Charles F. Seifert

Effects of multisource feedback and a feedback facilitator on the influence behavior of managers toward subordinates.

The authors compared a feedback workshop with both a no-feedback control group and a comparison group of managers who received a feedback report but no feedback workshop. The multisource feedback was based on ratings of a managers influence behavior by subordinates, peers, and bosses. Managers in the feedback workshop increased their use of some core influence tactics with subordinates, whereas there was no change in behavior for the control group or for the comparison group. The feedback was perceived to be more useful by managers who received it in a workshop with a facilitator than by managers who received only a printed feedback report.

Using Coaching to Enhance the Effects of Behavioral Feedback to Managers

A field experiment was conducted to assess whether coaching would enhance the effectiveness of a feedback workshop for store managers in a regional supermarket chain. The experimental group of managers received individual coaching several weeks after attending a feedback workshop. The control group of managers also attended a feedback workshop but did not receive the follow-up coaching. Each manager’s use of proactive influence tactics was rated by subordinates (department managers) before the interventions and several months afterward. Coaching significantly increased the use of collaboration with subordinates, but results for the other three “core” tactics were mixed. Likely reasons for the lack of stronger results are identified, and implications for practice are discussed.

The Effectiveness of Methods for Analyzing Multivariate Factorial Data

A Monte Carlo simulation was used to examine the effectiveness of univariate analysis of variance (ANOVA), multivariate analysis of variance (MANOVA), and multiple indicator structural equation (MISE) modeling to analyze data from multivariate factorial designs. The MISE method yielded downwardly biased standard errors for the univariate parameter estimates in the small sample size conditions. In the large sample size data conditions, the MISE method outperformed MANOVA and ANOVA when the covariate accounted for variation in the dependent variable and variables were unreliable. With multivariate statistical tests, MANOVA outperformed the MISE method in the Type I error conditions and the MISE method outperformed MANOVA in the Type II error conditions. The Bonferroni methods were overly conservative in controlling Type I error rates for univariate tests, but a modified Bonferroni method had higher statistical power than the Bonferroni method. Both the Bonferroni and modified methods adequately controlled multivariate Type I error rates.

Book Review: Harlow, L. L. (2005). The Essence of Multivariate Thinking: Basic Themes and Methods. Mahwah, NJ: Lawrence Erlbaum

The Essence of Multivariate Thinking is one of the books in the Multivariate Applications Series sponsored by the Society of Multivariate Experimental Psychology. In this publication, Lisa L. Harlow’s stated objective is to ‘‘make the topic of multivariate statistics more accessible to and comprehensible to a wide audience.’’ I believe she easily exceeds her objective. I found the layout of the book to be both logical and sequential. Before specifically discussing the various analytical methods covered in the book, three chapters are dedicated to covering some of the basic information required to understand multivariate methods. The book begins with a general overview of multivariate methods (including strengths and weaknesses), an introductory framework, and a frank discussion of the advantages and disadvantages of multivariate analyses. The second chapter discusses some of the main overarching themes associated with multivariate methods. Particular attention is placed on the multiplicity (e.g., examining multiple theories, constructs, measures, samples, and time points) associated with multivariate methods. The third chapter then provides an overview of several themes directly related to multivariate statistics, including data characteristics, measurement scales, roles of variables, incomplete information, missing data, descriptive statistics, and inferential statistics. Coverage of the topics is relatively brief and provides the basic information. In addition, where the degree of discussion is limited by the scope of the book (e.g., incomplete information and missing data), citations are provided for interested readers. Separate chapters are then devoted to each of the following multivariate methods: (1) multiple regression (MR); (2) analysis of covariance (ANCOVA); (3) multivariate analysis of variance (MANOVA); (4) discriminant function analysis (DFA); (5) logistic regression (LR); (6) canonical correlation (CC); and (7) principal components (PCA) and factor analysis (FA). A chapter on matrices and multivariate methods is also covered before discussing the multivariate grouping methods. Although the information presented in the chapter was direct and easy to follow, I am not sure it adds much value to the text. Also, besides bringing on some psychometric flashbacks, I don’t believe the information in the chapter was detailed enough to gain a true understanding of some of the complexities associated with matrix algebra. The chapter may also be intimidating for the intended audience (e.g., first-year graduate students and advanced undergraduates) and discourage them from reading the rest of the book. It may have been better to include the information in an appendix. That way, interested readers could explore the topic in more detail. When discussing each of the multivariate methods, the book consistently addresses the same 10-question framework. When reading the entire book, the 10-question framework does lead to some redundancies, however, the consistent format is beneficial if you are interested in exploring only one or a few of the different methods. The questions covered for each multivariate method include: