JOURNAL OF PERSONALITY ASSESSMENT, 1990, 54(1 & 2), 191-203Copyright © 1990, Lawrence Erlbaum Associates, Inc.

A New Measure of Weight Locusof Control: The Dieting Beliefs Scale

Stephen Stotland and David C. ZuroffMcGill University, Montreal

This article describes the construction and preliminary validation of a new scale ofweight locus of control, the Dieting Beliefs Scale. The 16-item scale demonstratedmoderate internal consistency and high test-retest reliability in a sample ofundergraduate women. Principal-components analysis suggested three factors.The three factors were interpretable and had distinct relations with a variety ofweight-related and psychological variables. The results suggest that weight locus ofcontrol is a multidimensional construct, and they provide a possible explanationfor the inconsistent findings concerning the relation between weight locus ofcontrol and dieting success. Implications for the study of dieting relapse and for theconstruction of treatment programs are discussed.

Internal locus of control (Rotter, 1966) has been proposed by several researchersas a potential predictor of success in weight-loss programs. Unfortunately, theresearch findings offer a rather confusing picture of the relation, with someresults positive (Blach &. Ross, 1975; Ross, Kalucy, &. Morton, 1983) and othersnegative (Gormally, Rardin, & Black, 1980; Tobias & MacDonald, 1977).Several reasons may be offered for this state of affairs, including: (a) the generalneglect of important variables specified by social learning theory other thanlocus of control, (b) the use of general as opposed to weight-specific measures oflocus of control, (c) the use of locus of control measures of questionablereliability and validity. An additional explanation for the inconsistent results inthis area concerns the nature of weight locus of control. Previous research hastreated it as a unidimensional construct, defined as internal versus externalbeliefs about the control of weight. Research concerning both general (Reid &Ware, 1974) and health locus of control (K. A. Wallston, B. S. Wallston, &.DeVellis, 1978) has discovered a multidimensional structure, and it is possiblethat the same is true of weight locus of control.

Several researchers have reported a significant relation between locus of

192 STOTLAND AND ZUROFF

control and weight loss. Balcb and Ross (1975) evaluated tbe relation betweenRotter's Internal-External (I-E) Locus of Control Scale and outcome in 34female participants in a 9-week bebavioral weigbt loss program. Significantcorrelations were found between the Rotter I-E scale and botb programadherence and weight loss.

Ross et al. (1983) administered the Reid and Ware (1974) revision of the RotterI-E scale to 133 women undergoing jaw wiring for massive obesity. Ross et al.revised the scale to include several items specifically related to beliefs concerninglocus of control of weight loss. A factor analysis of the scale suggested athree-factor solution, composed of a large, generalized Locus of Control factor;a Self-Control factor; and a Social Systems Control factor. These three factorswere considered separately in the prediction of weight loss, weight maintenance,and treatment compliance.

Results showed that scores on the General Locus of Control factor predictedtreatment compliance and weight maintenance. No relation was found betweenthis factor and weight loss during treatment, which the authors suggested mayhave been due to the specific controls in the weight-loss phase imposed by jawwiring. Neither of tbe other two scales was useful in predicting any of theoutcome criteria. The five weight-specific items did not load on any one of thefactored scales, but were spread across all three of them. The lack of cohesion ofthese items would tend to limit their predictive validity. This study, however,suggested tbat a general measure of locus of control may be significantly relatedto treatment outcome.

In contrast to these findings, several authors have reported no relationbetween locus of control and weigbt change. B. S. Wallston, K. A. Wallston,Kaplan, and Maides (1976) assigned overweight women to either a self-directedor a group (externally directed) program. The conditions differed in the mannerof dissemination of information about weigbt control strategies and in theamount of contact with the experimenters. Both I-E locus of control andhealth locus of control (HLC) were assessed. A significant interaction betweenHLC and treatment condition was found for subjects' ratings of satisfaction withthe treatment program, such that satisfaction was higher wben program andsubject type were consistent. No such relation was found for I-E classification.No significant main effects or interactions were found, for either scale, wbenweight loss was considered. Anotber negative finding was observed by Gormallyet al. (1980). They measured I-E locus of control at pretreatment in 40 femaleparticipants in a 16-week bebavioral treatment program. No relation was foundbetween the I-E dimension and weight loss.

