A psychometric analysis of the Mississippi Assessment Technique for Identifying Learning Disabled Adults (MATILDA)

Jessica Dorsett


The primary objectives of this study were to determine the psychometric characteristics of the MATILDA and to compare MATILDA scores with those of the Adult Learning Disability Screen (ALDS) in order to ascertain how well these instruments correlate in identifying adult literacy education students at risk for learning disabilities. The study further sought to ascertain percentages of agreement between MATILDA scores, ALDS Inventory Scale scores, ALDS Rating Scale scores, and Teacher-ratings as determined by official documentation. A sample of 432 participants enrolled in three literacy education programs in one mid-south state was used in this study. A principal component factor analysis utilizing both orthogonal and oblique rotations resulted in a six-factor solution. While the alpha reliability for the total scores on the MATILDA was adequate at .79, the internal consistency among scores on the individual factors was greater than .70 for only three of the six factors. Correlations between MATILDA factor scores and ALDS Scale scores were generally statistically significant at less than .001 ( p < .001); however, the largest correlation was only .30, indicating a relatively weak relationship. The multiple correlations between the MATILDA factor scores and the ALDS Rating Scale and ALDS Inventory Scale scores were modest at .373 and .358, respectively. Test results to ascertain percentages of agreement among the MATILDA scores, both Inventory and Rating Scale scores, and Teacher-ratings indicate that both the MATILDA scores and the ALDS Rating Scale score agreed equally as well with Teacher-ratings (58.80%); however, the Inventory Scale scores and Teacher-ratings had the highest percentage of agreement (85.42%). Results from this study point to directions for further improvement of the individual factors of MATILDA in order to improve the instrument's psychometric soundness.