||A Rasch analysis of L2-English achievement goals of female university students
RICHARD, Jean-Pierre Joseph
156 , 20150930
Achievement goals are cognitive representations of future-focused purposes learners adopt in academic situations to direct behavior towards approachment or avoidance of competence-related end states. In the 2×2 achievement goals model employed in this study, goals are defined as mastery and performance; and two valence dimensions, positive and negative, also co-occur. The definitions and dimensions led to the 12-item achievement goals questionnaire （AGQ） and revised form （AGQ-R）. Few studies have used the Rasch Model to investigate achievement goals, and results have been inconclusive or conflicting likely due to the limited number of items （k＝3） measuring each goal. Validation of an expanded version of the AGQ-R （k＝20） using the Rasch Model is the primary purpose of this study. Data were gathered from first-year L1-Japanese participants （N＝125） enrolled in six sections of one required L2-English course at one private, female-only university in the Kantō region.
Mastery-approach items had highest mean scores, statistically higher than the other three goal orientations. The complete questionnaire items （k＝20） were tested for unidimensionality: total variance explained was low, first contrast eigenvalue was large, with low disattenuated correlations, and theoretically distinct constructs loading separately, which suggest multidimensionality. Five other competing models, All-Mastery, All-Performance, All-Approach, All-Avoidance, and Trichotomous framework were likewise judged to be multidimensional. Lastly, achievement goals were tested separately: Performance- and mastery-approach items had best model fit. The Rasch Model provided evidence for the validity of the extended AGQ-R used in this current study; however, problems remain, in particular the need to write more items with greater levels of difficulty.