Tobias and MacDonald (1977) administered the I-E scale and a five-itemweight-specific locus of control scale to 96 undergraduate women participatingin one of five weight-loss treatment groups. Only the bibliotherapy and bebav-ioral contract groups achieved significant weight losses. No differences between

DIETING BELIEFS SCALE 193

groups in locus of control (either scale) were found at either pretreatment orposttreatment. Only the group in which treatment was designed specifically toinculcate beliefs that deficits in willpower and effort are primary causes of obesityshowed a shift to more internal weight locus of control beliefs; however, thisgroup was not successful in losing weight. The authors concluded that locus ofcontrol is not important in determining treatment success. However, twocriticisms of this conclusion can be made.

First, the unknown psychometric properties of their weight locus of controlsscale preclude definite conclusions about the construct itself. Furthermore,Tobias and MacDonald's analyses did not address the important issue of locus ofcontrol as a predictor variable. It is possible that a correlational analysis of therelation between locus of control and weight loss would have produced asignificant result. In other words, within-group differences in weight loss mighthave been associated with differences in locus of control.

A study by Saltzer (1982) is notable for being the first to report informationabout the reliability and convergent validity of a scale designed to measureweight locus of control. Saltzer examined the ability of a four-item Weight Locusof Control Scale (WLOC) to predict weight loss, relative to general measures oflocus of control (The Rotter I-E scale), and the Multidimensional Health Locusof Control Scale (MHLC; K. A. Wallston et al , 1978). The scale demonstratedmoderate test-retest reliability in an undergraduate sample. Internal consistency(Cronbach's alpha) appeared relatively low, however. Saltzer suggested that thismay have been attributable to the fact that the scale had only four items. Thescale was shown to correlate significantly, but only to a moderate degree, withgeneral locus of control scales, thus establishing some degree of convergentvalidity. Finally, the scale was found to be uncorrelated with a measure of socialdesirability.

Saltzer (1982) then assessed the relations among WLOC, MHLC, and weightloss in 115 female patients in a medical weight-reduction program. In addition tolocus of control, a value survey was included. It was hypothesized that internalswith a high value of health or physical appearance would be more successfulthan externals with similar values. Program success was defined in terms of anindividual's ability to achieve her weight-loss goal. The magnitude of thecorrelation between initial weight-loss goal and posttreatment weight wascompared across groups of subjects selected to be high on health or physicalappearance and either high or low on WLOC. The correlation was significantlyhigher in the group of internals with high value on health or physical appear-ance than in the group of externals with similar values. It was concluded thatgiven a high value of weight loss, WLOC is a significant predictor of success (i.e.,the ability to achieve one's weight-loss goal).

This study seems to provide some evidence of the value of a specific measureof weight locus of control. Criticisms can be made concerning the questionable

194 STOTLAND AND ZUROFF

reliability of the scale, and the unorthodox measure of outcome used in thesecond study. One wonders about the relation of WLOC and weight changesdefined in terms of percentage overweight, percentage body fat, or otheroutcome measures typically used in obesity treatment studies.

The research reviewed so far can be summarized as follows. First, the relationbetween generalized locus of control and weight loss is inconsistent. Further-more, the reasons why only some studies find a significant relation have notbeen identified. One possibility is suggested by Saltzer (1982), who found thatWLOC was a much better predictor when the value of weight loss wasconsidered. Rotter (1975) suggested that the failure to consider value is the mostfrequent conceptual problem on the part of locus of control investigators. Asecond possibility is that generalized measures of locus of control are too broadto make good predictions about weight loss (cf. Rotter, 1975). Generalized locusof control allows prediction of a wide range of behaviors, but at a low level.Rotter suggested that when one is seeking to predict behavior for a practicalpurpose, the cost of developing a domain-specific measure may be justified. Anadditional consideration is that the importance of generalized expectancy indetermining behavior decreases as an individual's experience in the situationincreases. The great familiarity of most overweight people with dieting suggeststhat more specific measures of expectancy may be necessary to predict dietingoutcome.

