Respect: Results of a pilot project designed to improve behaviour in English football

Abstract

The research on which this article reports arose from recognition by The English Football Association (FA) that poor behaviour in affiliated football was having widespread and deleterious effects on the game, at every level, including a loss of about 7,000 referees each year. In order to address these concerns, The FA implemented a programme of pro-social behaviour change, branded ‘Respect’, and commissioned research into a 3 month pilot project in a small number of County FAs during the spring of 2008. In designing the evaluation for the Respect Pilot the researchers attempted to adhere to best practice in programme evaluation by addressing both the process and outcomes of the intended change programme (Coalter, 2007; Pawson, 2006; Weiss, 1998) and by working with a logic model that could be adapted over time (Aspen Institute, 2003; Schmitz and Parsons, 2007). The four main stakeholder groups identified to take part in the Pilot were players, coaches, referees and spectators/parents, for three age groups – Under 10, Under 16 and Open Age. Three different interventions were tested at the matches: designated spectator areas, codes of conduct with sanctions and only the captain being allowed to talk to the referee. Interviews were conducted with the four stakeholder groups at three games in each age group. In addition, week-by-week behaviour assessments from 583 trial games were entered online by over 1,000 people. Overall, both enjoyment and behaviour scores among the Active (experimental) teams were perceived to be significantly better than that among the matched Control teams (with no interventions). The experimental teams also ranked all three interventions highly over the ten weeks, with the designated spectator areas rated best, followed by the signing of codes of practice second and only captains talking to referees third. The results are discussed in relation to the future plans for the Respect campaign and the efficacy of the original logic model.