ALTERNATIVES TO THE A-B PERSONALITY CONCEPT IN PREDICTING CORONARY HEART DISEASE
J. J. Ray
University of N. S. W., Australia
Abstract
"A-type" personality is said to lead to coronary heart disease (CHD) yet even the authors of the concept admit that "A type" comprises several independent components. Generally, these components seem to be well-known psychological constructs (such as achievement motivation and aggression) but a fairly new construct which might be called "Freneticism" is also included. Freneticism (as measured by factors "S" of the Jenkins Activity Survey (JAS), does not however in fact seem to predict CHD. A reanalysis of the data from an Australian study by Ray & Simons was therefore carried out to see if something could be saved from the one genuine conceptual innovation falling under the "A-B" umbrella. After appropriate partial correlations were carried out to allow for correlations among the predictor variables, it was found that a new measure of freneticism independent of "A-B" (the Ray & Bozek scale) did provide a significant prediction of CHD. "A-B" score (derived from the JAS), however, was found to be a significant predictor of CHD only insofar as it measured an element of authoritarianism. There are thus two quite independent predictors of CHD -- authoritarianism and Freneticism. The A-B concept and the scale used to measure it (the JAS), however, have no useful role in either of them.
In 1979 Jenkins, Zyzanski & Rosenman (1979) published their manual for the ` "Jenkins Activity Survey" (or JAS) as the culmination of a research program that stretched back to at least 2O years earlier. The JAS was intended primarily as a measure of "A-B" -- a personality type (or, as they call it, a "behaviour pattern'') that has been claimed to be particularly common among sufferers from coronary heart disease (CHD). Since this commercial publication of the JAS in 1979, interest in the "A-B" concept appears to have become widespread among psychologists -- even among social psychologists (e, g. Mathews, 1982 and Strube, Turner, Patrick & Perillo, 1983). The interest of the concept now seems, in other words, to extend well beyond its original cardiological applications. Close attention to what the concept of "A-B" means and whether the JAS measures it is then well-warranted.
The term "A-B" seems to be used by its authors in two different ways (See Ray, 1984b). It can mean either whatever personality predisposes to CHD or a particular set of personality traits. In the first usage the theory "A-type personality leads to CHD" is no theory at all but merely a proposal for a definition. Usually, however, the term is used for a set of specifiable traits including achievement motivation, aggressiveness and hyperactivity. The theory "A-type personality leads to CHD" thus could be expanded to: "Aggressiveness, achievement motivation, hyperactivity (etc) lead to CHD''. Talk about "A-B" is thus in a sense superfluous. Should we not be examining whether individual traits within the set of A-B lead to CHD? Could It not be that measures that mix up achievement motivation, aggression and hyperactivity all in the one index might be providing a prediction of CHD solely because of one component in that index? Might it not be that aggression alone could account for the relationships observed ? A recent literature review (Diamond, 1981) suggests that it might.
If, then, aggression/hostility could alone account for much of the relationship between CHD and personality, where lies the usefulness of the "A-B" concept? Aggression and hostility are familiar concepts to psychologists and well-known measures of them have long been available. Why do we need either the concept of "A-B" or the inventory (the JAS) needed to measure it? One possible answer is that the definition of "A-B" includes one concept that is in fact relatively new to users of personality tests -- a concept that has to do with the way time is handled. Note the following descriptions of the "A-type" from Jenkins, Rosenman Et Zyzanski (1974) : "Time-urgency, acceleration of common activities, restlessness, feelings of struggle against the limitations of time". The Jenkins group do of course avoid providing a name for this component of "A-B" but for ease of reference we might perhaps call this component of the "A-type" personality "Freneticism". This then is one of the traits that is said to lead to CHD. This is obviously an hypothesis well worth examining. Unfortunately, it cannot be tested simply by using "A-8" scores from the JAS. "A-B" scores might predict CHD because of the other elements (such as aggression) that they contain. To test the freneticism hypothesis we need a separate freneticism measure. Such a measure is available. Jenkins, Zyzanski & Rosenman (1979) obtain from the JAS a factor-score labelled "S" (for "speed and Impatience") that should be adequate for the purpose. Unfortunately, in the studies they themselves list on p. 14 of their manual, factor "S" does not seem to predict CHD! The most original aspect of their theory appears therefore to be false.