Three studies reviewed here included weight-specific locus of control scales(Ross et al , 1983; Saltzer, 1982; Tobias & MacDonald, 1977). Saltzer's seems tobe the best scale developed thus far, demonstrating superiority in predictingoutcome to health locus of control and generalized locus of control. However,the WLOC had low internal consistency. This may be due to its length, but itraises questions about whether weight locus of control may be a multidimen-sional construct.

This article describes the construction of a new measure of weight locus ofcontrol, its reliability and dimensional structure, and preliminary constructvalidity data. To establish the construct validity of our Dieting Beliefs Scale(DBS), we examined its relations with weight, dieting behavior, and relevantpsychological variables. In general, predictions were based on the idea thathigher scores on the DBS (i.e., more internal beliefs) would be related to reportsof success in weight control. It was predicted that DBS scores would besignificantly and negatively related with weight, the self-perception of having aweight problems, and binge eating. We assumed that heavier individuals andthose with problem eating behaviors have had the experience of not beingsuccessful in exerting control in this area. Significant positive relations with DBSwere predicted for tendency to diet, retrospective reports of weight loss, re-strained eating style, self-esteem, and WLOC. A nonsignificant relation be-tween DBS and social desirability was expected.

DIETING BELIEPS SCALE 195

METHOD

Scale Information

Following Rotter's (1966) definition of locus of control, weight locus of control wasdefined as the expectancy that one can affect or control, at least in part, one'sown weight. The belief that one's own behavior or attributes determine one'sweight is described as a belief in internal weight locus of control. The belief thatone's weight is due to factors outside his or her own control, such as luck, genes,fate, or social support, is labeled a belief in exterruxl weight locus of control. Four ofthe items on the DBS were patterned after those on the Health Locus of Control(HLC) Scale (B. S. Wallston et al., 1976). The 16 items are listed in Table 1. Itemswere balanced to include equal numbers of internal and external items, tocontrol for acquiescence bias. The scale was constructed as if it were measuring

TABLE 1Dieting Beliefs Scale

Please respond to the following statements by indicating how well each statement describes yourbeliefs. Place a number from 1 (not at all descriptive of my beliefs) to 6 (very descriptive of my beliefs)

in the space provided before each statement.

1 2 3 4 5 ÇNot at all Verydescriptive of descriptive of

my beliefs my beliefs

1. By restricting what one eats, one can lose weight.2. When people gain weight it is because of something they have done or not done.

* 3. A thin body is largely a result of genetics.* 4. No matter how much effort one puts into dieting, one's weight tends to stay

about the same.* 5. One's weight is, to a great extent, controlled by fate.* 6. There is so much fattening food around that losing weight is almost impossible.* 7. Most people can only diet successfully when other people push them to do it.

8. Having a slim and fit body has very little to do with luck.9. People who are overweight lack the willpower necessary to conttol

their weight.10. Each of us is directly responsible for our weight.11. Losing weight is simply a matter of wanting to do it and applying yourself.

* 12. People who are more than a couple of pounds overweight need professional helpto lose weight.

13. By increasing the amount one exercises, one can lose weight.* 14. Most people are at their present weight because that is the weight level that is

natural for them.15. Unsuccessful dieting is due to lack of effort.

* 16. In order to lose weight people must get a lot of encouragement from others.

Note: Items labeled with an asterisk (*) are scored in the reverse direction.

196 STOTLAND AND ZUROFF

a unidimensional but broad construct. The dimensional structure was examinedby means of principal-components analysis.

Respondents are asked to indicate how well each statement describes theirown beliefs, using a 6-point scale ranging from not at all descriptive of my beliefs (1)to very descriptive of my beliefs (6). The scale is scored in the internal direction, sothat higher scores represent more internal weight locus of control. Scores canrange from 16 to 96. In our sample, the scale mean was 67.5, with a standarddeviation of 8.7. Scores ranged from 45 to 86.

Subjects

One hundred undergraduate women enrolled in psychology classes at McGillUniversity volunteered to participate in the study. Subjects were asked if theywould be willing to complete several questionnaires concerned with "dietingbeliefs." Subjects were not selected for dieting status.

