How to check a computer program

The usual ways of checking computer programs have important flaws. A better way to check a program you have written is to have someone else write a program that does the same thing and then compare the output of the two programs over a wide range of input. Two examples are described.     

FACTOR: A computer program to fit the exploratory factor analysis model

Exploratory factor analysis (EFA) is one of the most widely used statistical procedures in psychological research. It is a classic technique, but statistical research into EFA is still quite active, and various new developments and methods have been presented in recent years. The authors of the most popular statistical packages, however, do not seem very interested in incorporating these new advances. We present the program FACTOR, which was designed as a general, user-friendly program for computing EFA. It implements traditional procedures and indices and incorporates the benefits of some more recent developments. Two of the traditional procedures implemented are polychoric correlations and parallel analysis, the latter of which is considered to be one of the best methods for determining the number of factors or components to be retained. Good examples of the most recent developments implemented in our program are (1) minimum rank factor analysis, which is the only factor method that allows one to compute the proportion of variance explained by each factor, and (2) the simplimax rotation method, which has proved to be the most powerful rotation method available. Of these methods, only polychoric correlations are available in some commercial programs. A copy of the software, a demo, and a short manual can be obtained free of charge from the first author.     

Evaluation of a group program designed to increase androgyny in feminine women

The effectiveness of a group treatment program designed to increase androgyny in feminine women was examined. The full treatment (FT) incorporated the principles of assertiveness training and consciousness raising, and was compared to a discussion-orientated alternate treatment (DO) and a wait-list control (WL). Fifty-one female subjects who scored feminine on the Bem Sex-Role Inventory participated in this study. Pre- and posttreatment measures of sex-role orientation, assertiveness, and the attainment of individual goals were obtained. Two-month and 1-year follow-up data were also collected. At the posttest, the FT group was more assertive than the DO and WL groups, and higher than the WL group in the attainment of individual goals. No posttest differences were found among the groups on sex-role orientation. These findings were reversed by the 2-month follow-up such that the FT was more androgynous and more masculine than the WL group, while the differences between groups in assertiveness and personal goal attainment had dissipated. The 1-year follow-up revealed the FT group to be superior to the WL group in androgyny, masculinity, and assertiveness.Appreciation is extended to N. J. Chehayeb, Brenda Barker, and Jimmie Eaker for their assistance in the data collection for this study. The study is based in part on the first author's doctoral dissertation submitted to Southern Illinois University under the supervision of the third author.     

Intention stability assessed using residual change scores moderates the intention-behaviour association: a prospective cohort study

Intention stability is considered to be one of the key pre-requisites for a strong association between intention and behaviour. It has been claimed, however, that studies examining the moderating impact of intention stability may be invalid, as they have relied on statistically inferior methods. Residual change scores have been suggested as a more appropriate method of measuring change (or lack thereof) in constructs. The aim of the current study, therefore, is to test whether intention stability, calculated using residual change scores, moderates the intention-physical activity behaviour association. A total of 163 participants (124 women, 39 men) completed questionnaires online at three time points separated by 14 day intervals. The moderating impact of intention stability was assessed using multiple linear regression followed up using simple slope analyses to identify the direction of any effect. The interaction of intention and intention stability was found to significantly improve the overall model fit. Intentions had a stronger positive association with behaviour when intentions were more stable than when they were more unstable. However, sensitivity analyses revealed that the association was not robust and reduced to non-significant with the removal of potential multivariate outliers. Future research should use residual change scores as the preferred method of assessing intention stability.     

A computer program to generate and measure random forms

A computer program is described that generates random form stimuli by a method parallel to that described by Attneave & Arnoult (1956) for construction forms with both angles and arcs in their perimeters. The program also performs a physical analysis of the forms by computing the values of a large number of physical variables describing each form. Already existing forms also may be analyzed by the program. Thus, it has the capability to perform physical analyses of approximations to natural forms.     

A computer program to find the best-fitting orthogonal factors for a given hypothesis

A modification of the quartimax computation for factor rotation is described in which a hypothesized factor pattern is given to the machine along with the data. The machine uses the pattern to select the subset of variables to which it will attend when rotating in a given plane, in order to find an orthogonal solution which closely fits the hypothesis. The program also provides a measure of the goodness of this fit. The program can utilize pattern matrices that reflect only partial hypotheses as to the nature of the factors, as well as those that specify highly determined simple structure.     

A ToolBook computer program to develop course objectives and assessment measures

College courses are generally developed because of current fashion in the field and/or student and faculty interest in some topic without reference to departmental or programmatic goals. Current calls and needs for accountability in higher education demand a more organized approach. A computer program, ToolBook 1.53, which provides information to the instructor about systematic development of courses and assessment of student outcomes, is described. The program supports the construction of outcome-oriented learning objectives and corresponding assessment techniques. Instructors who used this program to learn about how to construct such objectives showed an increase in knowledge and evaluated the program positively. Using the information provided in this program leads to an integrated approach to course design and development in which course objectives are linked to the college mission statement through departmental or program objectives.     

A survey of residual analysis and a new test of residual trend

A survey of residual analysis in behavior‐analytic research reveals that existing methods are problematic in one way or another. A new test for residual trends is proposed that avoids the problematic features of the existing methods. It entails fitting cubic polynomials to sets of residuals and comparing their effect sizes to those that would be expected if the sets of residuals were random. To this end, sampling distributions of effect sizes for fits of a cubic polynomial to random data were obtained by generating sets of random standardized residuals of various sizes, n. A cubic polynomial was then fitted to each set of residuals and its effect size was calculated. This yielded a sampling distribution of effect sizes for each n. To test for a residual trend in experimental data, the median effect size of cubic‐polynomial fits to sets of experimental residuals can be compared to the median of the corresponding sampling distribution of effect sizes for random residuals using a sign test. An example from the literature, which entailed comparing mathematical and computational models of continuous choice, is used to illustrate the utility of the test.     

Using temporal information to construct, update, and retrieve situation models of narratives

Four experiments explored how readers use temporal information to construct and update situation models and retrieve them from memory. In Experiment 1, readers spontaneously constructed temporal and spatial situation models of single sentences. In Experiment 2, temporal inconsistencies caused problems in updating situation models similar to those observed previously for other dimensions of situation models. In Experiment 3, merely implied temporal order information was inferred from narratives, affecting comprehension of later sentences like explicitly stated order information. Moreover, inconsistent temporal order information prevented the creation and storage in memory of an integrated situation model. In Experiment 4, a temporal inconsistency increased processing time even if readers were unable to report the inconsistency. These results confirm the significance of the temporal dimension of situation models.