Vocational Interest Testing
Vocational Interest Testing
Many a man finds himself at middle age fully committed to continuing until retirement in an occupation that has lost its appeal for him. His dissatisfaction is not eased when he observes the large number of persons who are delighted with their choice of occupation and who are willing to spend extra hours and excess energy in work. Counselors of young persons have sought to develop ways to increase the proportion of persons whose interests match the activities involved in their daily work. The development of measures of vocational interest has aimed at assessing the inclinations of the young person so that he may be assisted in the choice of an occupation that will sustain his interests, be personally satisfying, and keep him usefully employed throughout his working life.
The most successful efforts at such assessment have involved the use of interest and/or biographical inventories. These differ only in the type of information that is obtained as a basis for measurement. The essential feature of the best of such interest measures is the use of large numbers of successful workers in each of many occupations for purposes of comparison. The young person’s responses to these inventories are scored in such a way as to inform him of the degree to which his interests and preferences match those of persons in each of a number of occupations.
A person who completes a vocational-interest inventory expresses preferences about items concerning a field of work or recreation or about items dealing with values, needs, or personality characteristics. For example, he may be asked if he likes a specific activity, is indifferent to it, or dislikes it; or he may be asked to select out of three or four activities the one he likes most or likes least. Inventories of this sort are not measures of aptitude or ability.
Since scores can be easily influenced by the desire of a person to make a particular impression, the preferred use of the interest inventory is in a setting where the individual considers it to his advantage to obtain the most accurate description of himself.
A biographical inventory contains items related to the past activities of the individual. He may be asked to indicate whether or not he has had certain experiences or has participated in certain activities. His enumeration of these past activities reflects fundamental preferences for activities; thus the reported record will provide evidence that will be relevant to vocational choice. To enable vocational counselors to use the data obtained from these inventories, scoring keys have been devised that are based on a comparison of the responses of persons employed in a given occupation with the responses of the general population.
Background and history . When asked what field of work he prefers to enter, a young person’s response incorporates not only his preferences for job activities but also his understandings or misapprehensions about the nature of the employment setting, the likely economic rewards he associates with a given occupation, and the factors relating to the social prestige of various occupations. His thinking may also be colored by his perception of the likelihood of opportunities for employment in one field or another, his estimate of the educational or training requirements for a given position, and his estimate of the personal characteristics that the job requires in comparison to his estimate of his own characteristics. The development of measures of vocational interests has resulted from the need to obtain more systematic and more veridical descriptions of occupational preference than individuals can usually provide by self-report.
The earliest major work in the development of vocational-interest inventories of the sort now commonly in use was that of Strong. He developed the Strong Vocational Interest Blank (SVIB), primarily for use with college students, and prepared a number of scoring keys for those occupations into which college students normally go (Strong 1943). These keys were based on the administration of the SVIB to a large number of employed workers in a variety of occupations, including medicine, dentistry, engineering, chemistry, physics, mathematics, highschool teaching, production management, life-insurance sales, journalism, and advertising.
Other investigators have assisted in the development of other scoring keys for the SVIB or have developed additional inventories for this purpose. The most widely used of the other instruments in the United States is the Kuder Preference Record (e.g., Kuder 1959), which can be scored for a very large number of occupations (for example, electrical engineer, farmer, newspaper editor, accountant, druggist) or for a small number of interest areas (for example, computational, persuasive, mechanical). The Minnesota Vocational Interest Inventory (MVII) is another such device, but one that aims to assist the young person primarily interested in entering the skilled trades or other occupations normally engaged in by noncollege youth (e.g., Clark & Campbell 1965).
Scoring key development . The men-in-general reference group for the SVIB represents all persons employed in the occupations normally entered by college graduates. The percentage of persons in both the men-in-general group and a given occupational group who respond to each item’s alternative responses is determined. Whenever men in a given occupation (the “criterion group") respond much more frequently to a given alternative than do men in general (the “reference group"), that response is given a positive scoring weight. When the response by the criterion group is much less frequent than for the reference group, that response is given a negative scoring weight. The summation of the scoring weights obtained by an individual who has completed the inventory is his score on that particular key. Persons who complete the inventory, therefore, are able to determine the degree to which their interests are like those of persons in a variety of occupations by having their inventories scored with the keys of a large number of groups. Obviously such scoring is laborious. Normally it is done by machine at centers established in various parts of the United States.
There is thus available to an individual who completes an interest inventory a substantial amount of information beyond that which he previously had. He can be led to think about the world of work in terms of the presumed activities of a variety of specific occupations, since these constitute many of the items to which he has responded, and he can learn the degree to which his responses are like those of persons in various occupational groups. In addition, he can learn something about the degree of disparity between his interests and those of certain groups, for the scores on the interest inventory not only identify the occupations that attract persons like himself but also those occupations that attract persons markedly different from himself.
