The power of statistics in surfacing or indicating systematic patterns of equity or inequity rely on large sample sizes. Small samples tend to have high data variability and low statistical power, meaning that statistical analyses might not be able to identify significant effects even if they are present in the population. On the flipside, they might falsely identify an effect as significant. Small sample sizes also mean that there are greater chances for sampling error.  Furthermore, using percentages to report data with small sample sizes and using those percentages to infer meaningful conclusions or systemic patterns about a subgroup can be misleading, and lead to misguided actions that are not based in sound data analysis. Also, important demographic data that are not tracked (for example, gender identity and religion) are wholly invisible in most schools’ quantitative data. 

For this reason, using qualitative data is often the most powerful and analytically sound way to gather meaningful evidence about lived experiences of inequity, or to guide action in small settings.

Whether data are quantitative or qualitative, it is challenging to safeguard individual identities when reporting on small communities. People may speculate about qualitative findings and identify individuals (from a quote or the description of a traumatic experience, for instance) or be able to identify individuals from quantitative subgroup data (for example, if a survey reveals that “90% of nonbinary students feel unsafe at school”, and there are only 7 nonbinary students, then those students’ confidential feelings are put on public display).  

Within a setting where non-dominant students or participants already experience inequity or harm, the lack of anonymity in the process of an equity audit can create mistrust, strain relationships, and make individuals targets of unwanted attention or further harm.