Inspection of Table 2 reveals that this sample of college women was generallyof normal weight. Subjects' self-reported weights and heights were used to obtaina measure of Body Mass Index (BMI, kg/m^; Keys, Fidanza, Karvonen,Kimurag, &. Taylor, 1972). Subjects' previous weight loss was defined as thedifference between self-reported current and maximum BMI. Only 3 womenwere overweight (BMI > 25.0), and only one could be considered obese (BMI >27.3). Thirteen percent of the sample had been overweight at their highestreported weight and 2 subjects' highest weight placed them in the obese range.Despite the generally normal weight of the sample, a large proportion hadengaged in weight-loss diets. Twenty-three percent of subjects were currentlydieting, and 69% had engaged in previous diets.

Procedure

Subjects were tested in small groups of 3 to 5 women and asked to completeseveral questionnaires, including the DBS, questions about age, weight, height,and dieting history, the Restraint Scale (Herman &. Polivy, 1980), a self-esteemquestionnaire (Rosenberg, 1965), the Binge Scale (Hawkins <St Clement, 1980),the four-item WLOC scale (Saltzer, 1982), and the Marlowe-Crowne SocialDesirability Scale (Crowne &. Marlowe, 1960). Self-perception of having aweight problem was assessed by the question, "Do you see yourself as having a

BMIHighest BMI

M

21.022.6

Current

SD

2.062.40

TABLE 2and Highest Reported

Min

17.017.4

Max '

28.428.4

BMI

% BMÍ > 25

313

% BMI > 27.3

12

DIETING BELIEFS SCALE 197

weigbt problem?" Subjects responded on a 10-point scale, ranging from not at all(1) to serious problem (10). Confidence in reaching goal weigbt was measured ona 10-point scale ranging from not at all confident (1) to very confident (10).Self-rating of success at previous dieting attempts was measured on a 10-pointscale ranging from not at all successful (1) to very successful (10).

To examine the stability of the DBS, subjects were asked to fill out the DBSa second time about 6 weeks later. Forty-three subjects who could be located andwere willing to participate were included in this analysis.

RESULTS

Scale Reliability

Item-total correlations (with the item removed from the total) and coefficientalpha were computed to examine internal consistency. Of the 16 items, 13 baditem-total correlations of .2 or greater. Cronbach's alpha was .68. Deleting thefour items witb lowest item-total correlations improved alpha only slightly(Cronbach's alpha = .69), so all items were retained for further analyses.Test-retest reliability over approximately 6 weeks (N = 43) was .81.

Factor Analysis

A principal-components factor analysis conducted to examine the dimensionalstructure of the DBS. Three factors with eigenvalues greater tban 1.0 wereretained and subjected to varimax rotation. These three factors together ac-counted for 46.2% of the variance (Factor 1, 18.4%, Factor 2, 15.7%; Factor 3,12.1%). The factor loadings of the three rotated factors are presented in Table 3.

Six items bad factor loadings greater tban .40 for the first factor. These itemsappeared to reflect beliefs that weight is under the control of internal factors(e.g., willpower, effort, responsibility). Tbe five items loading above .40 on tbesecond factor concerned beliefs about properties of the individual that arebeyond his or her control (e.g., luck, genes, fate). Four items had factor weightsexceeding .40 for the third factor. These items appeared to reflect beliefs thatweight control is a function of environmental factors (e.g., encouragement fromother people, fattening food). Thus, both Factors 2 and 3 appeared to measureexternal beliefs, but items loading highly on Factor 2 included aspects of theindividual outside her control (e.g., genes), whereas those loading highly onFactor 3 described uncontrollable circumstances outside of the individual (e.g.,encouragement firom others). It should be noted tbat because of the scoringformat, high scores on Factors 2 and 3 represent a rejection of the importance ofexternal factors.

198 STOTLAND AND ZUROFF

TABLE 3DBS Factor Loadings After Varimax Rotation

Item Factor I Factor 2 Factor 3

1 .37 .19 -,122 .35 .15 .003 -.03 .74* -.074 ,17 .72* .315 .10 .71* ,136 -,15 .27 ,52*7 - .08 .05 ,78*8 .41* .48* -.219 .58* - .05 -.34

10 .82* .07 .0211 .74* - .01 -.0612 - .10 .23 ,1613 .59* -.35 ,2914 .06 .63* .41*15 .73* -.07 .0516 ,08 ,08 .69*

Note: Items loading greater than .40 on a factor are indicated by an asterisk.

