What is a limitation of using ordinal scales?

The correct choice highlights a significant limitation of ordinal scales, which is that while these scales allow for ranking of values, they do not ensure that the intervals between the ranks are consistent. For instance, if you consider a survey that asks participants to rank their satisfaction from "very unsatisfied" to "very satisfied," the difference in satisfaction between "satisfied" and "very satisfied" may not be the same as the difference between "neutral" and "satisfied." This variability in the magnitude of differences complicates the use of ordinal data for certain types of statistical analysis that rely on equal intervals or ratios.

In contrast, ordinal scales do permit ranking of items (which is why the choice stating that values cannot be ranked is inaccurate), but they fall short when it comes to accurately measuring the distance between those ranks. This limits the types of analyses that can be appropriately conducted on ordinal data. While it's also true that some numerical analyses are not appropriate for ordinal data, stating that they do not allow for any numerical analysis at all is misleading; certain non-parametric statistics can be performed on ordinal data, though they are constrained. Lastly, while larger sample sizes can help in achieving statistical power, they are not a specific limitation of ordinal scales themselves.