Norm Block Sample Sizes

Abstract

The representativeness, recency, and size of norm samples strongly influence the accuracy of inferences drawn from their scores. Inadequate norm samples may lead to inflated or deflated scores for individuals and poorer prediction of developmental and academic outcomes. The purpose of this study was to apply Kranzler and Floyd’s method for estimating norm block sample sizes via a review of the most prominent, individually administered intelligence tests. A rigorous, double-coding process was used to obtain these estimated sample sizes for 17 intelligence tests (10 full-length multidimensional tests, 4 nonverbal intelligence tests, and 3 brief intelligence tests). Overall, 47% of the tests failed to meet the minimum standard of at least 30 participants per norm block across age groups, and estimated norm block sizes were smallest for elementary school–age children. These results can inform intelligence test selection by practitioners and researchers, and they should be considered by test publishers when developing, revising, and reporting information about their tests.

Test Review: Wechsler Intelligence Scale for Children, Fifth Edition: Canadian 322 (WISC-VCDN) by D. Wechsler

“…The size of the norming sample, however, appears to be insufficient when compared with current guidelines and other commonly used cognitive assessment batteries (Kranzler & Floyd, 2013;Norfolk et al, 2015), which may bring into question the overall quality and trustworthiness of the norm-referenced scores that are generated from its normative sample (Kranzler & Floyd, 2013). The WISC-V CDN demonstrates a number of strengths, such as its relatively short administration time, strong psychometric properties, and engaging materials.…”

“…Norming blocks are the specific age ranges that are used to generate scaled scores for interpretative purposes. Norfolk et al (2015) reviewed the normative samples of 17 individually administered intelligence tests to determine whether they met the guidelines established by Kranzler and Floyd (2013). Kranzler and Floyd (2013) explain that if 3-month norming blocks are used to generate scores for interpretation, then the sample sizes of these blocks should be in line with the established norming sample size guidelines.…”

“…Norming blocks are important because many norm-referenced tests used to assess children and adolescents, such as cognitive assessment batteries, use 3-or 4-month intervals to generate scaled scores because of the rapid developmental changes that occur during the school years. According to Norfolk et al (2015), the American sample of the WISC-V did meet the medium standard of having a minimum sample of 66 participants per norming block, with a total sample of 2,200 participants. Norfolk et al (2015) reviewed the normative samples of 17 individually administered intelligence tests to determine whether they met the guidelines established by Kranzler and Floyd (2013).…”

“…Kranzler and Floyd (2013) explain that if 3-month norming blocks are used to generate scores for interpretation, then the sample sizes of these blocks should be in line with the established norming sample size guidelines. Norfolk et al (2015) reviewed the normative samples of 17 individually administered intelligence tests to determine whether they met the guidelines established by Kranzler and Floyd (2013). Their findings suggested that 47% of the tests reviewed failed to meet the minimum standard of having at least 30 participants per norming block.…”

“…Their findings suggested that 47% of the tests reviewed failed to meet the minimum standard of having at least 30 participants per norming block. According to Norfolk et al (2015), the American sample of the WISC-V did meet the medium standard of having a minimum sample of 66 participants per norming block, with a total sample of 2,200 participants. The norming sample for the WISC-V CDN , however, appears to have a norming block sample size of 26.7, which is below the minimum standard.…”