METHOD

 

The study was based on two measuring devices. The first was a brief (15-item) biographical questionnaire, discussed In detail below, which asked respondents for basic demographic information, academic field of interest, grade-point average, etc., and contained five questions about the use of marijuana and other drugs. Four levels of marijuana use were defined with this questionnaire: (a) students who reported smoking marijuana fairly regularly (frequent users), (b) students who reported using fnarijuana 10 times or less (occasional users), (c) students who indicated they had not smoked marijuana (nonusers), and (d) students who said they had not and never would smoke marijuana (principled nonusers).

 

The second measuring device was the California Psychological Inventory (CPI; Gough, 1964). The CPI was scored for the standard 18 scales plus an additional scale developed by the first author (Hogan, 1969).

 

The Ss for the study included 76 male undergraduates at the Johns Hopkins University and 72 undergraduate men from Lehigh University. Forty-two of the Hopkins Ss were tested individually in the campus student union; the remainder were tested as a group in an introductory psychology class. The Hopkins sample contained 23 frequent users, 14 occasional users, 16 nonusers, and 23 principled nonusers. Direct evidence was available that 26 of the 37 professed users had actually smoked marijuana. The veracity of the self-reports of the remaining 11 users could not be determined.

 

The students from Lehigh University (a school comparable to Hopkins in terms of size, geographic location, and the academic interests of the student body) were drawn from a large introductory psychology course. There were 14 frequent users, 9 occasional users, 28 nonusers, and 21 principled nonusers in this group. Although the investigator had fine rapport with these students, the validity of their self-reports is only presumptive.

 

RESULTS

 

The biographical questionnaire results are presented in Table 1, where nonsignificant findings concerning place of birth, size of home town, and plans after graduation are omitted (most students came from moderately large to large cities on the Eastern Seaboard, and the four groups had quite similar career aspirations). The students differed with regard to 4 of the 10 questionnaire variables, the strongest difference appearing in fraternity membership. Table 1 shows that frequent users are more often fraternity members than are occasional users, nonusers, and principled nonusers. The data from Table 1 also suggest a modest trend for freshmen and sophomores to cluster in the principled nonuser category, for frequent and occasional users to major in the humanities and social sciences, and finally, for nonusers to attain slightly poorer grades than the other three groups.

The last finding from the biographical questionnaire concerns the degree to which students who try marijuana experiment with other drugs. The data for users from both schools combined (N = 60) show that 61.7% had also used hashish; 18.3%, opium; 30%, amphetamines; 25%, LSD; and 1.6%, or one case, had tried heroin. Thus, there seems to be some connection between smoking marijuana and using other drugs, although no causal relationships may be inferred. Furthermore, within this sample no evidence is found for the common belief that marijuana users frequently experiment with heroin. The degree to which this latter finding would hold for less socially advantaged populations is obviously a matter for further investigation.

 

The data obtained from the CPI provide evidence about the personological determinants of marijuana use. The Hopkins and the Lehigh samples were classified into four groups according to usage. Studying the samples separately, a one-way analysis of variance was conducted for each of the 19 CPI scales. Table 2 presents the F ratios and eta coefficients resulting from this analysis. An examination of the data in Table 2 suggests the presence of a moderately similat pattern of relationships for each sample, with slightly stronger results appearing for the Hopkins group. The two samples were then combined and a one-way analysis of variance was again conducted for each of the 19 CPI scales. Ten of the 19 scales differentiated between the four levels of use at or beyond the .05 level of statistical significance (Table 3). Users as a group scored highest on Capacity for Status, Social Presence, Achievement via Independence, Flexibility, and Empathy. Users scored lowest on Sociability, Responsibility, Socialization, Communality, and Achievement via Conformance. The pattern was exactly reversed for the two combined groups of nonusers.

 

Although it is gratifying to know that the four groups can be differentiated with regard to their CPI scores, it is also important to inquire about the strength of the observed relationships. Dividing the total sample into groups of 60 users and 88 nonusers, a discriminant function analysis was conducted using the 10 scales significant in the previous analysis of variance. The resulting discriminant function accurately classified 82% of both users and nonusers (generalized Mahalanobis D square = 78.9, df = 10, p < .01) (cf. Morrison, 1967; Rulon & Brooks, 1968), a hit rate which clearly exceeds a chance level of accuracy.

 

The hit rate obtained with 10 variables may be compared with the discriminating power of the best single variable (Flexibility) which correctly classified 63.8% of the sample. For the best two variables (Flexibility and Socialization), the hit rate was 67.1%; for the best three (Flexibility, Socialization, and Empathy), it was 71.9%. The best four variables (Flexibility, Socialization, Empathy, and Social Presence) correctly identified 70.8% of the sample, and finally, by adding Communality to the preceding four variables, an accuracy level of 73.0% was obtained.

