Non-parametric tests are used in which situations?

Non-parametric tests are used in which situations?

• A. When the data are normally distributed and measured at interval level.
• B. When data violate normal distribution or are ordinal.
• C. Only for categorical data.
• D. Only for large samples.

Answer: (B)

Non-parametric tests are used when the data don’t meet the assumptions of parametric tests, especially when the distribution is not normal or the data are ordinal rather than interval/ratio. They require fewer assumptions about shape and scale, which makes them appropriate for skewed data, outliers, small samples, or when the measurement level is ordinal. Examples of situations where you’d choose non-parametric methods include comparing two independent groups with an outcome that’s not normally distributed, or assessing a monotonic relationship with ordinal data. It’s not that non-parametric tests are reserved for categorical data, and they aren’t only for large samples; parametric tests can be more powerful when assumptions hold and sample size is large, so non-parametric options are chosen mainly for non-normality or ordinal measurement.

Non-parametric tests are used when the data don’t meet the assumptions of parametric tests, especially when the distribution is not normal or the data are ordinal rather than interval/ratio. They require fewer assumptions about shape and scale, which makes them appropriate for skewed data, outliers, small samples, or when the measurement level is ordinal. Examples of situations where you’d choose non-parametric methods include comparing two independent groups with an outcome that’s not normally distributed, or assessing a monotonic relationship with ordinal data. It’s not that non-parametric tests are reserved for categorical data, and they aren’t only for large samples; parametric tests can be more powerful when assumptions hold and sample size is large, so non-parametric options are chosen mainly for non-normality or ordinal measurement.