Analysis On Construct Validity Of Five-Pattern Personality Inventory

Purpose: Five-Pattern Personality Inventory(FPPI) is a localized personality test,according to TCM theory. The purpose of this study is simulating construct validity ofFive-Pattern Personality Inventory used structural equation model. Based on the theory offive state of Yin and Yang of traditional Chinese medicine theory, the model is set up. Bymeans of the model parameter estimation, model evaluation and revision, find a modelfitting traditional Chinese medicine theory, and promote the development of Five-PatternPersonality Inventory.Methods: According to the method of psychological test validity analysis, This studyuse the structural equation model to analysis construct validity and verify the theoreticalconstruct validity. Sample data source for the2004ministry of science and technologyproject: people’s personality and physique investigation in our country. The subject of thesample selection process is based on China’s2000census data which is announced by thecensus office of the state council and the national bureau of statistics and the statisticsdivision of science and technology, population and society. combining with the clustersampling and stratified random sampling principle, we draw some of the age of urbanpopulation over the age of18. Considering the actual research needs, this study onlyextract all of Beijing area effective samples which is total of430as the data source ofstructural equation model, and use MPLUS computer software to implement this process.Results: The degree of fitting of sample data and the structural equation model(measurement model) which built according to TCM theory is as follows: absolute fitindex of model-RMSEA value is0.03, shows that model fitting is very good. Because atthe time of the scaling, as well as the revision of the norm in2008, we assessed the reliability of Five-Pattern Personality Inventory. The results proved that this inventory hadhigh reliability. The RMSEA value is ideal. This state from one aspect shows theFive-Pattern Personality Inventory has good construction validity. But relative fit index ofmodel such as CFI and TLI are less than0.8, from another side, illustrates the test modelneeded to be modify.According to the model fitting, we proceed model modification. Then a revisedsimple model is coming out and the path of observed variables and latent variables is less.Correction model results: absolute fit of index RMSEA value is0.034, represented modelfitting perspective beteen the sample data and assumptions, and the model fitting is verygood. Relative fit of index values were higher than0.8. The CFI value is0.818, and TLIvalue is0.807.This result stand for that the revised simple model is acceptable.Conclusion: First, Five-Pattern Personality Inventory has good construct validity.According to the results of fitting, there is a "five-pattern”. Even thousands of years passedaway, the ancient experience to classify people can still be the personality classificationbasis in today’s society. Theory of the Five-pattern Personality Inventory is based on thecontent of Yin and Yang in the five states of different people. In this category of Yin andYang, five-pattern who represents a state, they obviously have relationship, and it is notsimple linear correlation. In the light of the thinking mode of Chinese people to know theworld, Yin and Yang is a Tai Ji ball. Five-pattern people, which is obviously dividedaccording to different levels of Yin and Yang also should be solid. It is probability thateach pattern people contains not just a dimension or single a trait, that may have some orcertain characteristics more prominent and can be observed or feeling obviously by people.the observed or feel character is reflected in the Five-Pattern Personality Inventory.Second, Five-Pattern Personality Inventory has a space for further improvement anddevelopment. On the one hand, the category of Yin and Yang is very wide,all-encompassing, and clearly the FPPI item can not contain all of the content. On theother hand, on the basis of the result of fitting, the degree of fitting of the model can beincreased by removing or modifying some items. Through this step, we can delete sometopic like the inefficiency and stand out the highlight part which is the high efficiency, soas to improving the efficiency of the test. In this process, of course, should give attention toboth high efficiency and accuracy. Trough the comprehensive consideration, when itemsare removed we should retain representative characteristics of the five-pattern people as faras possible. In the simple revised model, the latent variable Tai Yang, kept five projects; Shao Yang kept nine projects; Yin and Yang He Ping retained seven projects; ShaoYinretained seven projects; Tai Yin kept eighteen items. Compared items with theFive-Pattern personality characteristics, we found that they are consistent.