Showing papers in "Educational and Psychological Measurement in 1979"
TL;DR: In this article, the authors presented a scale for self-esteem in social situations, which was rated by 265 subjects and subsequently subjected to an iterative item-total correlation and analysis in order to maximize factorial purity.
Abstract: Seventy six items describing self-esteem in social situations were rated by 265 subjects, and subsequently subjected to an iterative item-total correlation and analysis in order to maximize factorial purity. Factor analysis of the remaining 30-item scale revealed a single general factor that accounted for 39.7% of the total variance. A new group of 128 subjects completed the 30-items twice, and a retest reliability over 4-weeks indicated that the scale was highly reliable (r = 0.88). Normative data, based on the original 265 subjects, is provided in the appendices along with a description of the scale.
128 citations
TL;DR: This paper investigated the effects of two scoring formulas on the discriminant validity of the dimensions of divergent thinking and found that the utility of the percentage formula is limited to subjectively determined originality scores.
Abstract: Two multitrait-multimethod studies were conducted to investigate the effects of two scoring formulas on the discriminant validity of the dimensions of divergent thinking. In the first study, 39 fifth-grade students were given the Torrance Tests of Creative Thinking. In the second study, 60 college students were given Alternate Uses, Plot Titles, and Consequences. Findings indicate that the discriminant validity of subjectively determined originality scores is enhanced when the scores are divided by the total number of responses (yielding a percentage score). However, when the percentage formula was used to derive flexibility scores from the Torrance Tests or an originality score based on statistical rarity from the Alternate Uses test, the coefficient alpha reliabilities dropped to near zero. This finding suggests that the utility of the percentage formula is limited to subjectively determined originality scores.
105 citations
TL;DR: In this paper, the authors measured the size of an effect for fixed effects factorial analysis of variance (ANOVA) experimental designs, and provided a brief review of the literature.
Abstract: The present paper is concerned with measuring the size of an effect for fixed effects factorial analysis of variance (ANOVA) experimental designs. A brief review of the literature is provided, emph...
80 citations
TL;DR: In this article, an empirical investigation of the psychometric properties of a commonly-used instrument to measure cognitive complexity was designed, which reinforced the internal validity of cognitive complexity research using the instrument.
Abstract: An empirical investigation of the psychometric properties of a commonly-used instrument to measure cognitive complexity was designed. Bieri's (Bieri Atkins, Briar, Leaman, Miller, and Tripodi 1966) grid form of Kelly's (1955) Role Construct Reperatory (REP) Test was shown to have high and statistically significant test-retest reliability (p < .001) and convergent and discriminant validities (p < .05). These results reinforce the internal validity of cognitive complexity research using the instrument. Moreover, stable scores were found across student (N = 176) and nonstudent (N = 169) samples. Such characteristics as sex, level in the organizational hierarchy, year in college, and college major were not found to alter the distributions of scores. Thus, preliminary norm tables for scores were established which would be necessary in building arguments for external validity of research results.
58 citations
TL;DR: In this paper, five different shrinkage formulas were compared to see which most accurately reduced the positive bias in sample R2 values as estimators of the squared population multiple correlation coefficient (ρ2).
Abstract: Five different shrinkage formulas were compared to see which most accurately reduced the positive bias in sample R2 values as estimators of the squared population multiple correlation coefficient (ρ2). Artificial populations of known parameters were constructed and "Monte Carlo" sampling was done with a computer. The following parameters were examined: number of independent variables (5 and 15), sample size (400, 200, 100, 50, 25, and 10), and ρ2 (.2, .4, .6, and .8). On the basis of the data collected, one formula was eliminated and the conditional use of two others was recommended.
41 citations
TL;DR: In this paper, the determinacy problem is considered in the context of image correlation matrix factor analysis, with particular reference to determinacy matrix determinacy in the oblique case, and a tentative resolution is proposed.
Abstract: The factor analysis of the image correlation matrix is suggested. Whether to factor the image correlation matrix itself or to proceed via a new model with an alpha (common) factor analysis of it is mentioned, with particular reference to the determinacy problem. The transformation (rotation) problem, in the oblique case, has a new aspect, for which a tentative resolution is proposed. It is pointed out that the distribution of the images is sensibly multivariate normal, making for "better" factor analyses.
