Abstract
<p>Precise norm scores are essential for psychological diagnostics. While conventional norming estimates norms within subgroups, continuous norming uses regression to estimate norms based on the entire sample. Consequently, continuous norming should provide sufficient precision with smaller sample sizes than conventional norming. However, clear insights on the sample sizes required to achieve sufficient precision for continuous norming are scarce. To provide empirical insights into the required sample sizes of continuous norming, we present two illustrative examples using normative data from two intelligence tests to evaluate the precision of continuous norms across varying sample sizes, norm predictor values, and ability levels. Our examples show that continuous norming achieves a sufficient precision with substantially smaller samples than conventional norming (N = 1,000–1,500 vs. N = 2,500–3,000). However, precision declines at extreme percentiles and norm predictor values, emphasizing the need to tailor sample size planning to the specific test and its use.</p>