Top4.2 Reliability Test And Validity Test
According to Churchill’s (1979) paradigm for construct measurement, after defining the domain of values systems and the leadership styles measurement constructs and developing a measurement instrument, we need to test the reliability and validity of the constructs. The values orientation can be used to measure an individual’s personal behaviors and can be worked as the benchmark for leadership profile; while the leadership style of an individual is the real behaviors one has expressed on the post of being a leader in different contexts. In our study, we want to test if there are cultural influences (values orientation) on the leadership styles. if it meets the requirements of reliability and validity (Chen and Paulraj, 2004). Reliability reflects the consistency or repeatability of measurement, which is a necessary but not sufficiency condition for validity. Validity is concerned with the extent to which a specific set of items reflects a content domain (DeVellis, 1991).
4.2.1 Reliability Test
To evaluate the internal consistency of the scales for reliability, we use the internal consistency method that is estimated using Cronbach’s alpha and item-to-total analysis. Typically, reliability coefficients of 0.70 or higher and the item loading on the factors of 0.60 or higher are considered adequate (Hull and Nie, 1981).
For each of the items in this research, we calculated Cronbach reliability coefficient (alpha), and the item-to-total correlations for every item. As can be seen from Table 4 of Chapter 3, all Cronbach’s alpha values of the factors are well above the critical value and ranged from 0.65 to 0.78, meanwhile the item loading on the factors are all acceptable. These results confirm the proposed research models meet the requirements of the internal consistency of the scales for reliability.