41 citations
TL;DR: The Student's Perception of Ability Scale (SPAS) as mentioned in this paper was developed in response to a need for a reliable and valid instrument to measure subject-based academic self-concept in elementary school children.
Abstract: The Student's Perception of Ability Scale (SPAS) was developed in response to a need for a reliable and valid instrument to measure subject-based academic self-concept in elementary school children...
37 citations
TL;DR: In this article, the authors presented a chi-square statistic for testing pattern hypotheses on correlation matrices, which is based on a multivariate generalization of the Fisher r-to-z transformation.
Abstract: The program presented computes a chi-square statistic for testing pattern hypotheses on correlation matrices. The statistic is based on a multivariate generalization of the Fisher r-to-z transformation. Monte Carlo studies have demonstrated that this statistic has small sample performance which is superior to an analogous likelihood ratio statistic obtained via the analysis of covariance structures.
35 citations
TL;DR: In this paper, the Verbal and Mathematics portions of the College Board Scholastic Aptitude Test (SAT-V and SAT-M) and high school grade point average (HSGPA) were shown to have moderate predictive validity for the entire group.
Abstract: Records of 402 students (206 men and 196 women) in the Honors Program at Marquette University were studied. The Verbal and Mathematics portions of the College Board Scholastic Aptitude Test (SAT-V and SAT-M) and high school grade point average (HSGPA) were shown to have moderate predictive validity for the entire group. However, sex differences were seen. Only SAT-M for women was found to have practical usefulness in predicting success in the Honors Program.
24 citations
TL;DR: In this article, the authors compare the Newton-Raphson approximation to the maximum likelihood estimate with several other possible procedures, and conclude that it should be used when it yields admissible (positive) results.
Abstract: For some situations the beta-binomial distribution might be used to describe the marginal distribution of test scores for a particular population of examinees. To use this distribution it is necessary to estimate two parameters which characterize the beta-binomial model. One method of estimating these parameters is to approximate the maximum likelihood estimates using some appropriately chosen iterative technique. When the number of examinees is small, however, it is not clear that this method of estimation is justified since the iterative approximations might not converge to the maximum likelihood estimate. Using Monte Carlo techniques, this paper compares the Newton-Raphson approximation to the maximum likelihood estimate with several other possible procedures. It is found that the Newton-Raphson method should be used when it yields admissible (positive) results.
22 citations
TL;DR: In this article, the construct validity of class ratings of instructors measured by the Course-Faculty Instrument (CFI) was assessed, and two sets of variables were used which explicate aspects of the nomological network of instructor ratings: (a) three intrinsic variables of instructor behavior represented wanted sources of rating variance (e.g., perceived effectiveness of instructor communications), and (b) six extrinsic variables represented variance often considered to be undesirable, such as instructor sex).
Abstract: The construct validity of class ratings of instructors measured by the Course-Faculty Instrument was assessed. Two sets of variables were used which explicate aspects of the nomological network of instructor ratings: (a) three intrinsic variables of instructor behavior represented wanted sources of rating variance (e.g., perceived effectiveness of instructor communications), and (b) six extrinsic variables represented variance often considered to be undesirable, (e.g., instructor sex). Eighty-nine percent of the variance in the ratings of 129 instructors teaching 334 classes (from 10,738 student ratings) was explained by the intrinsic measures of instructor behavior. Three of the extrinsic variables correlated significantly with intrinsic variables and/or ratings (r's ≤ .23). However, these latter variables (experience, rank, and publishing activities) may be as suggestive of validity as bias.
TL;DR: A program is described to cover the vast majority of factor analytic requirements with nearly the same degree of ease as that afforded by use of the "little jiffy" programs.
Abstract: A program is described that answers the intermediate need for a computer procedure that can provide a reasonably sophisticated factor analysis for as many as 500 variables—a program to cover the vast majority of factor analytic requirements with nearly the same degree of ease as that afforded by use of the "little jiffy" programs.
