Open Access Dissertation
Date of Award
James L. Moseley
THE EFFECTS OF SELF DIRECTED TEAMS IN AN AUTOMOTIVE MANUFACTURING ENVIRONMENT
DAVID W. SHALL
Advisor: James L. Moseley, EdD, LPC, CHES, CPT
Major: Instructional Technology
Degree: Doctor of Philosophy
This study compares self-directed work structures to more traditional supervised work structures in order to determine if the expenditures and efforts required to implement self-directed work teams are warranted. Multiple internal performance metrics are examined in comparing plant work structures in various degrees of implementation between traditional work structures and self-directed work teams. The researcher collected data from multiple organizations within Ford Motor Company and four participating North American Ford production plants. Two Ford assembly plants and two Ford engine manufacturing plants were researched. Performance data from the 2004 production year were examined in each facility. Both assembly plants built the same Ford F-150 pick-up truck and both engine manufacturing plants produced the same V-6 engine in 2004. Data were collected to answer several questions including: 1) Does the presence of effectively rated self-directed work teams affect injury frequency; 2) Does the presence of effectively rated self-directed work teams affect injury severity; 3) Does the presence of effectively rated self-directed work teams affect unexcused absenteeism; 4) Does the presence of effectively rated self-directed work teams affect productivity; 5) Does the presence of effectively rated self-directed work teams affect cost performance; 6) Does the presence of effectively rated self-directed work teams affect external quality and customer satisfaction; 7) Does the presence of effectively rated self-directed work teams affect internal engine manufacturing quality; 8) Are Safety LTR, Safety SV, AWOL, Productivity, and Cost statistically significant predictors of customer satisfaction and, 9) Are Safety LTR, Safety SV, AWOL, Productivity, and Cost statistically significant predictors of work team effectiveness.
By comparing the performance metrics and customer satisfaction data between like plants with separate and different work structures, the researcher isolated the impact that work structures have on safety, cost, productivity, quality and employee morale. The hypothesis in this research suggests that significant performance differences exist between effectively rated self-directed work teams and more traditionally supervised work groups in automotive assembly and engine manufacturing plants. Furthermore the hypothesis suggests that dependent performance variables predict customer satisfaction and work team efficiency.
Several statistical procedures were used to answer the nine research questions which ranged from basic to theoretically experimental procedures. First, causal comparisons were drawn between plants with effectively rated self-directed work teams and plants with more traditionally supervised work structures to explore the relationship that the dependent performance metrics have with the independent work structures. Multivariate analysis of covariance was used to simultaneously test correlation between two independent predictor variables and several dependent variables. Second, a Hybrid Structural Equation Model (SEM) was utilized to further test and predict relationships between dependent and independent variables, but also within the dependent performance metrics. The technique allowed confirmatory and exploratory modeling to reveal the magnitude of performance variable interrelationships and predict their potential impact on customer satisfaction and work group efficiency. Statistical techniques increasingly dissected data with the goal of answering each research question with error-free statistical results.
Many inferences can be made from the analysis of descriptive statistics in this research, most of which indicate favorable performance results in plants with effective self-directed work teams over plants with more traditional work forces. The basic assumptions are challenged statistically with multivariate test of covariance, univariate tests, pair-wise comparisons, test of moderation, Z-tests and a hybrid structural equation model.
Pair-wise comparisons reveal five significant results in truck assembly plants. Effectively rated self-directed teams in Norfolk significantly outperformed their more traditionally supervised rivals in Kansas City in lost time case rate, severity rate and controllable employee absence. Furthermore, all of the effects are positive in nature and justify the effort required to implement self directed teams. Oppositely, in engine manufacturing plants, the more traditional workforce in Cleveland outperformed effectively rated self directed teams in Lima in terms of cost and customer satisfaction. Both findings were statistically significant and demonstrate adverse effects since improvements in work team effectiveness resulted in higher costs and lower customer satisfaction.
Tests of moderation and subsequent Z tests for truck assembly plants support four significant findings. In Kansas City work team effectiveness had explanatory power for lost time case rate and severity rate although the predictive nature of work team effectiveness on lost time case rate and severity rate are adverse since both rates increased. Z tests reveal significant differences in the regression lines for employee absenteeism and customer satisfaction. Results for absenteeism show mixed predictions where the traditional workforce in Kansas City experience favorable reductions in absence while self-directed work teams in Norfolk experience increased absence as work team effectiveness improved. The Z test for customer satisfaction reveal promise for self-directed work teams in both truck assembly plants since quality defects decrease as work team effectiveness improves.
Tests of moderation and subsequent Z tests for engine manufacturing plants support four significant findings. In Cleveland work team effectiveness demonstrates explanatory power for severity rate, cost and engine manufacturing quality. Work team effectiveness demonstrates positive predictive power over severity rate and engine manufacturing quality since injury severity and quality defects decrease as work team effectiveness improves. Conversely, cost predictably increases as work team effectiveness improves. Z tests revealed significant differences in the regression lines for employee absenteeism and engine manufacturing quality. Absenteeism results display mixed predictions where the traditional workforce in Cleveland anticipate an unfavorable increase in absence while self-directed work teams in Lima anticipate absence reductions as work team effectiveness improves. The Z test for engine manufacturing quality flaunted positive predictions for self-directed work teams in both engine manufacturing plants. As work team efficiency improves, engine quality defects are minimized.
The two final research questions asked if the dependent performance variables in the study were statistically significant predictors of customer satisfaction and work team effectiveness. Beta Coefficients from the Hybrid Structural Equation Model estimated that three variables influenced performance including safety lost time case rate, safety severity rate and productivity. The multivariable interaction of these dependent variables resulted in a statistical prediction that positive internal performance affects customer satisfaction but not work team effectiveness ratings.
This work adds relevant research findings to the body of literature in human performance improvement and instructional technology. Individuals contemplating an intervention involving teams or a work structure change are well served using this dissertation as a resource. To the extent possible the research follows Ford Motor Company's path along the human performance technology (HPT) model (Van Tiem, Moseley, Dessinger, 2004) that is endorsed by the International Society for Performance Improvement.
Shall, David Wayne, "The Effects Of Self-Directed Teams In An Automotive Manufacturing Environment" (2010). Wayne State University Dissertations. 114.