Examples of typical reference interval data distribution plots include all of the following except

Examples of typical reference interval data distribution plots include all of the following except

• A. ROC
• B. Nonparametric
• C. Parametric
• D. Bimodal

Answer: (A)

When you’re describing how a biomarker is distributed in a reference population, the goal is to show where most healthy values lie and how they spread. Nonparametric methods map the actual percentiles directly, without assuming any shape for the distribution, which is handy when the data aren’t normal. Parametric approaches assume a normal distribution and use the mean and standard deviation to define the reference interval, typically as mean ± a few standard deviations. If the data look or are known to come from two subgroups, a bimodal distribution acknowledges two peaks and may lead you to stratify the reference interval or use a mixture approach. A Receiver Operating Characteristic curve, in contrast, is about diagnostic performance: it plots sensitivity versus 1 minus specificity across thresholds to evaluate how well a test separates diseased from non-diseased individuals. It doesn’t depict the distribution of reference values in a healthy population, so it isn’t a typical plot used for establishing reference intervals. Therefore, ROC is the exception.

When you’re describing how a biomarker is distributed in a reference population, the goal is to show where most healthy values lie and how they spread. Nonparametric methods map the actual percentiles directly, without assuming any shape for the distribution, which is handy when the data aren’t normal. Parametric approaches assume a normal distribution and use the mean and standard deviation to define the reference interval, typically as mean ± a few standard deviations. If the data look or are known to come from two subgroups, a bimodal distribution acknowledges two peaks and may lead you to stratify the reference interval or use a mixture approach.

A Receiver Operating Characteristic curve, in contrast, is about diagnostic performance: it plots sensitivity versus 1 minus specificity across thresholds to evaluate how well a test separates diseased from non-diseased individuals. It doesn’t depict the distribution of reference values in a healthy population, so it isn’t a typical plot used for establishing reference intervals. Therefore, ROC is the exception.