To be or not to bey? Written Lexical Decision Accuracy in Neurotypical Aging and Left-Hemisphere Stroke Survivors
Candace M. van der Stelt1,2, J. Vivian Dickens1, Sarah F. Snider1, Sachi Paul1, Elizabeth Dvorak1, Andrew T. DeMarco1, Elizabeth H. Lacey1,2, Peter E. Turkeltaub1,2; 1Georgetown University Medical Center, Washington, DC, 2MedStar National Rehabilitation Hospital, Washington, DC
Alexia, an acquired reading disorder, is diagnosed using single-word oral reading assessments (Coslett & Turkeltaub, 2016). However, reading in daily life usually occurs silently through word-form recognition and comprehension. Word frequency, spelling-to-sound regularity, and imageability affect reading accuracy. Prior research indicates that these psycholinguistic effects may differ in oral versus silent reading (Fischer-Baum et al., 2014). Research on silent reading in alexia is vital to developing comprehensive diagnostic tools and understanding the neurobiological basis of reading. In order to understand silent reading deficits in chronic stroke, we must first investigate how demographic and psycholinguistic factors affect performance in neurotypical aging. Here, we investigate the psycholinguistic patterns of silent word-form recognition in stroke survivors and demographically-matched neurotypical adults through written lexical decision. Participants were 118 native English speakers. Fifty-nine were left-hemisphere stroke survivors (22F:37M; 24 Black:35 White; mean age=60.6 years (SD=11.8), mean education=16 years (SD=3.0); average years post-stroke=4.0, (SD=4.6)). The remaining 59 participants were neurotypical adults (34F:25M; 21 Black:38 White; average age = 60.2 years (SD=11.8); average education=17 years (SD=2.4)). Participants completed a 400-item written lexical decision task, consisting of 200 words crossed on frequency (high/low), imageability (high/low) and regularity (regular/irregular), and 200 length-matched pseudowords. Logistic mixed effects models, using the maximal random effects structure justified by the data, tested the effects of age, education, and psycholinguistic features on accuracy of lexical decision in both groups. Neurotypical participants’ mean accuracy was 97.4% (SD=2.4). Higher education related to higher overall accuracy (OR=1.56, CI=1.27-1.92, p<.001). Accuracy on real words was positively related to education (OR=1.63, CI=1.25-2.11, p<.001) and age (OR=1.39, CI=1.07-1.79, p=.013). We observed significant effects of frequency (OR=13.88 (CI=6.25-30.81), p<.001) and frequency × imageability (OR=0.13, CI=0.03-0.61, p=.010), such that low frequency words were less accurate, more so if they also had low imageability. Stroke participants’ mean accuracy was 88.5% (SD=10.5). Again, higher education related to higher overall accuracy (OR =1.45, CI=1.00-2.09, p=.048). Low frequency (OR=9.81, CI=6.46-14.88, p<.001) and imageability (OR=2.11 CI=1.52-2.93), p<.001) related to lower accuracy. Additionally, two interactions were observed: age × frequency (OR = 0.65, CI 0.49-0.86, p=.003) such that increased age related to higher accuracy on low frequency words; and education × regularity (OR=0.81, CI=0.66-0.99, p=.044) such that higher education related to higher accuracy on irregular words. Greater education and age relate to better lexical decision performance in neurotypical adults and stroke survivors, possibly due to the association of these factors with vocabulary size (Salthouse, 2019). These findings demonstrate the importance of considering demographic factors in alexia assessment. In stroke survivors, the specific relationships of these factors with low frequency and irregular words may reflect increased reliance on lexical representations for reading due to damage to the phonological system. Notably, the patterns observed in written lexical decision here are somewhat different from those previously observed in oral reading in a subset of this study’s cohort (Dickens, 2021). Future directions include direct comparison of lexical decision performance between stroke survivors and neurotypical adults, a comparison of oral reading and lexical decision performance, and lesion-symptom mapping of lexical decision deficits.
Topic Areas: Reading, Disorders: Acquired