Clearly, then, the days have passed when a single "A-B" score can be regarded as informative. What it measures is too multi-dimensional and some at least of the things it measures seem to be quite unrelated to CHD. Obviously, then, any relationship between "A-B" and other variables would be of uncertain implication. What is needed are studies which take the various proposed "components" of "A-B" separately and use different instruments to measure them. Two studies which did this were by Ray & Bozek (1980) and by Ray & Simons (1982). The first showed that what the JAS measured could be reduced to the familiar psychological constructs of dominance and achievement motivation plus a minor element of "Freneticism". The second study showed that of the three only aggressive dominance or "authoritarianism" (measured by the Ray (1976) "Directiveness" scale) predicted CHD. Thus most of what the Jenkins group have been measuring was found to be irrelevant to CHD. The finding with the "Directiveness" scale, it should be noted, does confirm Diamond's (1982) hypothesis that aggression is the central personality element in CHD prediction. As Heaven (1986) has shown, scales of dominance generally seem to measure a strong component of aggression as well as dominance per se.
A surprising and previously little-remarked feature of the Ray & Simons (1982) data, however, is that "A-B" was shown to be negligibly related to scores on the Ray & Bozek (1980) Freneticism scale. This suggests that either the JAS or the Ray & Bozek scale may be lacking in validity as measures of freneticism. Obviously, both cannot be measuring the same thing. See Ray (1984a). If we assume that it is the JAS that is providing a poor measure of "A-B", we are left with the interesting possibility that the theory put forward by the Jenkins group (concerning the involvement of freneticism in CHD) may simply not so far have been adequately tested. The consistent failures to confirm the theory may have been due to an inadequate measuring instrument. Clearly, alternative measures of freneticism should be tested tor their relationship with CHD.
The Jenkins group would no doubt wish to say again that the relationships with "A-B" reported in Ray & Simons (1982) and Ray (1984a) cannot be accepted because the selection of JAS items used by Ray & Simons (1982) to measure "A-B" is not exactly coterminous with any of the various selections of items that have been used by the Jenkins group from time to time. Ray (1984b) has however shown that the selection of items used in Ray & Simons (082) in fact correlates to the point of virtual identity with the selection currently favoured by Jenkins, Zyzanski & Rosenman (1979). This objection cannot therefore be used to deflect us from looking at the potential use of alternative measures of freneticism in CHD prediction.
METHOD
Following in the tradition of the many reanalyses of the "Framingham" and other data-bodies on CHD incidence, the simplest method of testing the possibility mentioned above seemed to be to carry out further analyses of the data described in Ray & Simons (1982) and Ray (1984a). These papers report a comparison of CHD sufferers and controls on a variety of personality tests and the Ray & Bozek (1980) Freneticism scale is one of those tests. The raw correlations between CHD incidence and personality variables reported in Ray (1984a) do not however tell us much by themselves because many of the predictor variables for CHD are themselves correlated. This must be allowed for in some way if fully informed conclusions are to be drawn.