 

 

Discussion

 

Some decrease in the hit rates reported above will obviously occur in other investigations; thus no claims can be made for the extent of the relationships observed in this first analysis. However, it is unlikely that future analyses will reveal any major change in the pattern of these associations, or in their essential components.

The 10 scales which were significant in the combined analysis of variance seem related in an important way to professed marijuana use; thus, it is proper to examine their social and interpersonal implications. Considerable information is available for the interpretation of these scales (Gough, 1964, 1968; Hogan, 1969). For discussion purposes, frequent users will be contrasted with principled nonusers. Occasional users and nonusers fall between these two groups on most measures. Frequent users are characterized by high scores on Capacity for Status, Social Presence, and Flexibility. Such persons tend to be self-confident, socially poised, skilled in interpersonal relations, and possess a wide range of interests. On the other hand, they also tend toward narcissism, self-aggrandizement, and overconcern with personal pleasure and diversion.

 

Frequent users received low scores on Socialization, Responsibility, Communality, and Achievement via Conformance. These scores suggest that frequent users will be hostile to rules and conventions, impulsive, somewhat irresponsible, and rather nonconforming. Finally, the users' high scores on Achievement via Independence and Empathy indicate that they have the sort of achievement motivation necessary for success in graduate school, and that they are socially perceptive and sensitive to the needs and feelings of others. Thus, undergraduate marijuana use is a complex syndrome in which social poise is offset by a somewhat assertive nonconformity, empathy by narcissism, and wide interests and achievement potential by impulsiveness and irresponsibility.

 

The profile for principled nonusers is similarly complex. The pattern of high scores on Responsibility, Socialization, Communality, and Achievement via Conformance indicates that these persons will be pleasant, responsible, considerate, and dutiful. On the other hand, they will tend to be rather conventional and lacking in spontaneity and verve. This trend is reinforced by low scores for Capacity for Status, Social Presence, Achievement via Independence, Flexibility, and Empathy, which further suggest that principled nonusers are perhaps too deferential to external authority, narrow in their interests, and overcontrolled. As in the case of frequent users, we see a complicated pattern, with responsibility and maturity balanced by authoritarian compliance, and devotion to duty balanced by a narrow interest span.

As is well known, Freud felt that a neurotic symptom could be understood only by examining the character structure of the person displaying the symptom; that is, the neurotic character rather than the symptom became the primary concern of psychoanalytic theory and therapy. Similarly, with regard to delinquent conduct, it is only in reference to the actor's character structure as a whole that single actions can be comprehended (cf. Fromm, 1947, p. 33). Thus, our understanding of undergraduate marijuana use would be increased if we could characterize the general moral posture of our Ss. There are two ways in which we can estimate this posture. The first method involves using a CPI-based "Social Maturity Index" (Gough, 1966; Gough, DeVos, & Mizushima, 1968) where social maturity is defined as follows:

 

Social Maturity, in its highest form involves the creation of new order, and hence the destruction of old order. The highly socialized individual can live by the rules, however oppressive. . . . The socially mature individual, on the other hand, although able to adapt to convention, is receptive to change and innovation, and under repressive conditions may set himself against the established order [Gough, 1966, p. 1901.

 

This social maturity index was originally defined by comparing the responses of 2,146 nondelinquents with those of 881 delinquents on the CPI and developing a six-variable regression equation to distinguish between the groups. In a cross-validating sample of 2,482 nondeliquents and 409 delinquents, the pointbiserial correlation between the index and the delinquency—nondelinquency criterion was .63. High scorers on the equation are described in peer ratings as rational, idealistic, wholesome, clear-thinking, and organized. Low scorers are seen as shallow, intolerant, nervous, temperamental, and frivolous. The constant and weights for the equation have been adjusted so that the mean score on the index in a normal population will be 50.0. In the original sample, the nondelinquent mean was 50.4, and the mean delinquent score was 42.7.

 

In the present study, the frequent users had a mean score on the social maturity index of 49.5, the occasional users' mean score was 50.5, the mean nonusers' score was 50.4, and for the principled nonusers, the mean score was 51.6 (F = 2.81, df = 3/144, p < .05). Thus, there is a dependable difference between the groups in terms of social maturity as measured by Gough's index, with principled nonusers receiving the highest scores. However, all groups are relatively mature when compared with the mean score of Gough's original delinquent sample.

 

The first author has outlined an alternative method by which the moral posture of these Ss can be estimated (Hogan, 1967). This method assumes that moral development proceeds along five important dimensions (moral knowledge, socialization, empathy, autonomy, and principles to justify moral decisions in conditions of normlessness). A first step toward assessing the character structure of an individual can be made after taking all of these dimensions into consideration.

