Agreement among four models used for diagnosing learning disabilities.

We compared the level of agreement among four models used to diagnose learning disabilities (LD), including the simple discrepancy, intraindividual, intellectual ability-achievement, and underachievement models. The sample included 170 clinic-referred university students. The simple discrepancy model diagnosed significantly more students with LD than the other three models. The highest degree of agreement occurred between the intraindividual and intellectual ability-achievement models (70%); the lowest level of agreement occurred between the simple discrepancy and underachievement models (48%). Finally, only two of the six comparisons among the four models demonstrated significant correlations. We conclude that even when discrepancy models diagnose similar numbers of students with LD, the same students are not diagnosed across different models.