Hitachi: Correlation and Causal Analysis Research
Hitachi Automotive Products, Los Angeles (HAPLA), based in Torrance, California, is a major manufacturer of remanufactured (remanned) electromechanical automotive parts such as alternators, starters, distributors, electronic control units, mass airflow sensors and linear flow valves. At HAP-LA, the production process and philosophy is greatly influenced by Japanese techniques such as Kaizen, 5S and Kanban. Sometimes, when changes occur in the production process after a Kaizen blitz, or restructuring, events occur that must be analyzed using correlation and causal analysis methodologies
Correlation and Causal Analysis Research
Before an analysis can be conducted, an understanding of both correlational research and causal analysis research must be met. A correlational analysis is one that determines that "variables occur together in some specified manner without implying that one causes the other" (Cooper, 2002, p. 44). An example of a correlational relationship is the determination that injuries occur in a factory after weekends or holidays.
In order to test the correlational relationship or hypothesis, the variables are tested and determined if they have either a positive or negative relationship. A positive relationship is one is which the variables increase and decrease at the same time (Wu, 2002a). The alternative is a negative correlational relationship, in which one variable increases while the other variable decreases. A correlation coefficient is a measure of strength and direction of the correlation between +1.00 and -1.00 (Wu, 2002a). A study of the variables should provide three different results; no relationship, a positive relationship or a negative relationship (Wu, 2002b).
Unlike correlation studies, which do not take into account the causal relationship between which variable is the cause and which is the result (Wu, 2002a), a causal research analysis determines if there is a measurable cause and effect relationship between the variables. In order to determine...