Adaptation of Web Pedagogical Content Knowledge Survey to Turkish
The aim of the study is adaptation of the web pedagogical content knowledge survey which was developed by Lee & Tsai (2008 to Turkish. In order to adapt the scale, firstly scale items were translated to Turkish by the researcher. Then translation form was further developed by consulting ten specialists. English and Turkish forms of scale were both filled by 30 undergraduate students with two-week intervals respectively. Two forms were accepted as equal since the correlation between them was
... etween them was 0.87. Turkish version of the scale was implemented on 232 students for validity and reliability studies. The scale has five factors, which was found as result of the exploratory and confirmatory factor analyses. In Turkish version of the scale alpha value was calculated as .94. Therefore, it can be concluded that Turkish version of the scale was reliable and valid. Key Words: web pedagogical content knowledge scale, web pedagogical content knowledge, pedagogical content knowledge SUMMARY Purpose and significance: When the literature is examined, web pedagogical content knowledge survey, developed by Lee and Tsai (2010) , is seen to be the most prominent scale developed to measure teachers' self-efficacy beliefs regarding the use of web. From this aspect, since there is not a scale in Turkish in this area, it is important to bring a scale into the field measuring teachers' selfefficacy beliefs toward web-based teaching. For this reason, this study aims to adapt scale into Turkish. Methods: The purpose of this study is to adapt web pedagogical content knowledge survey into Turkish. The sample of the study was 232 students from CEIT department of Sakarya University who were 2nd, 3rd and 4th year students during the spring semester of 2009-2010 academic years. After taking necessary permissions from Tsai, for the adaptation study of the scale, via e-mail, the adaptation study of the scale was started. Exploratory factor analysis (EFA) and confirmatory factor analysis (CFA) was performed to examine the original scale's structure approved by Turkish experts in Turkish culture. The Cronbach Alpha and the test-retest were calculated. Results: The scale was translated to Turkish by the researcher. After translation, the scale was prepared as a structure including original item, translated item and proposed form and given to experts for their opinion. After making all necessary adjustments, linguistic equivalence study of the scale was conducted. English and Turkish forms of scale were both filled by 30 undergraduate students with two-week intervals respectively. Two forms were accepted as equal since the correlation between them was 0.87. EFA was conducted and it was found that five-factor structure of the original scale was preserved when it was administered to Turkish students. The results of EFA also showed that eigen value of the Turkish form of the scale is 22.75 and it explains 75.8 % of total variance. All items in the Turkish form were found to be consistent with the sub-factors of the original scale; only the order of sub-factors has changed. First and second CFA were conducted to find out whether the Turkish scale was consistent with the original scale. Fit indexes were found to be χ 2 =1106.80 (df=395, p=.0000), χ 2 /df=2.80 RMSEA=0.07, RMR=0.08; SRMR=0.10, GFI=0.86, AGFI=0.80, IFI=0.91, CFI=0.91, NFI=0.89 and NNFI=0.91. When fit indexes were examined, it can be concluded that all values satisfied the criterion values or very close to them. Cronbach Alpha coefficient was found to be .94 for overall scale. The reliability values for the factors was .88, .91, .95, .90 and .92, respectively. The result of the test-retest analysis was found to be .86 which reveals there is a significant high correlation and support that the stability of the scale is high. Discussion and Conclusions: In this study, the scale was adapted to Turkish. The adapted Turkish form was found to be consistent with the original form regarding item-factor consistency and structure. Although the psychometric characteristics of the adapted Turkish scale found to be convenient for a scale, some values were lower than values of the original scale.