Identifying students with dyslexia in higher education

Area of Science:

• Psychology
• Education
• Linguistics

Background:

• Higher education institutions are seeing more students with dyslexia.
• Standardized diagnostic tools are increasingly necessary.
• Previous research on dyslexia assessment was limited to English and discriminant analysis.

Purpose of the Study:

• To investigate the minimal set of tests required for reliable dyslexia diagnosis in Dutch-speaking higher education students.
• To compare prediction analysis with traditional postdiction analysis for variable selection in dyslexia assessment.

Main Methods:

• A prediction analysis was employed using Dutch language data.
• The study focused on higher education students.
• Variable selection was based on predictive accuracy rather than post hoc analysis.

Main Results:

• A concise set of three tests (word reading, word spelling, phonological awareness) proved sufficient for diagnosing dyslexia.
• Prediction analysis identified a smaller, more relevant set of variables compared to postdiction analysis.
• Additional variables selected by postdiction analysis did not enhance the predictive power of the diagnostic model.

Conclusions:

• A limited number of tests can reliably diagnose dyslexia in higher education students.
• Phonological awareness, word reading, and spelling remain key indicators of dyslexia in adults.
• Prediction analysis offers a more robust method for selecting essential diagnostic variables than postdiction analysis.