The task of developing scoring keys of the sort just described is enormous. The cooperation of large numbers of employed persons is required. The development of appropriate items that will differentiate such occupations requires a careful prior examination of the characteristics of the world of work.
One might ask whether the same results might not be achieved by having persons well acquainted with the world of work make their own assessment of the way in which interest inventories ought to be scored. Such efforts have been made. Inventories developed with scoring keys based on the obvious content of items and employing the judgments of experts for scoring are enormously inferior to those having scoring keys developed by the more systematic approach.
An alternative, however, is to develop a key that describes the essential dimensions of vocational interests within a given individual. A measure of the degree of interest in mechanical things, for example, can be obtained by examining the way in which responses of an individual cluster into patterns. This sort of analysis of inventory responses has been attempted and has resulted in the use of scoring keys that are not occupational keys but keys measuring areas of interest. Inventories using such scores have the advantages of requiring a smaller number of keys and of being easier to score. They are harder to interpret, for they do not provide extensive information about the nature of interests of employed persons. However, the individual does have the opportunity to see himself described according to meaningful measures in comparison with other persons of his own age or status, so that he may see whether he exceeds them in terms of interest in various areas. If he then also has the opportunity to compare his scores with the average scores of persons employed in a wide variety of occupations, he may make an assessment of his likelihood of success or his likelihood of happiness in a given field.
Reliability and validity . Scores on interest inventories can be used for counseling only if we can be sure that they are stable over time and are related to other variables associated with occupational choice. The reliability of scores of vocational-interest measures is high, whether estimated by internal consistency or by test-retest. When inventories are administered with a short time lapse, the reported correlations between scores usually range between .80 and .95. Strong administered a retest of his inventory to a sample of graduates of Stanford University who had taken it as seniors 18 years before. He reported (1955) correlations that are almost as high (an average of .69) as those that have been obtained between intelligence test scores over the same interval of time. Reliabilities estimated on the basis of item intercorrelations, or internal consistency, have a high degree of association with the number of items used in the scoring process. These estimates, however, are also uniformly high for most of the commonly used inventories and scoring keys [see Psychometrics].
There are a variety of ways of estimating the validity of interest measures. One obvious method is to examine the degree to which the distributions of scores made by workers employed in different fields compare with each other. A variety of investigations have demonstrated that the degree of separation between occupational groups that is possible using the scoring keys of the most commonly used instruments is very great. In a typical comparison of two obviously dissimilar occupations one will find less than 5 per cent of persons in one occupation exceeding the average score obtained by workers in the second occupation when the two groups are scored on the key for the second occupation.
It is characteristic of measures that tap intellectual functions that the average intercorrelation of scores on a test battery is moderately positive and that negative correlations between measures are rare. This is not so for vocational interests. Very large negative correlations are obtained between scoring keys for occupations that are obviously different. Accountants and artists, for example, obtain very low scores on each other’s keys (Strong 1943); the correlation of the two keys is a high negative one (—.74).
Vocational-interest measures routinely show correlations very near zero with measures of intellectual functions, even when the latter are aptitude measures presumably related to success in the relevant occupation. This lack of relationship makes vocational-interest measures particularly useful in prediction because they add unique variance. This lack of correlation, however, also requires that caution be exercised in the interpretation of interest measures, for they cannot be treated as reflections of unobserved capabilities of the individual, but rather must be treated as reflections of inclinations or preferences that have no systematic relationship to combinations of aptitudes.
Longitudinal studies. Another method for estimating the meaning of vocational interest is to inquire about the likelihood of a person’s persisting in an occupation when his interest score is high as against the likelihood of his persisting in the occupation when his score is low. The most spectacular findings in this regard have been obtained by Strong in his 18-year follow-up study. He found that the likelihood of persistence in an occupation was markedly enhanced when a person’s interests were consonant with those of workers in that occupation. A similar study by Campbell indicated that students whose measured interests were similar to those of life-insurance salesmen were much more likely to enter a field related to selling than were their fellow students; his study was a long-term follow-up of students in the state of Minnesota who had completed the SVIB as high school seniors (1966a).
The stability of interests . The use of the comparison of occupational groups in the development of scoring keys rests on the assumption that occupational interests do not change with age and that the interests of a particular occupational group do not change over time. Both of these assumptions require testing. The first requires the administration of inventories to individuals at varying ages in order to see the degree to which their scores change. A variety of studies of this sort have been completed but have not yet been published. They lead to the generalization that students of college age as well as those in the 11th and 12th grades of high school have interests that are sufficiently mature and that have stabilized enough to be generally useful for prediction of later scores. At the 9th and 10th grade levels some young persons have achieved a maturation of interests sufficient to enable prediction of their scores at later times, but others have not. Unfortunately there is no certain way to identify those persons for whom interest measurements are appropriate and those who are not adequately mature, although undoubtedly many signs of immaturity may be used. These studies have demonstrated that the interests of persons in general move from an emphasis on the natural sciences toward social service and the social sciences during the period from early high school to the late college years.
