Paul E. Yeatts

Moving Beyond Univariate Post-Hoc Testing in Exercise Science: A Primer on Descriptive Discriminate Analysis

There has been a recent call to improve data reporting in kinesiology journals, including the appropriate use of univariate and multivariate analysis techniques. For example, a multivariate analysis of variance (MANOVA) with univariate post hocs and a Bonferroni correction is frequently used to investigate group differences on multiple dependent variables. However, this univariate approach decreases power, increases the risk for Type 1 error, and contradicts the rationale for conducting multivariate tests in the first place. Purpose: The purpose of this study was to provide a user-friendly primer on conducting descriptive discriminant analysis (DDA), which is a post-hoc strategy to MANOVA that takes into account the complex relationships among multiple dependent variables. Method: A real-world example using the Statistical Package for the Social Sciences syntax and data from 1,095 middle school students on their body composition and body image are provided to explain and interpret the results from DDA. Results: While univariate post hocs increased the risk for Type 1 error to 76%, the DDA identified which dependent variables contributed to group differences and which groups were different from each other. For example, students in the very lean and Healthy Fitness Zone categories for body mass index experienced less pressure to lose weight, more satisfaction with their body, and higher physical self-concept than the Needs Improvement Zone groups. However, perceived pressure to gain weight did not contribute to group differences because it was a suppressor variable. Conclusion: Researchers are encouraged to use DDA when investigating group differences on multiple correlated dependent variables to determine which variables contributed to group differences.

Shared Decision-Making in Nursing Homes Factors Associated With the Empowerment of Direct Care Workers

Objective: The advantages of empowering direct care workers (DCWs) within nursing homes (NHs) are well documented. Our objective is to identify factors that create DCW empowerment as this has not received adequate attention. Method: The data come from a larger study focused on the empowerment of DCWs in NHs. A self-administered instrument was completed by 372 DCWs within 11 NHs in the north Texas region. Cluster analysis and ordinary least squares (OLS) regression were performed. Results: Feedback from nurse management to DCWs was positively associated with shared decision-making (SDM) and ranked highest followed by information exchange, trust in management, and wages linked to performance. Discussion: SDM may be enhanced where nurse management shares relevant information with the DCWs, listens to their ideas, provides explanations when DCW suggestions are not used, and does so in a supportive environment. Organizational characteristics of importance include linking wages to DCW performance and providing an accessible training program.

The Use of Structure Coefficients to Address Multicollinearity in Sport and Exercise Science

ABSTRACT A common practice in general linear model (GLM) analyses is to interpret regression coefficients (e.g., standardized β weights) as indicators of variable importance. However, focusing solely on standardized beta weights may provide limited or erroneous information. For example, β weights become increasingly unreliable when predictor variables are correlated, which is often the case in the social sciences. To address this issue, structure coefficients, which are simply the bivariate correlation between a predictor and the synthetic Ŷ variable, should also be interpreted. By examining β weights and structure coefficients in conjunction, the predictive worth of each independent variable can be more accurately judged. Despite this benefit, researchers in the field of sport and exercise science have rarely reported structure coefficients when conducting multiple regression analysis. Thus, the purpose of the present article is to discuss problems associated with the sole interpretation of β weights and to demonstrate how structure coefficients can be incorporated to improve accuracy of interpretation. Additionally, a content analysis was conducted to examine current trends in reporting multiple regression results within sport and exercise science research.

Weight Control Behavior as an Indicator of Adolescent Psychological Well-Being.

BACKGROUND Adolescence is a critical time for the development of psychological well-being. Weight gain and the emergence of body image concerns during this period can lead to the development of negative psychological states. To explore this issue, we examined the relationship between weight control behavior (WCB; ie, trying to lose, gain, stay the same, or do nothing about weight) and levels of depression and self-esteem. METHODS Adolescents (508 boys, 502 girls; Mage  = 12.32 ± .88 years) completed a survey that assessed WCB, depression, and self-esteem. Descriptive discriminant analysis was used to analyze WCB group differences on psychological well-being. Multivariate post hoc analysis further examined group differences. Structure coefficients indicated the relative importance of each dependent variable in boys and girls. RESULTS Results indicated that, among both sexes, WCB was significantly related to depression and self-esteem. Individuals trying to lose weight had lower levels of psychological well-being than the other groups. CONCLUSIONS Adolescents trying to lose weight reported the lowest psychological well-being scores whereas those not doing anything to control weight reported the highest levels of psychological well-being. These findings have important implications for screening and education programs designed to monitor and support adolescent psychological well-being.