Using the theory of planned behavior to explore restaurant managers support for employee food safety training

Doctor of PhilosophyDepartment of Hotel, Restaurant, Institution Management and DieteticsElizabeth B. BarrettThis study examined the relationships among restaurant managers' attitudes, subjective norms, and perceived behavioral controls, using the Theory of Planned Behavior, that cause them to support or not support food safety training for their employees. The goal of this study was to investigate the facts that affect restaurant managers' willingness to support food safety training for their employees. Based on the results of the elicitation study, a pilot test focus group, and a pilot study, an 85 question instrument, that measured eight constructs and demographic information was developed. The instrument was administered via telephone, but response data was input using an online format. Data was entered electronically as it was collected. A total of 266 managers responded, but due to incomplete and missing data, 237 responses were usable. The final overall response rate was 17.9%. Results determined that overall intention to offer food safety training to employees was high. Restaurant managers had a positive attitude about food safety, placed importance on the beliefs of individuals they consider important, and felt they were in control about offering food safety training to their employees. Certified managers had more positive attitudes about offering food safety training and placed more emphasis on those individuals whom they considered to be important in their lives (subjective norms), but also perceived they had less control about offering food safety training to their employees. Most importantly, those managers who were certified had a higher intention to train employees than managers who did not have food safety certification. When comparing behavioral, normative, and control beliefs between chain and independent restaurant managers, only behavioral (specific attitudes) and normative beliefs (specific individuals whom they considered important) differed. In the multiple regression analysis exploring the prediction of behavior based on the respondents attitudes, subjective norms, and perceived behavioral controls, the model was significant (F=139.932, p [equal to or less than] .000) (Table 5.2). The significant independent variables in the model were the attitude mean composite score (Beta=-0.106, p [equal to or less than] .038) and the subjective mean composite score (Beta=0.727, p [equal to or less than] .000). Perception of control was not significant. Three multiple regression models were used to examine the relationships between the attitudes, subjective norms, and perceived behavioral controls as dependent variables and the factors for their indirect measures as independent variables. Of the three simple linear regression models, only two were significant. The regression of the behavioral belief factors on attitude composite score (Hypothesis 4) (F=16.714, p [equal to or less than] .000) and the regression of normative belief factors on the subjective norm composite score (F=11.896, p [equal to or less than] .000) were significant. The regression of control beliefs on perceived behavioral control showed no significance