The second assumption deals with change in the occupational interests of a group over a period of time. Some occupations certainly change: the airline pilot of today is a different sort of person from the aviator of World War i, and the physician of today is surely not the same as the general practitioner of thirty years ago. To what extent do changes in the nature of an occupation determine who enters that occupation? What basis do we have for estimating that a person today should enter an occupation when data used are based on a preceding generation? The most interesting study that bears on this question has been done by Campbell (1966b). Using Strong’s SVIB data on employees of the Federal Reserve System in the state of Minnesota in the early 1930s and comparing these scores with those of incumbents 35 years later, Campbell found that the profile of interests of the current group, in spite of the fact that the banking profession has ostensibly changed greatly in its attitudes and practices, was nonetheless almost identical to that obtained on the prior sample. Thus there is convincing evidence for this occupation of an astonishing consistency in the sorts of persons who are attracted into key positions. Further studies of this sort with other changing occupations are obviously required before we may make any general statement about the effects of change in the world of work.
Use of vocational-interest measures . Vocational-interest measures are not developed primarily for the purpose of predicting achievement; indeed, they do not predict achievement particularly well. This may be owing in part to the fact that achievement is normally measured in a school setting, where ability is a much more important variable than interest or inclination. In settings where the training situation is such that inclination is much more important, there may be some use for these measures. Clark, in a study using the MVII in a U.S. Navy technical school, found that persons of marginal abilities with high interest scores would achieve very well, whereas persons of marginal abilities with low interest scores would achieve quite poorly (1961). Studies with the SVIB in attempts to predict achievement in medical schools have not been so successful, perhaps because they enroll few students with marginal ability. [See Achievement Testing.]
Interest measures are intended, however, to predict the degree to which a person would find various sorts of work satisfying. If he selects an occupation that matches his interest scores, the likelihood that he will remain in it should increase. Studies of occupational satisfaction, however, have not given a great deal of support to this position. This may be due to the fact that occupational satisfaction has a number of dimensions, some of which are specific to particular jobs and some of which are specific to the occupation itself. It may also be, however, that satisfaction itself is not readily predictable and that whatever factors produce high degrees of satisfaction or low degrees of satisfaction as reported by individuals are not factors relating to the pleasure that a person gets from the activities that are specific to the occupation. Recent studies of the different factors producing satisfaction and dissatisfaction in work settings may yield light on this matter.
As reported earlier, however, it is possible to predict whether a person will stay in an occupation. That a person’s occupational choice can be predicted by interest measures, at least to some extent, supports the point that occupational choice reflects important characteristics of the person and of his perception of himself. Yet many jobs are considered as forms of employment to which many unfortunate persons are relegated because they are incapable of making any other contribution to society. Rewards in such occupations are presumed to be solely in wages and in the end of the work, which comes each day and each week. To what extent are choices of such occupations related to individual characteristics, and to what extent are they forced upon an individual as a result of his lack of capability for anything better? This question is fundamental to a decision about the degree to which it is appropriate to use vocational-interest measures with persons below the professional and skilled-trades occupational groups.
Darley and Hagenah conclude that the unskilled occupations and the more routine occupations provide no basis for differentiation on the basis of vocational interests (1941). Clark reports, however, that men enlisted in the navy, where one would not expect much opportunity for individual preference to affect assignment, do sort themselves out into various navy rating groups in ways that are related to their measured interest patterns. This was found to be true even when these measured interest patterns were unknown either to the men involved or to the persons who were influencing their classification and assignment. Clark’s work with skilled tradesmen and with retail salesclerks, warehouse men, milk-wagon drivers, and the like also demonstrated that these groups are fully as easily differentiated from each other on the basis of their measured interest patterns as are professional workers (1961). Thus it would appear that for a substantial portion of the work force, choices of occupation are made in the light of preferences for activity and of individual characteristics that are differentially suited to given fields.
These findings, however, are not contradictory to the generalization stated by Darley and Hagenah. The average intelligence of many of the groups with which Clark worked was above that of the general population. There are many occupational groups normally not thought of as professional or highly technical that attract persons of intelligence and presumably provide sufficient gratification and stimulation to lead them to persist in the occupation. For such groups, interest measures can be and have been developed and used. As yet, no such measures have been developed for the less intellectually attractive occupations.
