Are there any predictive factors in the perception of the risks from BSE and Salmonella?


The threat from diseases like BSE or Salmonella is unquantifiable: people have to make a risk judgment on incomplete information. Cognitive theories about anxiety predict that more anxious individuals will perceive the risk to be greater (Butler & Mathews, 1983; Mathews, 1990). A number of experimental paradigms have demonstrated such effects in clinically anxious individuals versus 'normals' (e.g., Eysenck, M., Mogg, May, Richards & Mathews, 1991; Butler & Mathews, 1983). Reports of such effects in non-clinical populations are less common (e.g., Eysenck, M., MacLeod & Mathews, 1987). There is also evidence to suggest that such effects occur in cognitions about food as well. Pliner & Hobden (1992), for instance, found a relationship between anxiety and food neophobia.

Two studies (Mogg, Glenwright & Kentish, 1990; Mogg, Potts, Miller & Bradley, 1991), however, suggest that such findings, at least in non-clinical populations, may be unsafe. They suggest that Social Desirability (MCSD; Crowne & Marlowe, 1964) may be underlying the effect. Social Desirability is the tendency to behave in a socially desired manner. It is thus also predictive of a subject's tendency to respond to response biases in an experiment. Further, individuals scoring high on MCSD tend to produce low scores on anxiety measures, not necessarily because they are unanxious but because anxiety symptoms are seen as socially undesirable.

Therefore, it was predicted that high anxiety individuals would be more rejecting of "BSE" and "Salmonella", though MCSD may mediate such an effect.


As for Part 1 .


The expected pattern of correlations was found between the psychometric measures. Trait STAI and MCSD were significantly correlated (r(42) = -0.366, p = 0.017), though state STAI and MCSD were not (r(42) = -0.184, p = 0.244). There were no significant differences for sex on STAI (trait or state) or MCSD.

The mean (standard deviation) of trait STAI scores was 40.36 (9.23); for state STAI scores, it was 29.07 (6.64). The trait STAI scores match published norms (Spielberger et al., 1983), however, the state scores were surprisingly low. (Compare Spielberger et al. (1983) for working adults aged 19-39: for men, 36.54 ± 10.22; for women, 36.17 ± 10.96.)

The population was median split on MCSD, EPQ "E", FNS, age and trait and state STAI. As above, repeated measures MANOVAs were performed, but with the addition of between-subject factors for each psychometric/demographic variable. The food within-subject factor was "BSE" versus "Salmonella". All 13 questions in the FRI factor were used. All 13 questions for both foods were analyzed in a single MANOVA to reduce the chance of type I errors.

There were significant main effects of FRI and significant food × FRI interactions throughout, but the between-subjects factors gave no significant main effects or interactions (ps > 0.05; n = 39).

For sex, there was a trend main effect (F(37, 1) = 3.75, p = 0.061). Post hoc t-tests were performed on all 13 questions for both foods. Only one of these was significant at a 5% level (question 3 for "Salmonella"), but women scored higher than men for all of the questions.

There were also no significant main effects or interactions of whether the subject had children or not, nor for smoking status.

As vegetarians will not eat steak anyway, a MANOVA was performed on just the answers to "Salmonella" with a between-subjects factor of vegetarian or not. (Vegetarian was defined loosely to be anyone who would not eat at least red meat.) There was no significant main effect for or interactions with this vegetarian factor (ps > 0.1).

Further analyses were performed with both anxiety and MCSD as between-subjects factors. These produced no significant effects (ps > 0.05).


The weak results for sex appear to be a general effect for all rejected foods in the study, not specific to "BSE" and "Salmonella", however I have yet to do full analyses for all the other foods.

This study provides no evidence for an anxiety/perceived threat relationship in cognitions about BSE and Salmonella. This could be because such relationships are inconsistent in non-clinical populations (Mogg et al., 1990, 1991), that general psychometric measures are unhelpful when trying to predict behaviours relating to specific, single foods or simply that this particular study had insufficient power.


Return to abstract/contents