KR-20 Reliability Coefficient Research Articles

Reliability of the Frostig Test of Visual Perception in a South African Sample

The Marianne Frostig Developmental Test of Visual Perception was administered to 31 preschool children (19 boys and 12 girls) whose mean age was 68 mo. An item analysis showed that 52% of the test items satisfied the requirements with regard to degree of difficulty and discrimination. KR-20 reliability coefficients varied between 0.31 and 0.58 for the different subtests, with r=0.72 for the full scale.

Estimation of the KR20 reliability coefficient when data are incomplete

Attention is focused on the estimation of the KR20 reliability coefficient when data are incomplete. In terms of point estimate, it is found that the analysis of variance procedure (Henderson's Method 1) would be most suitable among the three techniques studied. Confidence coefficients are provided for the KR20 when data are incomplete.

Effects of emotional state on sexual desire and sexual dysfunction

The present study explored effects of emotional states on sexual desire and dysfunction. Three basic emotion dimensions, pleasure-displeasure, arousal-nonarousal, and dominance-submissiveness, served as the independent factors. Factor analysis of a sexual desire and problems questionnaire yielded a single factor with sexual desire as one pole and sexual problems as the opposite pole. The resulting “sexual desire” scale had a KR-20 reliability coefficient of .98 and constituted a highly homogeneous dependent measure. Sexual desire was greater when subjects felt pleasure rather than displeasure, unaroused rather than aroused, and dominant rather than submissive. Desire increased with increasing arousal in pleasant states and decreased with increasing arousal in unpleasant states (Pleasure × Arousal interaction). Desire increased with increasing dominance in pleasant states and decreased with increasing dominance in unpleasant states (Pleasure × Dominance interaction). In unaroused states, submissiveness attenuated the polarization of sexual desire as a function of pleasure-displeasure (Pleasure × Arousal × Dominance interaction). The highest levels of sexual desire and freedom from sexual problems were associated with pleasant and dominant feelings (e.g., admired and vigorous or relaxed and leisurely); the lowest levels were reported for unpleasant and aroused feelings (e.g., hostile and hateful or pained and embarrassed). Except for the uniformly greater sexual desire of males than of females across all conditions, all findings were applicable equally to the two sexes.

Measures of Individual Differences in Temperament

Reliability and validity data are presented for three orthogonal measures of temperament: trait-pleasure, trait-arousal, and trait-dominance. The KR-20 reliability coefficients for the three scales were .91, .60, and .84, respectively. Intercorrelations among the three scales were insignificant and low. Thus, the expectation that these three measures could be used as a parsimonious base for the description of temperament was confirmed. The trait-pleasure scale was correlated positively with measures of social desirability, achieving tendency, extroversion, and affiliative tendency, and was correlated negatively with measures of trait-anxiety, neuroticism, and aggression. The trait-arousal scale was not related to the social desirability scale, and was correlated positively with measures of trait-anxiety, neuroticism, and aggression. The trait-dominance scale was also free of social desirability bias. It exhibited positive correlations with measures of dominance-submissiveness, autonomy, extroversion, achieving tendency, and aggression, and negative correlations with measures of trait-anxiety and neuroticism.

Factor Analysis and Reliability of the Mental Dexterity Test

The primary objective of the present study was to factor analyze a recently developed mental ability test in order to develop verbal and numerical subtesrs—then to establish the reliability of the subtests (and of the test as a whole) using 97 advanced college students as Ss. Using a varimax rotation, some 34 factors were extracted, the most prominent two factors being verbal and numerical in nature. Based mainly on factor-loadings, two subtests were developed—a 50-item verbal comprehension test and a 30-item numerical ability test. The KR-20 reliability coefficients for the verbal and numerical subtests were .89 and .82 respectively, whereas that for the entire 100-item test was .93.

The Mental Dexterity Test: A Study of Reliability and Validity

The objective of the investigation was to ascertain the reliability and concurrent validity of a new omnibus mental ability test which has considerably more ceiling than do most group intelligence tests, and which is proposed for use either as a speed or power device, once the appropriate norms are established. When administered to 184 college students under untimed conditions, the test was found to have a split-half reliability coefficient of .84 and a KR-20 reliability coefficient of .85. Regarding validity, the test correlated significantly with grades assigned in the classes which participated in the study (.26) and also with cumulative grade point average based on all courses taken at the university (.40).