The Algebraic Thinking Test (ATT) Instrument was developed to measure algebraic thinking skills in solving problems involving the topic of linear equations. This study meant to determine the quality of ATT instruments through the validity and reliability analysis using the Rasch Measurement Model. ATT comprises of nine closed ended questions with three constructs of algebraic thinking namely: Generalized Arithmetic (GA), Functions (F) and Modeling (M). The sample consisted of 120 students aged 12 to 13 years from one rural area school in Tuaran district, Sabah. Validity analysis found that the polarity of the items through the PTMEA-Corr value indicated that all ten items were >0.00 (+). Through item suitability analysis, all items are still retained as they meet the requirements in the range of one of the MNSQ, Outfit ZSTD and PTMEA-Corr outfit. In the Principal Components Analysis (PCA), the Raw Variance Explained by Measures was found 61.1% (excellent) and the Unexplained variance value in the first contrast below 15% indicate that the instrument has a strong dimensionality and has high construct validity. The reliability of Cronbach Alpha (KR-20) demonstrated a value of 0.90 (very high), item reliability of 0.98 (excellent) and respondent 0.86 (good) with an isolation value of item 6.29 and respondent 2.45. In conclusion, ATT has good validity and high reliability in measuring algebraic thinking skills among form one students in secondary schools.
Read full abstract