Refers to statistical complications or problems due to having few data. If your question is about a small sample relative to the number of variables, please use the [underdetermined] tag instead.

Only having access to a small sample may cause certain problems, for example: statistical tests may be underpowered; assumptions may be more difficult to verify; if data are not sufficiently normal, the Central Limit Theorem may not be able to ensure the normality of relevant sampling distributions, etc.