional help’, and a 10% (P = 0.05) rise in the number of people agreeing with the statement ‘Medication can be an effective treatment for people with mental health problems’. Over this short-term activity, changes were not evident for attitudinal or behaviour-related questions.

Another difference between our evaluation and that of Abraham et al is that we found familiarity with mental illness to be associated with less stigmatising responses. Therefore, our findings suggest the possibility of significant further progress via more openness, disclosure and social contact. It is clear from these studies that further investigation is needed to address the most effective dissemination and communication of anti-stigma messages. ^{Reference Clement, Jarrett, Henderson and Thornicroft4} Additionally, evaluation of the maintenance of changes over time and the additive effect of subsequent bursts of campaign activity will help us understand more about the effectiveness of this campaign in the long term. We are currently analysing data collected over the first year of the campaign.

Abraham et al also cite our paper comparing public attitudes in England and Scotland, ^{Reference Mehta, Kassam, Leese, Butler and Thornicroft5} and state: ‘Unfortunately, there have been reports that national anti-stigma campaigns are not particularly effective’. In fact, this paper shows the opposite, namely that ‘the results are consistent with early positive effects for the See Me anti-stigma campaign in Scotland’.