The obvious confounding variable is age. Age is of course a strong independent predictor of CHD and also has effects on personality. It makes you slow down and thus get lower scores on measures of freneticism and "A-B". When therefore the influence of age is removed by partial correlation, the correlation of the JAS "A-B" score with CHD rises to .148. The correlation between CHD and the Ray & Bozek (1980) Freneticisrn scale also rises -- to .174. The Directiveness scale is not significantly correlated with age so there is no point in partial correlations for that variable. Since the level of the correlation required for significance (p <. 05) with the given N (313) is .113, it may be seen that the new analyses reported so far show in fact three scales which significantly predict CHD -- the JAS, the Ray & Bozek Freneticism scale and the Ray (1976) Directiveness scale. The claim by the Jenkins group that the JAS "A-B" score predicts "CHD" does therefore stand replicated and two new scales have been added to the predictor set. As the "A-B" score was however highly correlated with Directiveness and as the Directiveness scale is the better predictor of CHD (p < .05), it is appropriate to partial out Directiveness from the correlation between CHD and "A-B". When this is done; the correlation drops again to non-significance -- .076. Partialling out Directiveness from the correlation between CHD and Freneticism, however, leaves the correlation still significant -- at .147. Clearly, then it is the aggressive dominance component (the Directiveness component) of "A-B" which is providing the prediction of CHD. The Freneticism scale, however, is a separate predictor in its own right.
DISCUSSION
The new analyses of the Ray & Simons (1982) data have confirmed the disutility of using a single "A-B" score to predict CHD or, presumably, anything else. When separate scales are used to measure the main "components" of "A-B", it is found that there is not one personality type leading to CHD but rather two quite independent personality types -- freneticism and authoritarianism (the latter being here defined and measured as aggressive dominance). Nor can further use of the JAS be justified. The measures it provides (of "A-B" and "S") appear related to authoritarianism and freneticism but in neither case do they predict CHD as well. Even insofar as "A-B" does predict CHD, it is here shown that it does so only insofar as "A-B" contains an element of authoritarianism. "A-B" and the JAS fail, then, not only on grounds of conceptual clarity but also as empirical CHD predictors.
Paradoxically, rejection of the JAS as an adequate measure of coronary-prone personality leads to resurrection of one of the more dubious aspects of the Jenkins group theory -- the relationship between freneticism (their factor "S") and CHD. When a better measure of freneticism than the JAS can provide is used, support is found for the theory. Their theory is found to be better than their measuring instrument.
Perhaps the most revealing aspect of the Ray & Simons data, however, is traceable to the way relationships in it were analysed. The use of correlation coefficients reveals that even the highly significant relationships were of very low absolute magnitude. There are personality correlates of CHD but personality as a whole is a factor of negligible importance in CHD prediction.
REFERENCES
Diamond, E. L. (1982) The role of anger and hostility in essential hypertension and coronary heart disease. Psychological Bulletin 92, 410-433.
Heaven, P. C. L. (1986) Directiveness and dominance. Journal of Social Psychology 126 (2), 271-272.
Jenkins, C. D., Rosenman, R. H. & Zyzanski, S. J. (1974) Prediction of clinical coronary heart disease by a test for the coronary-prone behavior pattern. New England Journal of Medicine 290 (23), 1271-1275.
Jenkins, C. D., Zyzanski, S. J. & Rosenman, R. H. (1979) Jenkins activity survey form C. N.Y.: Psychological Corporation.
Matthews, K.A. (1982) Psychological perspectives on the Type A behavior pattern. Psychological Bulletin, 91, 293-323
Ray, J.J. (1976) Do authoritarians hold authoritarian attitudes? Human Relations, 29, 307-325.
Ray, J.J. (1984a) Authoritarianism, A-B personality and coronary heart disease: A correction. British Journal of Medical Psychology 57, 386.
Ray, J.J. (1984) Confusions in defining A-B personality type: A rejoinder to Jenkins & Zyzanski. British Journal of Medical Psychology 57, 385
Ray, J.J. & Bozek, R.S. (1980) Dissecting the A-B personality type. British Journal of Medical Psychology 53, 181-186.
Ray, J.J. & Simons, L. (1982) Is authoritarianism the main element of the coronary-prone personality? British J. Medical Psychology
55, 215-218.
Strube, M, J., Turner, C.W., Patrick, S. & Perillo, R. (1983) Type A and type B attentional responses to aesthetic stimuli: effects on mood and performance. Journal of Personality & Social Psychology 45, 1369-1379.
FINIS
No comments:
Post a Comment