The use of vocational-interest measures in studying occupational choice and the world of work does permit some better understanding of how the world of work is sorted out in terms of individual characteristics. Earlier mention was made of the development of one type of scoring key by comparing occupational groups and of a different type of scoring key by studying the ways in which item responses of individuals correlate with each other. One would expect, if the world of work were divided in some rational manner related to the capabilities and motivations of persons, that these two sets of scores would have substantial correlations with each other. Helen Gee and W. T. Norman (see Clark 1961) developed a set of homogeneous keys for the MVII without reference to item content and found that these keys had highly orderly and meaningful relationships with empirical keys developed on the MVII by Clark. The fact that these relationships turned out to be highly orderly was directly a result of ignoring item content. When item content is used to determine the choice of items to be scored in homogeneous keys, this orderliness does not appear. This observation should make it clear that the field-of-interest measurement has not yet developed to a point where it is possible to identify and understand the major components of variance in interest scores and to relate these to our knowledge of the content of occupations and to other measures of psychological characteristics. Ultimately such orderly relationships should become known.
The nature of interest . What aspects of personality are being measured when interest inventories are used? Are these characteristics essentially inborn, or are they resultants of varieties of successful and unsuccessful experiences during youth? Information currently available is far from adequate but suggests that interests are acquired and that they have meaningful relations to other measures of the developing personality. Answers to these questions are important not only for vocational counselors but also for persons interested in all areas of psychological measurement and prediction. Measured interests, as psychometric variables, stand almost alone in the considerable amount of built-in validity that they possess. As reported earlier, criterion groups can be separated very sharply by interest measures. These differences are far greater than differences ordinarily obtained by using personality tests, intelligence tests, or tests of special aptitudes. The use, then, of interest measures in the broader areas of personality study surely is indicated and should add a great deal to understanding in that field.
A most promising area for study is related to factors producing these observed differences among persons. Roe’s work (1956) on the influence of developmental factors is important, as is the work of Holland (1962), Tyler (1964), and Super (1949; 1957). These investigators have provided conceptual frameworks that place interest measures in an orderly relation to life experience and to other facets of the developing personality. Thus, interest measures may some day permit a better review of earlier experiences and a better understanding of the effects of childhood and family experiences on satisfaction in work and on other aspects of adult life.
Kenneth E. Clark
[Directly related are the entriesAchievement testing; Aptitude testing; Individual differences; Intelligence and intelligence testing. Other relevant material may be found inCounseling psychology; Industrial relations; Occupations and careers; Personality measurement; Professions.]
Berdie, Ralph F. 1960 Strong Vocational Interest Blank Scores of High School Seniors and Their Later Occupational Entry. Journal of Applied Psychology 44: 161-165.
Campbell, David P. 1966a Occupations Ten Years Later of High School Seniors With High Scores on the SVIB Life Insurance Salesman Scale. Journal of Applied Psychology 50:369-372.
Campbell, David P. 1966b Stability of Interests Within an Occupation Over Thirty Years. Journal of Applied Psychology 50:51-56.
Clark, Kenneth E. 1961 The Vocational Interests of Nonprofessional Men. Minneapolis: Univ. of Minnesota Press.
Clark, Kenneth E.; and Campbell, D. P. 1965 Manual for the Minnesota Vocational Interest Inventory. New York: Psychological Corporation.
Darley, John G.; and Hagenah, Theda (1941) 1955 Vocational Interest Measurement. Minneapolis: Univ. of Minnesota Press. → First published as Clinical Aspects and Interpretation of the Strong Vocational Interest Blank.
Fryer, Douglas 1931 The Measurement of Interests in Relation to Human Adjustment. New York: Holt.
Holland, J. L. 1962 Some Explorations of Theory of Vocational Choice: I. One and Two Year Longitudinal Studies. Psychological Monographs 76, no. 26, serial no. 545.
Institute ON THE Strong Vocational Interest Blank, University of Minnesota, 1955 1960 The Strong Vocational Interest Blank: Research and Uses. Edited by Wilbur L. Layton. Minneapolis: Univ. of Minnesota Press.
Kuder, G. F. 1959 Manual for the Kuder Preference Record Occupation Form D. Chicago: Science Research Associates.
Roe, Anne 1956 The Psychology of Occupations. New York: Wiley.
Strong, Edward K. Jr. 1943 Vocational Interests of Men and Women. Stanford Univ. Press.
Strong, Edward K. Jr. 1955 Vocational Interests 18 Years After College. Minneapolis: Univ. of Minnesota Press.
Super, Donald E. 1949 Appraising Vocational Fitness by Means of Psychological Tests. New York: Harper.
Super, Donald E. 1957 The Psychology of Careers. New York: Harper.
Tyler, L. E. 1964 The Antecedents of Two Varieties of Vocational Interests. Genetic Psychology Monographs 70, no. 2:177-227.