Comparing personality constructs of risk-taking behavior

https://doi.org/10.1016/S0191-8869(98)00048-8Get rights and content

Abstract

Canadian subjects (N = 120) completed four psychological scales. Three of the scales were directly related to risk-taking behavior. These were the Sensation Seeking Scale (SSS V; Zuckerman, 1979), the Tension Risk Adventure Inventory (TRAI; Keinan et al., 1984) and the Telic Dominance Scale (TDS; Murgatroyd et al., 1978). In addition, subjects completed the Desire for Control Scale (DCS; Burger and Cooper, 1979) to investigate the role of desire for control in risk-taking. The results indicated (1) significant differences between age and sensation seeking (2) a four-factor structure for the TRAI following factor analysis with this Canadian sample (3) numerous intra- and inter-scale relationships obtained from correlation analyses of and (4) four major dimensions arising from PCA of the four inventory scales.

Introduction

Real life experience indicates that many people have the desire to experience challenges, face situations of some uncertainty, or improvise solutions to problems under difficult circumstances. However, large individual differences in peoples desire to take risks or to avoid possible hazards have been recognized in psychology for some time (Eysenck, 1967; Hodgdon et al., 1982; Slovic et al., 1982; Wilde, 1986; Zuckerman, 1979). With topics like decision-making under uncertainty, the prevention of accidents in traffic and industry, working in extreme environments (e.g. in space or underwater), deviant or negativistic behavior, risk or adventure sports and activities, gambling and unsafe sex, the notion of risk-taking behavior often plays a central role (Apter, 1992; Hoyos, 1980; Kerr, 1994; Zimolong and Trimpop, 1992).

Perhaps the most widely known research on risk-taking behavior is Zuckermans work on sensation seeking. Based on initial results from his sensory deprivation experiments (Zuckerman, 1979) and subsequent work on many aspects of risk-taking in everyday life (Donovan et al., 1985; Kalichman and Rompa, 1995; Zaleski, 1984; Zuckerman, 1983a; Zuckerman and Neeb, 1980; Zuckerman, 1994). More recently, Zuckerman, 1994has revised his definition of the personality trait of sensation seeking as the seeking of varied, novel, complex and intense sensations and experiences and the willingness to take physical, social, legal and financial risks for the sake of such experiences.fn2 Zuckerman (Zuckerman, 1980, Zuckerman, 1983b, Zuckerman, 1994) also focussed on the biological components of sensation seeking, arguing that biological disposition, reactivity and the presence of biological correlates of sensation seeking are genetically determined (Fulker et al., 1980).

Over a period of thirty years (1964–1994), Zuckerman developed the Sensation Seeking Scale (SSS; Forms I–VI) which, over time, underwent several revisions. A four-factor structure (thrill and adventure seeking, experience seeking, disinhibition, boredom susceptibility) was originally obtained from factor analysis of an American samples SSS scores (Zuckerman, 1974). This four factor structure was confirmed by research on a British sample (Zuckerman et al., 1976) and found further support in the results of Ball et al. (1984)Australian study, Rogers (1987)and Rowland and Franken (1986)studies in Canada and Birenbaum and Montags (1987)study of an Israeli sample. Zuckerman, 1994asserts that Lisrel analyses reveal a hierarchical factor model with a broad general factor contributing to the variance in all of the subscales, together with narrower more specific factors in the subscales, best explains the relations among subscales (p. 55).

In addition to the reported factor structure, Zuckerman, 1979, Rowland and Franken (1986)and Ball et al. (1984)reported similar findings with regard to sex and age differences, with men and younger people (age 17–21) scoring higher on sensation seeking. As regards age differences, Zuckerman, 1994found elevations in sensation-seeking between 9–14 years of age, reaching a peak in late adolescence and declining steadily thereafter and longitudinal studies suggest that these changes seen in cross-sectional studies are real age changes rather than generational differences.

Overall, with the exception of experience-seeking (which comprise some culture-specific items) the other three sensation-seeking subscales display good cross-gender and cross-cultural replicability (Zuckerman, 1994).

The SSS has been used to study the personality characteristics of participants in dangerous or risky sport activities. Correlations of high SSS scores with increased preference for and participation in high risk sports were found for activities such as parachuting (Hymbaugh and Garrett, 1974; Rowland et al., 1986), hang-gliding and auto-racing (Straub, 1982), motorcycling (Brown et al., 1974), scuba-diving (Heyman and Rose, 1980), mountain climbing (Robinson, 1985; Fowler et al., 1980; Freixanet, 1991) and downhill skiing (Bouter et al., 1988). Similar findings were reported for physical contact sports (Zuckerman, 1983b) and for participation in a greater variety of sports (Rowland et al., 1986).