TL;DR: In this paper, the influence of reading ability and item format on objective test performance was investigated for nine and tenth grade students in a basic health science course, randomly assigned to a true-false, multiple choice, or matching testing condition.
Abstract: As achievement in basic reading skills declines, the content validity of tests in special subjects (e.g., science or social studies) may be reduced because examinees cannot read the test items. This empirical study was designed to investigate the influence of reading ability and item format on objective test performance. Subjects were ninth- and tenth-grade students in a basic health science course, randomly assigned to a true-false, multiple choice, or matching testing condition. Results indicated that both reading ability and item format significantly influenced test performance. Examinees performed best on items presented in a matching format. Implications were drawn for evaluating content validity.
TL;DR: In this article, the Life Experience Inventory (LEI) was used for identification of creative electrical engineers and a point-biserial correlation coefficient of.82 was found between inventory scores and a dichotomous criterion of possession or lack of possession of one or more patents (held or pending).
Abstract: The construction and validation of the Life Experience Inventory (LEI) for identification of creative electrical engineers is described. For an initial-validation sample of 100 electrical engineers in a large industrial corporation, a point-biserial correlation coefficient of .82 was found between inventory scores and a dichotomous criterion of possession or lack of possession of one or more patents (held or pending), whereas in a cross-validation sample of 98 electrical engineers from the same firm, the corresponding coefficient was .62. Application of an approximate median cutting score on the LEI for the cross-validation sample revealed for the group above the median score a success rate of about 83% in terms of those electrical engineers who would meet the patent criterion at a cost of losing within the group placing below the median score about 29% who would have met the patent criterion. It was concluded that life experiences (as sampled by items emphasizing factual information rather than those ref...
TL;DR: The six subtests comprising the Metropolitan Readiness Tests were factor analyzed separately for white females, white males, black females, and black males as discussed by the authors, and a single factor best described the structure of the test for each group.
Abstract: The six subtests comprising the Metropolitan Readiness Tests were factor analyzed separately for white females, white males, black females, and black males. A single factor best described the structure of the test for each group. Coefficients of congruence for the factor between each pair of groups always exceeded .92. Thus, the factor structure of the Metropolitan Readiness Tests is essentially invariant with regard to race and sex variables.
TL;DR: In this paper, data obtained from two administrations of the Piers-Harris Children's Self Concept Scale to a single group of 159 elementary pupils were factor analyzed to determine the stability of factors from one test to another.
Abstract: Data obtained from two administrations of the Piers-Harris Children's Self Concept Scale to a single group of 159 elementary pupils were factor analyzed to determine the stability of factors from one test to another. Of ten factors from each test administration extracted and rotated by means of the varimax method, five in each test sample were found to have sufficient item content commonalities to warrant joint interpretation and labeling. The findings suggest that the factorial dimensions of the Piers-Harris scale can vary considerably even within one study group. Researchers therefore would do well to determine the factorial structure of their own test data as well as to consider carefully the variance accounted for by each factor if they wish to draw meaningful conclusions from findings involving changes in self-concept in terms of particular constructs. Implications of further factorial validity research of the Piers-Harris scale are discussed.
TL;DR: In this paper, a single-stage, non-Bayesian solution based on a strong true-score model is proposed to determine how many items to include on a mastery test.
Abstract: A problem of considerable importance in certain educational settings is determining how many items to include on a mastery test. Applying ranking and selection procedures, a solution is given which includes as a special case all existing single-stage, non-Bayesian solutions based on a strong true-score model. Illustrations are given for the binomial and compound binomial distributions.
TL;DR: In this article, the item development, reliability, and validity of a scale to measure attitude toward reading among elementary school pupils are described, and results suggest that this scale will yield accurate and v...
Abstract: The item development, reliability, and validity of a scale to measure attitude toward reading among elementary school pupils are described. Results suggest that this scale will yield accurate and v...
TL;DR: A literature review was carried out to determine existing knowledge regarding the relationship of GRE scores and graduate grade point average (GGPA) as discussed by the authors, and the results showed that GRE scores were positively associated with GPAs.
