Show/hide main menu

Analytical studies

Messages in Anti-Stigma Campaigns (MOSAIC): A consensus development study


Mental health-related stigma and discrimination are widespread and have a significant adverse impact on the lives of people with mental ill health. Mass media anti-stigma campaigns have been widely used to try to reduce mental health-related stigma and discrimination. The types of messages used in these campaigns vary and reflect different models e.g. biomedical messages, high prevalence messages, recovery messages, anti-dangerousness messages etc.

There are potential advantages and disadvantages with use of each type of message and disagreement about which types of messages are best. Consensus development methods, such as the nominal group technique, are a process for making policy decisions and their objective is to synthesise judgements when a state of uncertainty exists. As there is uncertainty about types of messages for use in mass media anti-stigma campaigns, there is a need for a consensus development study to measure, and if possible develop, consensus on this matter. Participants in consensus development studies are required to be in some way ‘expert’ on the matter under discussion. This may include formal experts with relevant specialist knowledge, as well as stakeholders who have direct experience of the issue in question.  


To develop and measure consensus on which types of messages should be included in mass media campaigns to reduce mental health-related stigma or discrimination.  


A nominal group technique study with two rounds of electronic voting was used to measure and develop consensus among multi-disciplinary experts and stakeholders in the field of anti-stigma and discrimination, attending an international conference on mental health stigma. Prior to the conference the research team extracted verbatim messages from national English language mass media mental health anti-stigma campaigns, and reviewed the theoretical and empirical literature on types of messages in such campaigns. Using these sources the team selected 10 types of message, with exemplar messages for each, reflecting a range of different message types.

Participants viewed PowerPoint slides, each containing the name of a type of message and some exemplar messages. Using the voting pads, they rated the type of messages on a 1-9 point scale to indicate whether they thought each type of message should be included in anti-stigma campaigns. The first round of rating was followed by audience members being presented with a summary of the research evidence on the effects of using different types of messages in mental health-related anti-stigma campaigns. A facilitated discussion followed, which was audio recorded to enable a qualitative analysis of the discussion. The audience was then asked to re-rate the types of message.   


A panel of 32 experts attending an international conference on mental health stigma participated in a consensus development exercise. The experts included stigma researchers, people implementing anti-stigma programmes, service users, informal carers and practitioners. There was high consensus (≥ 80%) regarding the inclusion of two of the message types presented – (i) 'recovery-oriented' and (ii) 'see the person' messages, and reasonable consensus (≥ 70%) regarding (iii) 'social inclusion / human rights' and (iv) 'high prevalence of mental disorders' messages. Ratings differed according to whether the participant was a psychiatrist or had personal experience of mental ill health.

Analysis of the qualitative data revealed four themes: (i) benefits of messages countering the ‘otherness’ of people with mental ill health; (ii) problematic nature of messages referring to aetiology; (iii) message impact being dependent on the particular audience; and (iv) need for specific packages of messages. The study supports the use of 'recovery-oriented' messages and 'see the person' messages. Social inclusion / human rights messages and high prevalence of mental disorders messages also merit consideration. The study has now been published (Clement et al 2010).


Sitemap Site help Terms and conditions Privacy policy Accessibility Modern slavery statement Contact us

© 2017 King's College London | Strand | London WC2R 2LS | England | United Kingdom | Tel +44 (0)20 7836 5454