Overall, Zuckerman, 1994summarised those who engage in risky sports are interested in a broad range of thrill-seeking outdoor activities that provide novel and stimulating experiences, but they are not necessarily social sensation-seekers, nor are they necessarily boredom susceptible.

It has also been shown that sensation seeking subjects tend to choose jobs involving high levels of change, flexibility, or some kind of risk (Zuckerman, 1979). A study of individuals in risky professions (e.g. fire-fighters, mountain-rescue personnel), risky sports (race car drivers, mountain climbers) and a non-risk control group showed that the risk sport group scored higher on the SSS than the professional risk group, which scored higher than the non-risk group (Zaleski, 1984). Musolino and Hershenson, 1977showed that subjects with occupations that placed them or others in jeopardy (e.g. air-traffic controllers, pilots) scored significantly higher on the SSS than civil servants and college students. Kish and Donnenwerth (Kish and Donnenwerth, 1969, Kish and Donnenwerth, 1972) reported negative correlations between SSS scores and clerical interests, and positive correlations with scientific interests. Thus, the SSS has proved useful in identifying differences in risk-taking behavior between different subject groups and has become relatively well-established in the psychological literature. The value of distinguishing the various facets of sensation-seeking however, was underlined by Goma et al. (1988)who found that although firemen scored higher than students on total, thrill and adventure seeking and experience seeking subscales, they did not differ on Disinhibition.

A perhaps less well-known measure, the Tension Risk Adventure Inventory (TRAI), was developed by Keinan et al., 1984as a global measure of risk-taking behavior for use in the selection of candidates for high risk occupations. Like the SSS, it is a trait measure. Findings from previous studies (Musolino and Hershenson, 1977) had suggested that risk-taking was related to a preference for certain occupations. Keinan et al., 1984used three separate Israeli subject samples, which included occupational groups such as security service agents, commando soldiers and pilots and compared them with less risky occupations like bank employees, to evaluate the reliability and validity of the instrument (see Measures). To the authors knowledge, no comparisons of the TRAI with other scales or on other subject samples has been attempted.

A third measure, the Telic Dominance Scale (TDS; Murgatroyd et al., 1978) is also a scale which can be used to measure risk-taking in the form of telic (arousal avoiding) or paratelic (arousal seeking) dominance in individuals. The scale has its origin in reversal theory (Apter, 1982). The TDS has three subscales planning orientation, serious mindedness and arousal avoidance. The scale has proved useful in showing that paratelic dominant arousal seekers are more likely than their telic dominant, arousal avoiding peers to engage in risky activities like dangerous or risky sports (Chirivella and Martinez, 1994; Kerr, 1991; Kerr and Svebak, 1989), sports characterized by explosive action (Svebak and Kerr, 1989), delinquency (Bowers, 1985), opiate addiction (Doherty and Mathews, 1987), gambling (Anderson and Brown, 1987) and work as undercover and field narcotics agents (Girodo, 1985).

Martin et al. (1987)examined the relationship between stress and telic/paratelic dominance and found differences in stress response. Paratelic dominant subjects were found to respond positively to moderate levels of stress whereas telic dominant subjects response to stress increased linearly with increasing levels of stress. Interestingly, Sarason, 1978and Cohen (1982)reported similar findings with the SSS (Zuckerman, 1979, Zuckerman, 1994). Sensation seeking was found to act as a moderator variable between negative life changes and subjects distress, with lower sensation seekers experiencing more stress than higher sensation seekers, for similar life changes. In fact, the TDS (Anderson and Brown, 1987) and the TDS subscales arousal avoidance and planning orientation (Cowles and Davis, 1985; Murgatroyd, 1985) have been found to correlate highly significantly and negatively with Zuckerman, 1974SSS, Form IV.

When examining particular personality scales, their relationship with other personality measures is often a point of concern. Significant correlations between other personality measures and the SSS and TDS have been found in several studies. For example, Eysenck and Zuckerman (1978)found a significant correlation between the impulsivity dimension from the Eysenck Personality Inventory (EPI; Eysenck and Eysenck, 1976) and the thrill and adventure seeking subscale of the SSS, while Murgatroyd, 1985cites a number of studies which have obtained significant correlations between TDS arousal avoidance and EPI introversion. Conversely, in an analysis between Eysencks broad impulsivity, empathy and venturesome scales and SSS Form V, Corrulla (1989)found SSS total score correlated low but significantly with venturesome, but was unrelated to impulsivity. Campbell and Palus (1987)reported correlations of the magnitude of +0.38 (impulsivity) and +0.77 (venturesome) with SSS total scores.

