Cooper B. Holmes

Sample Size in Psychological Research over the Past 30 Years

The American Psychological Association (APA) Task Force on Statistical Inference was formed in 1996 in response to a growing body of research demonstrating methodological issues that threatened the credibility of psychological research, and made recommendations to address them. One issue was the small, even dramatically inadequate, size of samples used in studies published by leading journals. The present study assessed the progress made since the Task Forces final report in 1999. Sample sizes reported in four leading APA journals in 1955, 1977, 1995, and 2006 were compared using nonparametric statistics, while data from the last two waves were fit to a hierarchical generalized linear growth model for more in-depth analysis. Overall, results indicate that the recommendations for increasing sample sizes have not been integrated in core psychological research, although results slightly vary by field. This and other implications are discussed in the context of current methodological critique and practice.

Color preference as a function of the object described

Noting the inconsistent research on the relationship between color and personality, the authors present data on one of the possible reasons for the problem. The problem addressed is that color/personality research almost always presents a color patch or patches and draws inferences from the subjects’ choices, ignoring the possibility that color preference is related to the object in question. College students were asked to give their favorite colors for a number of items (e.g., automobile). The results clearly show that color preference is a function of the object described. That is, color preference cannot be asked independently of an object if that color preference is to have interpretive significance.

Correct Classification, False Positives, and False Negatives in Predicting Completion of the Ph.d. from GRE Scores

Combined Verbal and Quantitative GRE scores were obtained from the records of 24 former students of a masters degree program (from a total of 128 students) who had successfully completed a doctorate in psychology or who had withdrawn from a psychology doctoral program. Success rate by classification with the GRE was calculated using both a cut-off of 1000 and a cut-off of 1100. The results indicated a high false negative rate, that is, students whose GRE scores would not predict success but who obtained a Ph.D.

Relationship between Personality Traits and Fear of Death

Personality traits and reported levels of death fear were compared for 75 individuals. The subjects were undergraduate men and women in psychology who completed the 16 PF and the Death Anxiety Scale. Five personality traits were associated with a high fear of death; however, none were to a degree that would allow the statement that people with a high fear of death are necessarily abnormal. Those with a high fear of death were relatively more easily affected by feelings, less trusting, less self-assured, less socially precise, and more tense than people with a low fear of death.

The barnum effect in luscher color test interpretation

A series of statements from the interpretation section of the Luscher Color Test were selected randomly and put into paragraph form. College students read these interpretative statements under several conditions to assess how accurately the statements described themselves. On a scale from 1 to 9, the ratings were all 5 or higher, which indicates a Barnum effect from the Luscher Color Test.

SAMPLE SIZE IN PSYCHOLOGICAL RESEARCH

This study was conducted to provide information on the typical sample size employed in psychological research, as it is reported in selected American Psychological Association journals. All the articles in those journals for the years 1955 and 1977 were read to ascertain both the total n for the study, and the size of each of the groups that made up that total n. The mean, SD, median, 25th percentile, 75th percentile, modes, and range were calculated for each year, both for total n and for individual group n. A comparison of the 1955 data and the 1977 data showed no significant changes in sample size between those years. Certain problems in determining sample size, as it was reported, are presented in this article.

Relationship between the luscher color test and the MMPI

Administered the short form of the Luscher Color Test and the booklet form of the MMPI to 42 graduate students in a counseling program. For every S a report that described personality was written from the test results, i.e., each S had two reports, one from the Color Test and one from the MMPI. These reports were written independently by two separate examiners. Two independent raters then read the reports on each S and noted the degree of agreement between the two reports. The overall results showed very little agreement between the Color Test and the MMPI. Reasons for the poor rate of agreement are discussed.

Relationships among scores on the Stanford-Binet IV, Peabody Picture Vocabulary Test-Revised, and Columbia Mental Maturity Scale

This study investigated the correlations among the Stanford-Binet Intelligence Scale (4th Edition), the Peabody Picture Vocabulary Test-Revised, and the Columbia Mental Maturity Scale. A kindergarten class of 21 children (10 boys, 11 girls) took each test, and the intercorrelations of the total scores and the correlation between the Binet IV Vocabulary subtest and the Peabody were calculated. Only the relations between Binet IV and Peabody and between Binet IV Vocabulary and Peabody were found to be statistically reliable.

Validity of five MMPI alcoholism scales

Assessed five alcoholism scales derived from the MMPI for their ability to classify accurately two groups of alcoholics (N = 120) and a group of nonalcoholic psychiatric patients (N = 60) in an inpatient treatment setting. The alcoholic group was divided into those who were self-committed and those who were court-committed. Depending on how the data were treated, only one, or none of the scales was accurate.

Zung Self-Rating Depression Scale Scores of Psychiatric Outpatients by Age and Sex

This study was designed to estimate the correlations of the Zung Self-rating Depression Scale with age and sex in a psychiatric outpatient sample. 378 male and 671 female clients of 13 to 89 yr. of age completed the scale as part of their admission to an outpatient psychiatric clinic. The scores were classified according to sex and age (in 10-yr. groups). Analysis showed some statistically significant differences according to age and sex, but the magnitude of those differences was small. Comparison of these data with a previously published set of data from a nonpsychiatric sample showed, as would be expected, the psychiatric group scored consistently higher.