Description

The present study aims to provide normative verb generation behavioral data to subsequently investigate word retrieval processes in individuals with neurological disorders such as epilepsy of stroke-induced aphasia. Verb generation is a classically used paradigm to probe word retrieval in clinical settings, however normative data is not readily available to compare the performance of individuals with neurological disorders possibly affecting word retrieval processes to that of control participants. In the verb generation task, individuals are asked to produce a verb in response to a noun. Some nouns are associated with many possible verbs (eg. “ball” can be associated with “kick”, “throw”, “bounce”, etc.), whereas other nouns are associated with fewer verbs (e.g., “chair” which usually is associated with “sit”). Entropy in this context refers to the degree of variability in the set of verbs associated with a noun. Generating a verb in response to a noun with low entropy typically leads to shorter reaction times as compared to nouns associated with high entropy. We collected behavioral data from 42 participants, (4 male, 38 female, an average age of 28.1, with a standard deviation of 10.6). using an online data collection platform. Our stimuli consisted of 50 common nouns with varying associated entropy values. Entropy values for these nouns were calculated using data from a separate control group, which included 34 participants, (9 male, 24 female, and 1 unspecified, an average age of 28.3, with a standard deviation of 9.4). We used a correlation model to assess if there was a correlation between the entropy values and reaction times derived from our 42 online participants, and we excluded incorrect responses from participants in our analysis. As expected, we found a strong positive correlation between reaction time and entropy values, with a correlation coefficient of 0.75, replicating the finding that increased selection needs lead to increased reaction times in word retrieval. This result indicates that our normative data set shows the expected entropy effect and will subsequently be used to characterize the performance of individuals with neurological disorders on the same task.