Evaluating a Group-Based Sustainability Intervention Using the Theory of Planned Behaviour

Meeting the challenge of anthropogenic climate change will require widespread adoption of more sustainable behaviours. However, although attitudes towards sustainable behaviours are positive, actual change is lagging behind. Three studies explored the success of a classroom intervention programme that was intended to support individual change towards more sustainable behaviour in the domains of energy conservation and consumer responsibility. It was expected that identification with the small action groups used in the programme would have a positive effect. The theory of planned behaviour (TPB) (Fishbein & Ajzen, 1977) and the social identity perspective (Turner, 1999) were used as a framework for analysis. Studies one and two examined the success of the intervention programme across two iterations. Behaviour measures used in study one were inadequate but effective measures were developed for the second study. The intervention programme was very successful in achieving behaviour change and improving attitudes towards and intentions to perform sustainable behaviour. The TPB was supported by both studies, although there were unexpected inconsistencies in the variables predicting intent. Contrary to expectations, there was no effect found for group identification. Differences were also found between those participants who chose to focus on energy conservation and those with a focus on consumer responsibility, suggesting that the consumption group approached environmental behaviour in a more holistic way. Study three was a qualitative analysis of diary entries by participants in study one. A participant narrative of sustainable behaviour was constructed and related to attribution theory, particularly the Martinko and Thomson (1998) synthesis model. The narrative substantially matched the TPB but some problematic aspects of the intent construct in the TPB were identified. There was also evidence of a positive effect of group membership that had not been captured by the group identification variable. Potential interpretations and consequences of these findings were discussed.