Abstract: A literature review was carried out to determine existing knowledge regarding the relationship of Graduate Record Examinations Aptitude Test (GRE) scores and graduate grade point average (GGPA). Bu...
TL;DR: In this article, the Least Preferred Coworker Scale (LPC) scales of 287 airmen were analyzed and it was shown that low LPC is more closely related to cognitive complexity than is high LPC.
Abstract: Cognitive complexity has been offered (Mitchell, 1970; Foa, Mitchell, and Fiedler, 1971) as an alternative conceptualization of Fiedler's (Fiedler, 1967) measure of leadership (i.e., the Least Preferred Coworker Scale—LPC). In order to establish the validity of this proposition, the completed LPC scales of 287 airmen were analyzed. ANOVA tests for linear trend revealed that, contrary to current theorizing, low LPC is more closely related to cognitive complexity (as operationalized by LPC item-variability) than is high LPC.
TL;DR: In this paper, a sample of more than 500 freshmen at California State University, Northridge (CSUN) who had elected to take the general English composition course (English 155) during their first semester as well as for male and female subsamples of comparable size, comparisons were made of the predictive validity of the Californian State University and Colleges English Placement Test (CSUC-EPT) (Educational Testing Service, 1977a, b), the Test of Standard Written English (TSWE) of the Admissions Testing Program of the College Board, 1974-1978,
Abstract: FOR a sample of more than 500 freshmen at California State University, Northridge (CSUN) who had elected to take the general English composition course (English 155) during their first semester as well as for male and female subsamples of comparable size, comparisons were made of the predictive validity of the California State University and Colleges English Placement Test (CSUC-EPT) (Educational Testing Service, 1977a, b), the Test of Standard Written English (TSWE) of the Admissions Testing Program of the College Board (Educational Testing Service, 1974-1978, 1978a, 1978b), and the Verbal and Mathematics portions of the College Board Scholastic Aptitude Test (CEEB—SAT—Verbal and CEEB-SAT—Mathematics) (Educational Testing Service, 1948-1978) with respect to each of four criterion measures: (a) Grades in Fall Semester 1977 English 155—Written Expression, (b) Fall Semester 1977 GPA, (c) Spring Semester 1978 GPA, and (d) 1977-1978 Academic Year (Total CSUN GPA. Statistical findings suggested these conclusio...
TL;DR: The dependence of the item discrimination index (D) on the item difficulty index (p), and the relationship of D and p to the phi coefficient (ϕ) are delineated.
Abstract: The dependence of the item discrimination index (D) on the item difficulty index (p), and the relationship of D and p to the phi coefficient (ϕ) are delineated. A table of critical values of D for significance at the .10 and .05 levels, with the number of examinees in the upper group (n, ) on total test score ranging from 10 to 25 and p ranging from .10 to .90, is presented. This table may be used to determine an acceptable magnitude of D for a given p and n1 or to find whether a particular combination of p and D will yield a significant ϕ.
TL;DR: In this article, nine college instructors were rated by their students on five performance dimensions, and randomized thirds of each class used either standard graphic, behaviorally anchored, or mixed standard scales to make their ratings.
Abstract: Nine college instructors were rated by their students on five performance dimensions. Randomized thirds of each class used either standard graphic, behaviorally anchored, or mixed standard scales to make their ratings. Analyses of the ratings indicated (a) somewhat less positive leniency when the mixed standard scales were used rather than the graphic or behaviorally anchored rating scales, (b) generally lower interdimension correlations for ratings made on mixed standard or behaviorally anchored scales than on standard graphic scales, and (c) no significant relationship between mean instructor ratings and instructor leniency in assigning grades. Implications for scale validity were also considered.
TL;DR: In this paper, an examination was made of the predictive validity of the GRE-V, GRE-Q, and GRE-T test scores, which were correlated with performance in Educational Foundations (FED) courses which involved at least one course in research and statistics.