Zuckerman, 1994reported correlations in the expected direction between the SSS and Cattells (1957)16PF test, the California Personality Inventory (CPI; Gough, 1957), and the Minnesota Multiphasic Personality Inventory (MMPI; Hathaway and McKinley, 1951). Also, Burkhart et al. (1978)reported negative correlations between SSS scores and dimensions of anxiety. Murgatroyd et al., 1978reported correlations, also in the expected direction, between the TDS and the trait measure of anxiety developed by Spielberger et al. (1970)and between the TDS and Robinson, 1961Need for Achievement measure.

Kohn et al. (1982)and Goldman et al. (1983)reported significant negative correlations of high SSS scores with the reducing component of the Augmenter-Reducer Scale (Vando, 1974) and Cowles and Davis, 1985found a significant negative correlation for the TDS with the same scale. Neary and Zuckerman (1976)reported positive correlations of sensation seeking with the strong nervous system type proposed by Nebylitsyn (1972). More recently, however, Zuckerman, 1994commented Sensation-seeking is weakly related to all three of the Pavlovian dimensions as measured by the subscales of the Strelau Temperament Inventory (STI), positively with strength of excitation and mobility and negatively with strength of inhibition (p. 97). Finally, Furnham (1984)and Ortega and Pipal (1984)reported positive correlations of sensation seeking with measures of Type A behavior (Jenkins et al., 1967), but Apter, 1984reported no significant relationship between the TDS and Type A behavior.

Piët, 1987 has reported that professional stuntmen experienced their profession as risky but controlled and that they spent a lot of their time making the situation controllable, while still maintaining the pleasantness of the high levels of arousal intrinsic to the activities. The importance of perceived control, familiarity, security and predictability to induce the paratelic (arousal seeking) state has also been stressed by Apter, 1982. In the case of paratelic dominant arousal seekers, enjoyment of high arousal associated with taking risks is likely to take place under circumstances where they have a relatively high degree of subjectively perceived control.

Some research on desire for control and risk-taking behavior has been carried out using the Desire for Control Scale (DCS; Burger and Cooper, 1979) which assesses individual differences in the level of motivation to exert personal control over ones life. The scale has five subcomponents: a general desire or wish for control, decisiveness, preparation-prevention control, avoidance of dependence, leadership. Burger and Schnerring (1982)and Burger and Smith (1985)studied gambling behavior and found that the desire for control scores predicted the type of game that problem gamblers bet on and the highest amount of financial loss during a gambling year. Other research on risk-taking behavior in gambling (Strickland et al., 1966; Burger, 1986) as well as driving (Montag and Comrey, 1987) has shown that risk-taking is influenced by the desire for personal control or belief in personal control. Due to the apparent importance of personal desire for control in risk-taking, the DCS was included in this study along with the TDS, the SSS V and the TRAI. As far as could be ascertained, no previous comparisons between SSS, TDS and TRAI and DCS measures have been reported.

In this study, the authors were interested in examining a number of aspects of these different risk-taking-related scales, including: (1) the role that age might play in risk-taking behavior (common observations suggest that young persons are more inclined to take risks than older individuals); (2) the internal structure of the TRAI and in particular, whether a single personality construct of risk-taking could be identified in the present Canadian subject sample; (3) the intra- and inter-scale relationships between personality scale subcomponents (SSS V, TRAI and TDS), and (4) the underlying factorial structure between these scales.

Section snippets

Participants

Male subjects (N = 120), aged between 16 and 29 years (M = 22.5 years) were recruited by advertisement (poster) in the City of Kingston, Canada. Subjects took part in a study examining risk-taking behavior in experimental situations (Trimpop, 1994) in which they could receive a cash reward ($40–70) depending on their performance. As a part of this larger study, subjects also completed the four psychological questionnaires at focus in this study (see Measures).

