Validation and standardization of the Psycholinguistic Assessments of Language Processing in Aphasia (PALPA)

Abstract

ABSTRACT Aims This study aimed to standardize and assess the psychometric properties of a selected protocol of subtests from the Psycholinguistic Assessments of Language Processing in Aphasia (PALPA), a battery of tests that measures a wide range of language abilities underpinned by a cognitive neuropsychological model of language processing. Methods One hundred healthy participants were recruited (mean age 64 ± 10.4 years) as part of a large population-based study assessing cognitive change in Amyotrophic Lateral Sclerosis (ALS) and were assessed using the PALPA protocol. Reliability was measured by means of internal consistency and test-retest reliability. Validity was assessed through an inter-correlational analysis between PALPA subtests and a correlational analysis between PALPA subtests and external variables. A supplementary validation study (n = 35) of the Auditory Sentence – Picture Matching subtest of the PALPA against the TROG-2, another test of grammatical and syntactic processing, was carried out. Results The PALPA protocol had good internal consistency (Cronbach’s alpha coefficient = 0.78). Intra-class correlations indicated moderate to excellent test-retest reliability (ICCs ranging from 0.62 to 0.93), with significant agreements (p <.05) between time points in most cases. Validity of PALPA subtests was demonstrated by significant positive correlations among PALPA subtests (especially between those measuring the same language construct, ranging from 0.41 to 0.74). Moreover, the PALPA subtests correlated equivalently with external variables such as age, years of education and IQ, which further supported construct validity. Conclusion This study has proven adequate reliability and validity of a standardized protocol of the PALPA. Based on results from the assessment of psychometric properties of the PALPA protocol studied, adjusted IQ-based PALPA scores were generated.