Holland's Theory of Personality and Occupation in a Road Safety Context

Monash University Accident Research Centre – Report #145 - 1998

Authors: N. J. Pronk & W. A. Harrison

Full report in .pdf format [713KB]

Abstract

This report details a study which aimed to confirm and extend earlier work conducted by the second author using a larger, more representative sample. A sample of 3699 drivers conducted in a telephone survey provided information about their occupations and driving record. The occupational information was used to generate personality orientation information which was then analysed in terms of its association with self-reported driving data. It was concluded that it is possible to use personality information derived this way to identify subgroups of offending drivers (confirming Harrison, 1996, 1998c), and discriminant models predicting both self-reported speeding and drink-driving included personality information derived from occupations.

Executive Summary

This study aimed to confirm and extend earlier work conducted by Harrison (1996, 1998c) who used a theory of the relationship between occupation and personality as the basis for an investigation of the relationship between personality and drink-driving. The current study used a larger (3,699 participants) and more representative (drawn from a telephone survey of drivers) sample to confirm Harrison's original findings and to extend them in a number of ways.

The study was based on the same theoretical approach to personality, derived from Holland (1985) and others. The Holland theory describes personality (as related to occupational choice) in terms of differing strengths in six personality orientations. Harrison had shown that some groups of drivers defined in terms of their Holland orientations were over-represented in a sample of drink-drivers.

The study had four main aims:

  • To describe the distribution of Holland occupational codes in the sample of drivers interviewed in a telephone survey. It was concluded that the driver sample was generally representative of the wider population with the exception that the quota sampling used in the survey resulted in a larger number of rural participants than expected based on the population.
  • The second aim was to use the present data set to investigate the concerns raised about Harrison's (1996, 1998c) study. The results suggested that the concerns raised by Harrison (1996, 1998c) were well founded. This outcome suggested that the present data set would be a more appropriate normative data set for Harrison's original analysis.
  • The third aim of the present study was to compare the personality orientations of Harrison's (1996, 1998c) sample of drink drivers with the present sample used as a normative sample. The results largely confirmed Harrison's original findings. The results reported by Harrison and confirmed here are largely consistent with the literature concerning drink-driving and personality.
  • The final aim was to investigate the possibility that self-reported drink-driving and speeding in the present sample were related to other information collected from each respondent at the time of the survey. This analysis suggested that speeding and drink-driving were both related to personality characteristics identified from the Holland codes, and that in the case of speeding there were a number of additional characteristics associated with the behaviour.

The results of this study confirm the potential value of the application of the Holland model to road safety research, and more importantly provide further evidence that it may be possible to target countermeasures more effectively with information about the psychological characteristics associated with unsafe driver behaviour.

Sponsoring Organisation: Monash University Accident Research Centre