 

It was not possible to obtain scores for our Ss on all five dimensions; however measures of two of the key variables (socialization and empathy) are found in the CPI. The four combinations of high and low scores on these scales form a simple but interesting typology. According to the model, persons who score high on both Socialization and Empathy would be regarded as morally mature, all other dimensions being equal. Persons high on Socialization but low on Empathy would be considered well-socialized rule followers, but somewhat deficient in charitable or benevolent tendencies. The Ss scoring low on Socialization but high on Empathy would be careless of rules and conventions, but disposed to adopt a "broad moral perspective," that is, to consider the implications of their actions for the welfare of others (Hogan, 1969). Finally, delinquents would score low on both scales. Compared with men in general (Table' 4), principled nonusers receive high Socialization and low Empathy scores. By the same standard, frequent users are low on Socialization and high on Empathy. Occasional users and simple nonusers score between the "extreme" groups. Thus, in terms of the definition outlined above, neither extreme group is morally mature. Rather frequent users and principled nonusers seem to represent two different moral postures, each with its own particular strengths and weaknesses. Occasional users and simple nonusers share these features to a lesser degree.

A comment might be in order concerning the incidence rate of marijuana use in our samples. At the outset of the study the authors tested users exclusively. Thus, the percentage of users in the total sample (40.5%) probably overestimates the correct base rate for these groups. For example, 31.9% of the Lehigh students, who represent an approximate random sample of undergraduates at that university, reported smoking marijuana. The 34 Hopkins students who were tested as a class were a cross-section of students at their university, and there were 11 users in the group (32.4%). Thus, the best estimate of the base rate of marijuana use at these two institutions is about 32%. It will be interesting to compare this percentage with usage data from other schools as the information becomes available.

 

CONCLUSIONS

 

The preceding discussion leads to four conclusions. First, professed marijuana use at two small eastern universities can be predicted with fair accuracy. However, the degree to which the methods and findings would apply to other groups with high marijuana usage (ghetto blacks, Mexican Americans in the Southwest) is obviously an open question. For these groups, the dynamics of the behavior apd the character structure of the participants will undoubtedly differ markedly from the present sample, thus different results should obtain. Second, users and nonusers are indistinguishable with regard to their secondary education, extracurricular activities, or athletic participation. They do differ, however, in terms of fraternity membership, and to a lesser degree in their academic major, year in school, and scholastic achievement. Third, marijuana users show a personality pattern which is somewhat at variance with many popular stereotypes. In comparison with nonusers, they are more socially skilled, have a broader range of interests, are more adventuresome, and more concerned with the feelings of others. Conversely, and in accordance with general opinion, they are also im pulsive and nonconforming. However, it is the complexity of this pattern we wish to emphasize. While users are in some ways antisocial, they are characterized by other rather valuable traits as well, that is, interpersonal sensitivity and intellectual curiosity. Finally, we would like to argue that, in the long run, the character structure of nonusers is not necessarily superior to that of users. This judgment is based on the fact that frequent users and principled nonusers receive rather similar scores when compared with delinquents on a well-validated index of social maturity, and both appear less than morally mature when assessed in terms of two scales specifically designed to predict moral behavior. In general terms, marijuana use is a solipsistic or self-regarding activity which is perhaps more properly classified as amoral than immoral, and current disapproval of its use may reflect a cultural emphasis rather than a truly "moral" judgment.

 

REFERENCES

 

FROMM, E. Man for himself. New York: Holt, Rinehart & Winston, 1947.

Govan, H. G. Manual for the California Psychological Inventory. (Rev. ed.) Palo Alto, Calif.: Consulting Psychologists Press, 1964.

Cocci, H. G. Appraisal of social maturity by means of the CPI. Journal of Abnormal Psychology, 1966, 71, 189-195.

GOUGH, H. G. An interpreter's syllabus for the California Psychological Inventory. In P. McReynolds (Ed.), Advances in psychological assessment. Vol 1. Palo Alto, Calif.: Science and Behavior Books, 1968.

GOUGH, H. G., DEVos, G., & Mrzusinms, K. Japanese validation of the CPI social maturity index. Psychological Reports, 1968, 22, 143-146.

HOGAN, R. Moral development: An assessment approach. Unpublished doctoral dissertation, University of California, Berkeley, 1967.

HOGAN, R. The development of an empathy scale. Journal of Consulting and Clinical Psychology, 1969, 33, 307-316.

Monism, D. F. Multivariate statistical methods. New York: McGraw-Hill, 1967.

RULON, P. J., & Bkooxs, W. D. On statistical tests of group differences. In D. K. Whitla (Ed.), Handbook of measurement and assessment in behavioral sciences. Reading, Mass.: Addison-Wesley, 1968.

WEE, A. T., ZINBERG, N. E., & NELSEN, J. M. Clinical and psychological effects of marijuana in man. Science, 1968, 162, 1234-1242.