Login if you have a Gold, Gold plus or Platinum package; individual journal subscriptions; or an APA username and password. User ID or e-mail: Password:.
Journal of Consulting and Clinical Psychology IWi, Vol. 44, No. 4, 580 585
Development and Validation of the Health Loeus of Control (HLC) Scale Barbara Strudler Wallston George Peabody College Kenneth A. Wallston, Gordon D. Kaplan, and Shirley A. Maidcs Vanderhilt University The Health Locus of Control (HLC) Scale is an area-specific measure of expectancies regarding locus of control developed for prediction of health-related behavior. Two experiments show discriminant validity of the HLC in contrast with Rotter's Internal-External Locus of Control Scale. In Study 1, HLC internals who value health highly sought more information than other subjects. In Study 2, subjects in weight reduction programs consistent with their locus of control beliefs (as assessed by the HLC scale) were more satisfied with the programs than were "mismatched" subjects. Normative data on the HLC arc provided.
In a recent discussion of the directions and implications of locus of control research (Rotter, 1966), Strickland (Note 1) identified the relationship between a belief in internal control and physical health or well-being as an important emergent area. Although she cited 11 studies in this area in which "positive" relationships had been found, she neglected to mention reported instances of "negative" findings (e.g., Marston, 1970; O'Bryan, 1972). In addition, it should be recognized that, as with most research areas, results that do not confirm hypotheses tend not to be reported in the literature. One possible explanation for contradictory findings may be found in social learning theory, from which the construct of locus of control stems (Rotter, 1954). According to this theory, it is assumed that increasing an individual's experience in a given situation will lead to the development of specific expectancies. These expectancies subsequently play a greater role in determining one's future behavior in that situation than more generalized expectancies. It stands to reason that research This research was supported in part by Research Grant NU00426 from the Division of Nursing, National Institutes of Health. We gratefully acknowledge Ann Cowles' assistance during the scale development and validation and Stephen I. Abramowitz's helpful comments on a draft of this article. Requests for reprints should be sent to Kenneth A. Wallslon, School of Nursing, Vanderbilt University, Nashville, Tennessee 37240.
whose aim is the prediction of behavior in specific situations could profit from the use of more specific expectancy measures. In a recent article, Rotter (1975) recognized the value of such measures "if one's interest is in a limited area and particularly if one is seeking some practical application where every increment in prediction is important" (P. 59). Wallston and Wallston (Note 2) have discussed the difficulty of predicting behavior in a specific area such as health when using measures of generalized expectancies such as Rotter's (1966) Internal-External Locus of Control (I-E) Scale. The present research was based on the assumption that a healthrelated locus of control scale would provide more sensitive predictions of the relationship between internality and health behaviors. This article (a) describes the development of one such instrument, the Health Locus of Control (HLC) Scale, and (b) demonstrates the differential functional utility of this new measure over the traditional, more generalized T-E scale. Data obtained on the HLC with several populations are presented along with two studies in which subjects were independently classified as internals or externals on both the HLC and I-E scales. SCALE DEVELOPMENT Using a 6-point, Likert-type format, an item pool consisting of 34 items written as face-valid measures of generalized expectan-
580
HEALTH LOCUS OF CONTROL (HLC) SCALE TABLE i HEALTH Locus OF CONTKOL SCALE ITEMS Item 1. If I take care of myself, I can avoid illness. 2. Whenever I get sick it is because of something I've done or not done. 3. Good health is largely a matter of good fortune. 4. No matter what I do, if I am going to get sick I will get. sick. 5. Most people do not realize the extent to which their illnesses are controlled by accidental happenings. 6. 1 can only do what my doctor tells me t.o do. 7. There are so many strange diseases around that you can never know how or when you might pick one up. 8. When I feel ill, I know it is because I have not been getting the proper exercise or eating right. 9. People who never get sick are just plain lucky. 10. People's ill health results from their own carelessness. 11. I am directly responsible for my health.
Direction" I 1
K E
E E
r K I
r
" I = internally worded. K = externally worded. The scale i,s scored in the external direction, with each item scored from 1 (strongly disagree) to 6 (strongly agree) for the externally worded items and reverse scoied for the intctnally wotded items.
cies regarding locus of control related to health was administered to 98 college students in a small southern university. All subjects received psychology credit for their participation. Subjects also completed Rotter's I-E scale (Rotter, 1966), the Marlowe-Crowne Social Desirability Scale (Crowne & Marlowe, 1964), and they provided demographic data. An item analysis was run, and items were selected using the following criteria: (a) item mean close to 3.S, the midpoint; (b) wide distribution of response alternatives on the item; (c) significant item-to-scale correlation (r > .20); and (d) low correlation with the Marlowe-Crowne Social Desirability Scale. An attempt was made to maintain the balance between items worded in the internal and external direction. From the original pool, 11 items were chosen for the final scale (see Table 1). The 11-item devised scale has a potential range of 11 to 66. For the original sample,
581
the mean was 35.57, and the standard deviation was 6.22. Alpha reliability of the 11 items chosen by the above criteria was .72. In addition, the HLC does not reflect a social desirability bias, as evidenced by a —.01 correlation with the Marlowe-Crowne Social Desirability Scale. Concurrent validity of the HLC is evidenced by a .33 correlation (p < .01) with Rotter's 1-E scale for the original sample. The new scale, therefore, shares 10% common variance with the more established measure of locus of control. The overlap with the I-E scale was kept purposely low to enhance its discriminant validity, thus meeting the requirement that a new test not correlate too highly with measures from which it is supposed to differ (Campbell & Fiske, 1959). Additional Scale Information Table 2 presents normative data on the 11item HLC scale, which has been administered to a variety of subjects. Consistent with other findings on Rotter's I-E scale (Lefcourt, 1966), a group of older, primarily black, hypertensive outpatients were more external than college student samples or more middleaged, primarily white subjects recruited from the community. Within each of our samples, there have been no significant differences in HLC scores between males and females. Item analyses of the HLC for three subsequent college samples (with approximately 100 subjects in each sample) have resulted in alpha reliabilities of .40, .50, and .54. The alpha reliability of the community sample of equivalent size was .54. In the two studies described below, subjects' beliefs in locus of control were assessed TABLE 2 NORMATIVE HLC DATA HLC
Sample
N
College students 1X5 College students
