James D. Linden

A Factor Analysis of Teacher Trainee Responses to Selected Personality Inventories

THE &dquo;competent teacher&dquo; has recently been defined by the National Commission on Teacher Education and Professional Standards of the National Education Association (1955) as an intelligent, socially adequate, personally desirable, and professionally able individual. Of these four, the personally desirable qualities are of primary concern in the present study. The personally desirable teacher is said to be competent if he likes to work with children, demonstrates a warm outgoing personality, inspires emulation by children, and is emotionally mature, physically fit, patient, and understanding. That it is generally agreed by individuals engaged in the professional preparation of teachers that certain personality variables are important prerequisites for success as a teacher is seen in a recent study by Stout (1957). &dquo;Emotional stability&dquo; was ranked as most important among nine suggested criteria used for the selection and retention of teacher candidates. While there might not be any disagreement with the rating given this criterion, the problems involved in securing valid measures of such a personal characteristic are extremely complex. In the attempt to identify prospective teacher candidates who exhibit personality traits felt to be desirable for success in teaching,

Use of Multiple Discriminant Function Analysis in the Evaluation of Three Multivariate Grouping Techniques

eminently confusing. Either valuable information is lost by disregarding the multidimensionality or interrelatedness of a category of variables, or misclassifications occur, allowing individuals with distinctly different variate profiles into the same class (Goldstein and Linden, 1967). Assignment of individuals to such groups on the basis of simultaneous consideration of two or more variates requires that attention be given to two general problems : (1) selection of a measure of profile similarity which adequately represents the information originally generated, and (2) selection of a gouping technique which reflects with most precision the real and distinguishing differences between subjects.

A Critique of the Hand Test

THE problem of predicting overt aggressive behavior in the prison inmate, the hospitalized patient, and other institutionalized individuals has been investigated in several studies (Finney, 1955; Gluck, 1955; Rust, 1948). The results have been, in general, discouraging. This thorny problem has recently received new attention in conjunction with the publication of a semistructured projective technique, the Hand Test (Wagner, 1962). The authors of this instrument maintain that they have developed a test which predicts this behavior in the individual case and they attempt to support this claim by citing numerous research studies which are summarized in a recent monograph (Bricklin, Piotrowski & Wagner, 1962). A cursory perusal of this research reveals that these claims

SOCIAL DESIRABILITY, ADJUSTMENT AND EFFECTIVENESS

Measures of social desirability were correlated with measures of adjustment and effectiveness which were methodologically independent, i.e., structured personality tests. Different relationships were found in the two groups. Responding in a socially desirable manner seems to be complexly related to positive psychological health.

A 7094 Fortran Program for the Computation of Tetrachoric Correlations1

A variety of computer programs in use today depends upon correlation matrices for the solution of problems. Many of the original data utilized in the construction of these matrices are in dichotomous format. Because of the unavailability of a program for the computation of the tetrachoric correlation, Pearson product-moment correlations generally have been employed which provide spurious estimates of relationships among the data involved. AR a consequence of the need for a program that would compute tetrachoric correlations to provide more appropriate item reliability information and appropriate input for factor analytic work, the authors developed the program reported here.

Computer Procedures for Determining Significance With Range Tests Over Differences Between Means

a significant F value presents many experimenters with a distinct problem in procedure. The educational and psychological literature is replete with cases of multiple t-tests violating all assumptions of independence underlying this statistic. Winer (1962) has shown that as the number of t-tests increases, the collective, level of significance also increases as: 1 (1 a)&dquo; where a is the level of signifi-

7094 Template Programs for Scoring Dichotomous Response Format Tests with Special Reference to the Minnesota Multiphasic Personality Inventory

SAVINGS in time and effort have stimulated the development of many procedures for the electronic processing of tests and questionnaires (Shuman, 1966). From the earliest mark sensing devices which simply counted responses, scoring programs have appeared with increasing regularity. That many programs have been specific to a given test or one user’s system has made the employment of the procedure by others difficult, if not impossible. Users of popular tests such as the Strong Vocational Interest Blank (SVIB) or the Minesota Multiphasic Personality Inventory (MMPI) are well