Abstract: An examination was made of the predictive validity of the Graduate Record Examinations Aptitude Test, Verbal, Quantitative, and Total scores (GRE-V, GRE-Q, and GRE-T) and undergraduate grade-point average (UGPA). Criterion variables consisted of graduate grade-point average (GGPA), performance on the Master's Comprehensive Examination (MCE), and grades in required area and core courses. Subjects were 107 students enrolled in the Department of Counselor Education Master's degree program at a large Southwestern university. Scores on the GRE-V were correlated .24 (p < .01) with both MCE scores and GGPA, but were not correlated significantly with any other criterion variable. GRE-Q test scores were correlated .31 (p < .001) with performance in Educational Foundations (FED) courses which involved at least one course in research and statistics. The GRE-T measure was a valid predictor of grades in FED courses and of GGPA. Although each predictor variable was correlated with at least one of the criterion variable...
TL;DR: This article found that the GRE-V part was correlated significantly with GGPA and GRE-Q was not, whereas GRE-T score was correlated.24 with the GGPA, and it would appear that theGRE-V portion largely contributed to this correlation.
Abstract: This study was concerned with a determination of the predictive validity of the Graduate Record Examinations Aptitude Test: the verbal section (GRE-V), the quantitative portion (GRE-Q) and the total score (GRE-T). The criterion variable was the graduate grade point average (GGPA); subjects were 135 students enrolled in a Master of Arts in Counseling Program (MAC) at an urban university. The GRE-V part was correlated significantly with GGPA (.26), whereas the GRE-Q was not. The GRE-T score was correlated .24 with GGPA. It would appear that the GRE-V portion largely contributed to this correlation. These findings raise questions regarding the adequacy of the predictive validity of the GRE Aptitude Test for this sample of students for the criterion measure employed.
TL;DR: The Raven Coloured Progressive Matrices and a Piagetian battery were administered to 120 hearing-impaired children, ages 6-11 years, under one of two non-elaborative conditions as discussed by the authors.
Abstract: The Raven Coloured Progressive Matrices and a Piagetian battery were administered to 120 hearing-impaired children, ages 6-11 years, under one of two nonelaborative conditions—(1) standard, or (2) ...
TL;DR: In this paper, the authors explore the nature of these problems and suggest strategies for circumventing them by using factor scale scores as weights to create composite variables representing the variables included in a factor analysis.
Abstract: Factor scale scores are sometimes used as weights to create composite variables representing the variables included in a factor analysis. If these composite variables are then used to predict some dependent variable, serious theoretical and methodological problems arise. This paper explores the nature of these problems and suggests strategies for circumventing them.
TL;DR: In this article, a function of the age at entrance into the program was highly negatively related to both scholarship and professionalism for 61 doctoral recipients, and the scores from the Graduate Record Examinations (GRE) Aptitude Test and Advanced Education Test gave somewhat lower correlations than did the grade-point averages.
Abstract: Averaged rankings by faculty members on scholarship and professionalism were used as criterion variables for 61 doctoral recipients. A function of the age at entrance into the program was highly negatively related to both criteria. Undergraduate and graduate grade-point averages showed statistically significant validities of moderate magnitude. The scores from the Graduate Record Examinations (GRE) Aptitude Test and Advanced Education Test gave somewhat lower correlations than did the grade-point averages. Use of the GRE for various purposes is discussed.
TL;DR: A computer program called Program CONGRU as discussed by the authors analyzes the overall conjoint agreement among many observers for their classifications on categorical scales, and a special feature assesses the significance of observers' congruence in assignments to each nominal category considered.
Abstract: A computer program entitled Program CONGRU is presented which analyzes the overall conjoint agreement among many observers for their classifications on categorical scales. A special feature assesses the significance of observers' congruence in assignments to each nominal category considered.
TL;DR: The Classroom Behavior Rating Scale (CBRS) as mentioned in this paper was developed to provide school psychologists with a measure of learning-related classroom behaviors of elementary school pupils. The initial item pool co...
Abstract: The Classroom Behavior Rating Scale (CBRS) was developed to provide school psychologists with a measure of learning-related classroom behaviors of elementary school pupils. The initial item pool co...