Social Desirability as a Variable in the Locus of Control Scale

Abstract

The present series of studies based on Ss in Catholic institutions attempted to explore several related questions concerning the social-desirability variable in the Internnl-External Locus of Control scale (I-E). Results of Smdy I indicated that a majoriry of the I-E items have probability of endorsement values which depart significantly from a hypothetical value of P = S O , both for standard instructioc and social-desirability insrruction conditions. Study I1 demonstrated a small relationship between S's I-E score and S's tendency to agree with I-E items of high response-preference value ( r = .20). This relationship was especially obvious for those Ss designated as internally controlled. In Study I11 a significant relationship was shown between the social-desirability scale values for I-E items and the probability of I-E item endorsements ( T = .43). Moreover, a sizeable number of internal items were rated a significantly more socially desirable than the corresponding external items. It was suggested that the I-E scale may be contaminated by social desirability and that consequently the validicy of the scale as a measure of locus of control is questionable.The Internal-External Control of Reinforcement Scale (I-E) was constructed by Rotter (1966) as a self-rzport measure of generalized belief that a person can control his own destiny. The scale consists of 23 keyed items and 6 buffer items presented in a forced-choice format under the guise of a personal reaction inventory. S's score is determined by the number of external choices he makes, so that the range of possible scores extends from 0 to 23. A high score represents frequent endorsement of items attributing causality to environmental factors (e.g., chance, fate, luck, or powerful others), whereas a low score represents choices reflecting the attribution of causality to personal factors (e.g., skill or capacity). Thus, individuals are presumed to vary along a unidimensional continuum of "locus of conuol," n-ith the anchor points labeled external and internal.The I-E scale has received substantial research attention in recent years. Rotter (1966) and Lefcourt (1966) have presented well-organized surveys of the research literature and have amply documented the reliability and validity of the scale in a variety of experimental and natural field settings. In contrast, correlational studies of the I-E scale and other personological variables have often yielded disappointing resulcs. According to Rotter, che inability to establish suonger relationships may stem, at least in part. from the influence of sinlation 'The author wishes to extend his zppreciation to Randall Clouser and Frank Pepe for their assistance i n collecting and analyzing the data.

Differentiating Among Internality, Powerful Others, and Chance

“…Several investigators (e.g., Hjelle, 1971) have pointed to the finding that Rotter's I-E Scale is contaminated by social desirability, which affects validity. Levenson (1972) found correlations of the I, P, and C Scales with the Marlowe-Crowne Social Desirability Scale of .09, .04, and -.10, respectively.…”

Competence, Contingency, and the Development of Perceived Control

“…Adults generally regard internal attribu tions as more desirable than external ones, and scores on the Rotter scale are significantly correlated with scores on the Edwards Social Desirability Scale [Cone, 1971;Hjelle, 1971;Joe, 1972;Lefcourt, 1976; see counter-evidence from Gozali and Bialer, 1968, using the Bialer scale]. Very young children, by contrast, may be less influenced by social desirability effects; they may be less likely to comprehend all of the LOC items and less likely to perceive that internal responses are 'more socially desirable'.…”

Illness beliefs and locus of control

“…Social desirability was included in the Swedish version because the response set of this factor has been pointed out as a possible confounding variable in the assessment of forced-choice self-report scales [11]. The instrument comprises 50 statements with a four-point forced-choice scale with the following alternatives: (1) strongly disagree, (2) disagree, (3) agree and (4) strongly agree.…”