Impaired Instance acquisition as a Cause of the Comorbidity of Learning Disorders in Young Adults

Chiara Valeria Marinelli ¹ Giuliana Nardacchione ^1* Marialuisa Martelli ² Vincenza Tommasi ^1* Marco Turi ³ Paola Angelelli ³ Pierpaolo Limone ⁴ Pierluigi Zoccolotti ²

• ¹ University of Foggia, Foggia, Italy
• ² Sapienza University of Rome, Rome, Lazio, Italy
• ³ University of Salento, Lecce, Apulia, Italy
• ⁴ Pegaso University, Naples, Campania, Italy

Introduction: The “instance theory of automatization” suggests that automaticity relies on acquiring specific instances that enhance performance, preventing the slower application of procedures. It has been proposed that a low ability in instance acquisition may be the key cause of the comorbidity among learning disorders. We investigated performance on a learning task to test the hypothesis that difficulties in acquiring and consolidating instances would be linked with comorbid learning disorders. Methods: We examined the individual rate of learning of 143 young adults with typical development (32M, 111F, mean age: 20.3) and 59 with specific learning disorders (SLD; 12M and 47F, mean age: 20.9).Results: Both groups significantly reduced their response times across learning trials (following a power trend) without generalization to untrained items, indicating that learning occurred through instance acquisition. Initially, participants with SLD performed worse than the controls. However, they reduced their times by about 96 sec with practice, even though their “endpoint” (asymptote) remained slower than controls. Group differences were related to these two scaling values, not the power curve coefficient. Subsequently, we reclassified the sample into three groups based on the type of deficit: one without procedural/instance deficits ("Control" group), one with selective deficits in “procedural” tasks ("Poor procedural" group), and one with deficits in instance-based tasks ("Poor instance" group). The poor instance group not only showed deficits across all tasks requiring instance retrieval (i.e., arithmetical facts and lexical representation retrieval) but was also slower (86 sec) in the learning task compared to the other groups (58 and 70 sec, respectively; at least p < .01). The “Poor procedural” group behaved similarly to the “Control” group. Conclusion: Results support with the notion that a low ability to acquire and consolidate instances may contribute to the comorbidity of learning disorders.