Measures

The four questionnaires

Data analysis

In terms of statistical analysis, first, linear discriminant analysis and univariate F-tests were used to test for differences in subject responses related to age. Second, factor analysis was carried out on subjects TRAI scores. Third, Pearson product-moment correlations were used to compare SSS V, TRAI, DSC and TDS data. Fourth, factor analysis techniques were undertaken on all of the scales to identify possible multidimensional-factors and fifth, ANOVA techniques (2×2) were used to examine

Discussion

With respect to age differences, older subjects did score significantly lower on SSS V (and TRAI) than their younger peers, a finding which was in agreement with the results of previous research work undertaken by Ball et al. (1984), Zuckerman and Neeb, 1980and Rowland and Franken (1986). No age differences were apparent for telic dominance although there is some evidence (Murgatroyd, 1985) that telic dominance increases with age suggesting perhaps, that as individuals get older they may

Unlinked Bibliographic References

Björg-Akesson (1990), Zuckerman, 1971

References (89)

  • M.J Apter

    Journal of Research in Personality

    (1984)
  • M Birenbaum et al.

    Personality and Individual Differences

    (1987)
  • L.M Bouter et al.

    Personality and Individual Differences

    (1988)
  • J.M Burger

    Journal of Research in Personality

    (1986)
  • E.C Chirivella et al.

    Personality and Individual Differences

    (1994)
  • L.H Cohen

    Personality and Individual Differences

    (1982)
  • W.J Corulla

    Personality and Individual Differences

    (1989)
  • A Furnham

    Personality and Individual Differences

    (1984)
  • P Horvath et al.

    Personality and Individual Differences

    (1993)
  • C.D Jenkins et al.

    Journal of Chronic Diseases

    (1967)
  • J.H Kerr

    Personality and Individual Differences

    (1991)
  • J.H Kerr et al.

    Personality and Individual Differences

    (1989)
  • R.F Musolino et al.

    Journal of Vocational Behavior

    (1977)
  • T.B Rogers

    Personality and Individual Differences

    (1987)
  • G.L Rowland et al.

    Personality and Individual Differences

    (1986)
  • L.H Strickland et al.

    Journal of Experimental Social Psychology

    (1966)
  • S Svebak et al.

    Personality and Individual Differences

    (1989)
  • G.J.S Wilde

    Accident Analysis and Prevention

    (1986)
  • Z Zaleski

    Personality and Individual Differences

    (1984)
  • M Zuckerman et al.

    Personality and Individual Differences

    (1980)
  • G Anderson et al.

    Journal of Gambling Behavior

    (1987)
  • Apter, M. J. (1982). The experience of motivation. London: Academic...
  • Apter, M. J. (1992). The dangerous edge. New York: Free...
  • I.L Ball et al.

    British Journal of Psychology

    (1984)
  • Björg-Akesson, E. (1990) Measuring sensation seeking. Göteborg Studies in Educational Sciences, 75. Göteborg, Sweden:...
  • A.J Bowers

    British Journal of Clinical Psychology

    (1985)
  • L.T Brown et al.

    Journal of Consulting and Clinical Psychology

    (1974)
  • J.M Burger et al.

    Motivation and Emotion

    (1979)
  • J.M Burger et al.

    Motivation and Emotion

    (1982)
  • J.M Burger et al.

    Personality and Social Psychology Bulletin

    (1985)
  • B.R Burkhart et al.

    Journal of Consulting and Clinical Psychology

    (1978)
  • Campbell, J. B., and Palus, J. P. (1987). Intercorrelations and correlates of impulsivity measures. Paper presented at...
  • Cattell, R. B. (1957). Personality, and motivation structure and measurement. New York: Harcourt, Brace and...
  • Cowles, M., and Davis, C. (1985). Strength of the nervous system and reversal theory. In M. J. Apter, D. Fontana and S....
  • O Doherty et al.

    Personality and Individual Differences

    (1987)
  • D.M Donovan et al.

    Journal of Studies on Alcohol

    (1985)
  • Eysenck, H. J. (1967). The biological basis of personality. Springfield, IL:...
  • Eysenck, H. J., and Eysenck, S. B. G. (1976). Psychoticism as a dimension of personality. London: Hodder and...
  • S Eysenck et al.

    British Journal of Psychology

    (1978)
  • C.J Fowler et al.

    Psychiatric Research

    (1980)
  • M.G.I Freixanet

    Personality and Individual Differences

    (1991)
  • D Fulker et al.

    Journal of Research in Personality

    (1980)
  • Girodo, M. (1985, June). Telic and paratelic modes in operational undercover and field narcotics agents. Paper...
  • D Goldman et al.

    Journal of Personality and Social Psychology

    (1983)
  • Cited by (0)

    View full text