Extraversion and performance: a response to Eysenck.

Eysenck (1974) questioned the adequacy of our sample for a meaningful test of the relation between extraversion and performance. We do not agree that the correlation between the Neuroticism and Extraversion scores of the Maudsley Personality Inventory (MPI) is "greatly in excess of that which is usually found [p. 308]" or that our finding of a negative correlation between impulsivity and neuroticism is atypical. Table 1 gives the results of correlational analyses of MPI scales for six additional samples from the University of Texas. All but one (Sample 5) of the samples in Table 1 showed significant (p < .05) negative correlations between Neuroticism and Extraversion. None of these correlations was significantly different from the Neuroticism X Extraversion correlation for the subjects used in the Cohen and Horn (1974) experiment. In addition, for the two other studies in which impulsivity and sociability were scored, significant negative Impulsivity X Neuroticism correlations were obtained. Neither of these two correlations was significantly different from that found in Cohen and Horn (1974). The accusation of having randomly chosen parametric values for the performance tests is not only false but begs an important issue. As Eysenck (1974) pointed out, while some authors find support for Eysenck's theories concerning extraversion and performance, numerous investigators have failed to obtain results in line with the theory. The "bewildering variety of results reported in the older literature" (we would add the current literature as well) not only suggests the importance of appropriate parameter values but raises the question of whether the results are simply random fluctuations around a correla-