Relation to Saltzer's (1982) WLOC Scale

The DBS total score correlated .62 with the four-item WLOC scale. WLOC alsocorrelated .30 with Factor 1, .61 with Factor 2, and .22 with Factor 3. Allcorrelations were significant (p < .01). Because WLOC correlates most stronglywith Factor 2, it is reasonable to assume that WLOC primarily measures therejection of uncontrollable factors within the individual as causes of beingoverweight.

Relations to Weight-Related Variables

The relations between DBS total and factor scores, WLOC, and weight-relatedvariables are shown in Table 4. DBS total and WLOC have similar patterns ofcorrelations, demonstrating significant relations with BMI, dieting, lifetimedieting, and self-ratings of success at previous dieting. WLOC is also signifi-cantly related to retrospective reports of weight loss and confidence in one'sability to reach current weight-loss goals. In examining the relations to factorscores, it appears that Factor 1 is related to present weight (BMI) and currentlybeing on a diet, whereas Factors 2 and 3 are related to ratings of previous dietingsuccess and current confidence in reaching weight-loss goals. Factor 2 is alsosignificantly related to retrospective reports of weight loss. The significant

DIETING BELIEFS SCALE 199

TABLE 4Relations Between Weight Locus of ContTol and Weight/Dieting Variables

Variafcie

BMPChange in

BMI'Dieting

(currently)*"Dieting

(lifetime)*"Gonfidence in

reaching goal weight'Success at

previous diets'*

WLOG

.30**

.20*

.18

.35*

.51*

.33**

DBS Totai

.27**

.16

.29**

.41**

.08

.25*

Factor I

.33**

.09

.25*

.20*

-.24

-.08

Factor 2

.002

.20*

.07

.12

.18

.37**

Factor 3

-.01

.04

.19

.01

.43*

.35**

•n = 99. \ = 100. 'n = 23. ""n = 69.*p < .05. **p < .01.

relation between Factor 1 and BMI suggests that the increased tendency to dietassociated with this factor may be related more to dissatisfaction with one'sweight than to confidence in succeeding at weight loss.

Relations to Psychological Variables

The relations between DBS total and factor scores, WLOC, and psychologicalvariables are presented in Table 5. It was predicted that weight locus of controlwould be related to greater experience and success in dieting, less binge eating,and higher self-esteem. The results partially support these predictions but areclarified by an examination of DBS factors. WLOC and DBS total are signifi-cantly related to the self-perception of degree of weight problem, whereas only

TABLE 5Relations Between Weight Locus of Control and Psychological Variables

Variafcie

Self-perceptionof weight problem"

Restraint*"Restraint factors

Diet Goncern*"Weight Fluctuation'

Self-esteem*"Binge eating**Social desirability'

WLOG

.29**

.20

.19

.14

.31**

.03

.29**

DBS Totai

.29**

.22*

.25*

.20

.05

.18

.17

Factor 1

.37**

.28**

.14

.28**-.16

.21*

.12

Factor 2

.11

.04

-.08.09.19.11.05

Fartor 3

-.10-.01

.06-.08

.21*-.04

.19

'n = 99. ""n = 95. 'n = 91. ''n = 100. 'n = 90.*p < .05. **p < .01.

200 STOTLAND AND ZUROFF

Factor 1 is related to this variable. DBS total and Factor 1 are significantlyrelated to restraint. Factor 1 is also related to the Weight Fluctuation factor ofthe Restraint scale (r = .28, p < .01). WLOC and Factor 3 are significantlyrelated to self-esteem. Response to the question, "Do you ever binge eat (i.e., eatin an excessive uncontrolled manner)?" was significantly related to Factor 1 (r =.21, p < .05). Thus the pattern that emerges when the factors are examined isthat Factor 1 is related to negative expectancies and experiences with weightcontrol, whereas Factors 2 and 3 are related to positive expectancies. Anadditional finding was that social desirability was significantly related to WLOC(r = .29, p < .01) but not to DBS.

