Safety in Hong Kong's construction industry: changing worker behaviour

The behavioural safety management techniques of performance goal setting and the provision of posted performance feedback were tested for their effectiveness in bringing about improved safety performance on seven public sector building construction sites in Hong Kong. Goal setting and feedback, in combination, were introduced to three different categories of site safety: Housekeeping, Access to Heights and Bamboo scaffolding. Performance in these three categories plus a control category (Personal Protective Equipment) was measured for a total period of thirty four weeks. A multiple-baseline research design was used. After a baseline period of measurement, goal setting with feedback were introduced to the three different performance categories at staggered intervals. Results were analyzed using Autoregressive-Integrated-Moving Averages modelling techniques which are particularly applicable to the analysis of temporally ordered scores. The results of the analyses suggest that the behavioural safety techniques were highly effective in bring about improved performance in site Housekeeping during the intervention period. In addition to this, some results suggested a two-stage intervention effect in Housekeeping whereby performance improved significantly both immediately prior to the introduction of the intervention and with the intervention itself. In contrast to the Housekeeping findings, significant improvements in Access to Heights were only observed on two out of seven of the experimental sites when the techniques were introduced to the Access to heights category and there was no significant improvement in Bamboo scaffolding performance, with the introduction of the techniques, on any site. In addition to the measurement of safety performance, worker attitude factors were assessed in interviews. Structured and formal interviews with workers were held in the site offices of each of the seven sites. During these interviews workers were questioned directly about the level of affective commitment felt to the main contractor and their ganger. The results of the site office interviews indicated unexpectedly high levels of affective commitment to both main contractor and ganger. A second interview instrument was developed and administered on site, as opposed to in the site office, in order to verify the site office interview results. The on site levels of affective commitment felt towards main contractors and gangers are, in fact, very low. This difference suggests that great care should be exercised, when attempting to measure worker attitude factors, to ensure that the method of measurement does not distort worker responses to interview questions. There was evidence to suggest that worker commitments are based on compliance with requirements in order to receive monetary rewards and the perception that they are reliant on their main contractor or ganger for future or long term employment.