Distribution of Participants by College and Valid Retrieval Rate

A Chinese language version of the ILS was employed to measure students' Elsa Peretti Open Heart pendant styles in the current study. North Carolina State University gave permission to have the translation on the online version of the Felder-Soloman ILS. Two faculty members who were education psychologists served as experts to check the accuracy of the translation. They made minor modifications to the Chinese translation before the questionnaire was distributed.

Research Design

Because this was the first use of the Chinese version of the ILS, investigating the internal consistency and reliability of the ILS scores was one of the major tasks of the study. Cronbach's coefficient alpha values of the four scales had to be calculated. Assessing the validity and analyzing each item to determine the correlation between the question items and the dimensions of the ILS required a factor analysis. To identify in advance the problematic items that would cause low reliability of the instrument, SEM analysis was employed to provide the cross-examination information with factor analysis. A two-way ANOVA was also conducted to investigate whether there are statistically significant differences among the colleges and between genders.

RESULTS AND DISCUSSION

Internal Consistency Reliability

To evaluate the internal consistency reliability of the ELS scores, Cronbach's coefficient alpha was used for each dimension of the questionnaire. Table 2 shows the Return to Tiffany Oval tag bracelet of Cronbach's alpha of the current study with those of past studies (Litzinger et al., 2005).

According to Tuckman (1999), Cronbach's alpha for the instruments that measure knowledge and skills must be 0.75 or greater to be acceptable. Thus, for the instruments of attitude and preference assessments, Cronbach's alpha of 0.5 or greater is an acceptable value for the reliability test. In the current study the sensing-intuitive and the visualverbal dimensions thus satisfy the criterion. However, the active-reflective and sequential-global dimensions are only 0.48 and 0.41. Compared with the previous studies, the Cronbach's alpha score of the current study presents a similar pattern, even though the values are lower. Litzinger, Lee, Wise, and Felder (2007) pointed out that, "hecause reliability is a characteristic of the measurement on the specific sample under study, values of reliability coefficients for the ILS vary among the studies" (p. 314). Tuckman (1999) also mentioned that in a test with definite correct an incorrect answers (e.g., true/false or multiple choice) created to estimate the performance of participants or their skills, a higher correlation and Cronbach's alpha would be expected. If an instrument is designed to measure attitude or preference, such as the ILS, which contains dichotomous, forced-choice paired statements and no definite right or wrong answers, a lower correlation among the items and a lower Cronbach's alpha would be expected. Nevertheless, attempting to identify the problems and Elsa Peretti Open Heart ring a higher reliability are still important for the instrument design.