【正文】
ear work role, and happy teaching environment). Facet measures target on specific aspects of their jobs. Most researchers adopt the facet approach to measure satisfaction (., Job Descriptive Index [JDI。 Smith, Kendall,amp。Hulin, 1969], Minnesota Satisfaction Questionnaire, and Warr’s Job Satisfaction Scale [WJSS。Warr, Cook,amp。Wall, 1979]) because not only can the domains of contributing to teaching satisfaction be found, but the overall level of satisfaction is obtained through simply adding up all the domain scores. However, the facet approach has several methodological problems. Problems of the Facet Approach First, some researchers assume that the whole is equal to the sum of its parts. An example is the JDI, which sums scores of five subscales despite serious problems with this procedure. In fact, the JDI scale was not designed to be aggregated across the five dimensions. The JDI was constructed to measure five distinctively different areas. Furthermore, these five dimensions are only moderately correlated (.25 to .45。 Smith, Kendall, amp。 Hulin, 1969) and represent at least five factors. Improper aggregation would distort the findings and interpretations eventually. Second, these facet scales usually involve too many items for practical use. For example, the JDI is a 72item modified adjective checklist. The Minnesota Satisfaction Questionnaire (Weiss, Dawiss, England, amp。 Lofquist, 1967) is a 100item instrument measuring 20 facets of job satisfaction. A respondent needs to plete 100 items for a single measurement of satisfaction. They are timeconsuming in survey Administration. Although the WJSS,