Share This

Anomia (word-finding difficulties) can arise when a person has a stroke, dementia or other neurological disorder affecting the left (typically) hemisphere of the brain. There are lots of theories underlying the process of word retrieval, many of which have not been adequately able to account for the spreading activation that requires a person to generate a lexical and phonological word form. The author of this paper flags that cognitive psychometric models, which rely on decision theory, can complement the more mechanistic models, such as the computational model, in our understanding of word retrieval (or naming). Probability can provide us with some insight to the probability of a person generating a word, but this is complicated by the fact that the process is dependent on multiple independent abilities. This issue could be dealt with by using a data driven approach - but asking a person to name thousands of items is not feasible and, instead, collating a small sample set across subtests is more achievable. This can allow for a weighted sum or compensatory model. Bayesian statistics provide a way to assign probabilities to the model. This modelling does simplify a set of observations and should be interpreted on a probability scale. Additionally, there are many practical barriers to applying this type of model in a real-world setting. The author of this article emphasises that mathematical approaches to multicomponent analysis, such as the MPT-Naming model, could have a real impact on diagnosis and management of anomia in the future.

Disentangling the Psycholinguistic Loci of Anomia with Cognitive Psychometric Models
Walker G M.
Share This
Anna Volkmer

UCL, London, UK.

View Full Profile