DISCUSSION

The 16-item DBS demonstrated moderate internal consistency and hightest-retest reliability. Factor analysis suggested the presence of three readilyinterpretable factors, an internal factor and two external factors. The DBScorrelated highly with Saltzer's (1982) four-item WLOC scale; Factor 2 wasparticularly highly correlated with the WLOC.

The global measures of weight locus of control, the WLOC, and the newDieting Beliefs Scale, were significantly related to several weight-related andpsychological variables in college women. More internal scores on the WLOCwere positively related to body mass, likelihood of being on a diet, confidence inreaching goal weight, retrospective reports of weight loss, self-perception ofsuccess at previous diets, self-perception of weight problem, self-esteem, andsocial desirability. Individuals with higher internal scores on this scale thusperceived a need to lose weight and reported confidence in their ability to do so.This attitude is consistent with reports of higher general self-esteem. Thepositive correlation with social desirability suggests that the weight-controlbeliefs represented in the WLOC describe a culturally approved attitude: Thatis, women are expected to demonstrate control over eating and weight. How-ever, WLOC demonstrated some ability to predict dieting success in a previousstudy (Saltzer, 1982) and was related to self-reports of previous dieting success inour study. Internal weight locus of control on the DBS was related to body mass,dieting, self-perception of success at previous diets, self-perception of weightproblem, and cognitive restraint. High scorers on this scale appear to be moreconcerned about their weight and more active in attempting to control it.Global measures of weight locus of control thus appear to reflect a high level ofconcern with one's weight. A difference between WLOC and the DBS is thatWLOC is more related to confidence in weight control and to reports ofprevious weight loss. The two scales appear to be measuring somewhat differentaspects of the weight-locus-of-control construct. The difference between thescales is clarified by an examination of the factor structure of the DBS.

DIETING BELIEFS SCALE 201

Examination of relations between BMI and the DBS factors indicated thatonly the relation with Factor 1 was significant. Thus, heavier subjects were morelikely to endorse internal beliefs about weight. Factor 1 also demonstratedsignificant correlations with tendency to diet (present and lifetime), and with theWeight Fluctuation factor of the Restraint scale. This suggests that individualsscoring highly on Factor 1 have had more experience than most in losing (andgaining) weight. Factor 1 also correlated significantly with the self-perception ofhaving a weight problem and the tendency to binge eat. Those scoring high onthis factor were heavier and had negative feelings about their weight, which mayhave contributed to the feeling that they "should" diet and "should" be able tocontrol their weight and to the adoption of internal beliefs about weight control.Interestingly, Factor 1 was not related to ratings of success at previous diets or toconfidence in reaching goal weight. Thus, Factor 1 was related to the tendencyto diet, periodic fluctuations in weight, and negative feelings about one's weightbut not to success at dieting or weight control.

Factor 2, which seemed to represent the rejection of factors within theindividual but outside volitional control, was not related to BMI or dietingtendency; however, it was significantly and positively related to ratings ofsuccess at previous diets and to retrospective reports of weight loss. Factor 3appeared to refiect the rejection of the belief that external factors are responsiblefor weight. This factor was not related to BMI or dieting tendency but wassignificantly related to confidence in reaching goal weight, ratings of success atprevious diets, and self-esteem. Thus, Factors 2 and 3 had fairly similar patternsof association with the other variables of concern. In contrast to Factor 1, thesefactors seem to refiect positive expectancies and experiences in weight control.

Unidimensional measures of weight locus of control suggested that internalswere more positive about the prospects for weight control. This is consistentwith the view of previous authors who attempted to predict weight-loss successwith locus of control. That line of research produced an unsatisfying pattern ofresults and little evidence that locus of control was a useful predictor of successat weight loss. Our results suggest that weight locus of control is best thought ofas a multidimensional construct. Future research attempting to predict weightloss should examine its relation with the three factors of the DBS.

It appears that certain types of weight-related locus of control beliefs areassociated with negative self-image and "guilty" (though not necessarily success-ful) dieting, whereas other beliefs are related to positive expectancies aboutdieting. This suggests that treatment programs for obesity should attempt tomodify extreme internal weight locus of control beliefs, because these beliefs maybe related to the kind of perfectionistic approach to dieting thought to predis-pose one to dietary breakdown (cf. Polivy &. Herman, 1987). In addition,treatment programs should encourage the patterns of beliefs refiected in Factors2 and 3.

