Statistics 23

First, we employed classical test analysis and examined reliability, means, and standard deviations for the questionnaire fields with SPSS Statistics 23.0. In the case of frequency of strategy use, we aimed to find out how strategy use was perceived by our sample. We also compared the students’ strategy use vis-à-vis their English language achievement and attitude using an independent sample t-test. To interpret effect size, we followed Wei et al.’s (2019) (link) and Wei and Hu’s (2019) benchmark: under 0.005 is small, 0.01 is typical or medium, 0.02 is large, and is 0.09 very large. We used R² unsquared; thus, the benchmark for the effect size index is 0.07, 0.10, 0.14, and 0.30, which, respectively, represents small, medium, large, and very large cut-off values. We applied path analysis to map the possible relationships and effects of our variables. We studied the goodness-of-fit indices by applying various cut-off values for many fit indices, including the Tucker–Lewis index (TLI), the normed fit index (NFI), the comparative fit index (CFI), the root mean square error of approximation (RMSEA), and Chi-square values (Kline, 2015 ). TLI, NFI, and CFI were regarded as eligible with a cut-off value of 0.95, and RMSEA values indicated an acceptable fit of 0.8 (Kline, 2015 ).

SPSS Statistics 23.0 was used to measure the statistical differences, estimated by one-way ANOVA followed by LSD and Duncan’s multiple range test. p < 0.05 was used to indicate statistical significance. Quantitative data were expressed as mean ± SD.