Psychometric Properties of the Shipley Block Design Task: A Study With Jamaican Young Adults

Abstract

We examined the structure of the new Block Patterns (BP) test from the Shipley Institute of Living Scale-Second Edition in a sample of Jamaican young adults. To date, very little has been published on the properties of this subtest's items and scores. The BP test is similar in design to the Block Design subtest found in many cognitive ability assessments but uses a matching format that minimizes the need for excess materials and time. We analyzed the BP items using item response theory (IRT) methods. Although designed to measure a single construct, the analyses from this study found that the BP subtest is likely measuring more than a single construct, which confounds the interpretation of the instrument's scores. Before the subtest is used clinically, more research should be done to purposefully investigate the effects of ancillary variables on its scores. Keywordsitem response theory, block design, block pattern, Shipley Institute of Living Scale-Second Edition, JamaicaFor almost a century, psychologists have been using nonverbal measures to assess the cognitive ability of individuals with limited education or limited language proficiency (Boake, 2002). Samuel Kohs (1920) developed one of the first nonverbal tests when he created the Block Design (BD) Tests. He initially developed the test by adapting a children's game (Color Cubes), which required individuals to reproduce a series of mosaic designs using colored cubes. As a student of Lewis Terman, Kohs developed his test to be a measure of "intelligence" akin to the Binet but minimizing the use of language. He proposed that successfully completing the tasks involved both analytic (e.g., breaking down the presented designs into smaller units) and synthetic (e.g., reassembling the smaller units with the colored cubes) thinking (Kohs, 1923).

Teacher-Led Math Inquiry: A Cluster Randomized Trial in Belize

“…Upon commitment from all schools, students were assessed using the Shipley Block Pattern (BP) subtest from the Shipley Institute of Living Scale-Second Edition (Shipley, Gruber, Martin, & Klein, 2010), which is a measure of visualspatial cognitive ability. An investigation of the psychometric invariance properties of the BP test with a Caribbean sample and a U.S. sample is provided by Beaujean et al (2017). We implemented the BP test because it is very brief, easy to administer, was designed to use a minimal amount of language (requiring no reading except for a few sentences in the instructions, which are accompanied by an illustration), and, finally, we anticipated that successful use of manipulatives would likely invoke a visual-spatial ability component and consequently correlate with the BP measure, more so than general math ability, which has implications for prerandomization blocking (described below).…”