Some limitation of the present findings should be mentioned. The factor

202 STOTLAND AND ZUROFF

structure of the DBS needs to be cross validated in an additional sample. It willalso be important to examine, prospectively, its ability to predict dietingbehavior and success in weight control. Finally, it is unknown whether thesefindings will generalize outside a college population.

Our sample was composed primarily of normal-weight women. The highprevalence of dieting in this sample, consistent witb previous findings (e.g.,Dwyer &. Mayer, 1970), appears to be more related to tbe self-perception ofbeing overweight (probably culturally determined) than to actually being over-weight. Whether the DBS will demonstrate comparable psycbometric propertiesand relations witb dieting and psycbological variables in an overweight sampleremains to be investigated.

REFERENCES

Balch, P., & Ross, A. W. (1975). Predicting success in weight loss as a function of locus of control:A unidimensional and multidimensional approach. Joumci of Consulting and Clinical Psychology,43, 119.

Crowne, D. P., &. Marlowe, D. (1960). A new scale of social desirability independent of psychopa-thology. Journal of Consulting Psychology, 24, 349-354.

Dwyer, J. T., &L Mayer, J. (1970). Potential dieters: Who are they? Journal of the American DieteticAssociation, 56, 510-514.

Gormally, J., Rardin, D., & Black, S. (1980). Correlates of successful response to a behavioral weightcontrol clinic. Journal of Couaseling Psychology, 27, 179-191.

Hawkins, R. C , & Clement, P. (1980). Development and construct validation of a self-reportmeasure of binge eating tendencies. Addictive Behaviors, 5, 219-226.

Herman, C. P., & Polivy, J. (1980). Restrained eating. In A. J. Stunkard (Ed.), Obesity (pp. 208-225).Philadelphia: Saunders.

Keys, A., Fidanza, F., Karvonen, M. J., Kimura, N., &. Taylor, H. L. (1972). Indices of relativeweight and obesity. Journal of Chronic Diseases, 25, 329-343.

Polivy, J., & Herman, e . P. (1987). Diagnosis and treatment of normal eating. Jourrwl of Consultingand Clinical Psychology, 55, 635-644.

Reid, D. W., &. Ware, E. E. (1974). Multidimensionality of internal-external control: Implicationsfor past and future research. Canadian Journal of Behavioural Sciences, 5, 264-271.

Rosenberg, M. (1965). Society arui the adolescent self-irrwge. Princeton, NJ: Princeton University Press.Ross, M. W., Kalucy, R. S., &. Morton, J. E. (1983). Locus of control in obesity: Predictors of success

in a jaw-wiring programme. British Jourrwl of Medical Psychology, 53, 49-56.Rotter, J. B. (1966). Generalized expectancies for internal versus external control of reinforcement.

Psychological Monograi)hs, 80(1, Whole No. 609).Rotter, J. B. (1975). Some problems and misconceptions related to the construct of internal versus

external control of reinforcement, journal of Consulting arui Clinical Psychology, 43, 56-67.Saltzer, E. B. (1982). The Weight Locus of Gontrol (WLOO) Scale: A specific measure of obesity

research. Journal of Personality Assessment, 46, 620-628.Tobias, L. L., &1. MacDonald, M. L. (1977). Internal locus of control and weight loss: An insufficient

condition. Jourrwl of Consulting and Clinical Psychology, 45, 647-653.Wallston, B. S., Wallston, K. A., Kaplan, G. D., & Maides, S. A. (1976). Development and

validation of the health locus of control (HLC) scale. Jourrud of Consulting and Clinical Psychology,44, 580-585.

DIETING BELIEFS SCALE 203

Wallston, K, A,, Wallston, B. S,, Si. DeVellis, R. (1978), Development ofthe multidimensionalhealth locus of control (MHLC) scales. Health Education Monographs, 6, 160-170,

Stephen StotlandDepartment of PsychologyMcGill University1205 Docteur Penfield AvenueMontreal, Quebec H3A lBlCanada

Received May 10, 1988Revised August 24, 1988

Copyright of Journal of Personality Assessment is the property of Taylor & Francis Ltd and its content may not

be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written

permission. However, users may print, download, or email articles for individual use.