Performance of Canadian adults on the Graded Naming Test

Abstract

Background: The Graded Naming Test (McKenna & Warrington, 1983) is designed to assess mild naming deficits. In order to do so, it uses relatively uncommon stimuli such as cowl and bellows. There are two studies reporting scores obtained by adults in England, but there is no published study of test–retest reliability. There is no study documenting performance of Canadian adults on this test.Aims: This study examined the performance of English‐speaking Canadian adults on the Graded Naming Test (GNT), in order to: (1) compare their performance to that of the two normative samples; (2) determine whether the 30 test items met the psychometric standards set in the test manual; and (3) assess the test–retest reliability of the GNT.Methods and Procedures: A total of 30 adults, aged 20 to 52 years, named the 30 pictures making up the GNT. The unusual cueing procedures specified in the test manual were not followed. Of the 30 participants, 28 returned for a second test, 7 to 15 days later.Outcomes and Results: The mean score of 17.5 was below that reported in the two studies of English adults. Scores were significantly correlated with level of education, and with estimated frequency of occurrence and age of acquisition. Item difficulty differed in many ways and for many pictures from that reported in the test manual. Problems with multiple names for several pictures and with the instructions for cueing are discussed. The mean score improved significantly on retesting.Conclusions: The Graded Naming Test requires further investigation, including clarification of administration and scoring procedures, and studies of larger samples of adults in other countries to determine local norms.