80ResultsReliability CheckingParticipantsData AnalysisDescriptive StatisticsTwenty-one Japanese undergraduates （five males, 16 females; six sophomores, six juniors, nine seniors） in the Department of English Language and Literature, School of Education, who participated in EMI, completed this survey. Initially, 30 students responded to this questionnaire, but nine of them mistak-enly answered the task assuming an EAP class instead of an EMI class, so those responses were excluded from the analysis. Students who responded to the questionnaire were paid 1,000 JPY as an honorarium by the university.Based on their average TOEFL Institutional Testing Program （ITP） scores （M = 531, SD = 49, 15 valid data points）, their English proficiency roughly corresponds to the CEFR B1 level （Council of Europe, 2001）. Seventeen participants reported that they had no previous experience taking content-based courses in English at an outside educational institution （e.g., international school, immersion school, or International Baccalaureate school abroad, or in Japan）; two had enrolled in such institutions for six months to one year, one for two to three years, and another for more than five years.First, after the questionnaires were collected, data anonymization and data cleaning （e.g., removing invalid data） were conducted, along with coding to quantify the frequency and difficulty of responses for each task. In line with Tahara et al. （2021）, we converted “low,” “relatively low,” “relatively high,” and “high” to “1,” “2,” “3,” and “4,” respectively, and “no experience” to “0.” Second, descriptive statistics were used to calculate the mean scores for the frequency and difficulty of each task in the questionnaire to describe and summarize the main characteristics of the dataset. Then, the overall mean scores of the task frequency and difficulty were used as the threshold for their high and low, respectively, and four groupings were made from each combination of high/low frequency and high/low difficulty （e.g., high-frequency and high-difficulty group, high-frequency and low-difficulty group）. Finally, each task in the questionnaire was sorted into four groups based on mean frequency and difficulty scores.Cronbach’s alpha coefficients for frequency and difficulty were calculated to assess the consistency of the measurements. The alpha coefficient for frequency is （α = .95） and （α = .95） for difficulty. Both coefficients meet acceptable levels for Cronbach’s alpha, above 0.7 to 0.8 （Larson-Hall, 2010, p. 171）.Table 1 summarizes the means and standard deviations for each task by category in EMI. In the before-and-after-class category, in addition to the task of researching specialized content in English using the Internet （Q1: M = 2.2, SD = 1.1 for frequency; M = 2.7, SD = 0.6 for difficulty）, reading tasks of specialized textbooks written in English, such as reading chapters （Q5: M = 3.3, SD = 0.6 for frequency;
元のページ ../index.html#86