HOW MANY FMV DLOM STUDY TRANSACTIONS ARE NEEDED TO HAVE A STATISTICALLY VALID SAMPLE (I.E. AFTER NARROWING HITS WITH SEARCH CRITERIA)?
Statistically valid, in this context (and in layman’s terms), means that the mean or median or whatever other measure you draw from the sample will tend to be close to the corresponding measures drawn from the entire population — using consistent search criteria. Generally speaking, the sample size necessary to generate a statistically valid mean will vary depending on the standard deviation of the sample. Statisticians often discuss whether a particular measure is statistically significant over certain thresholds of confidence (a function of the StDev). Practitioners that use the FMV restricted stock study often generate indications that are meant to be representative for their subject companies based on several search criteria, which can tend to produce small sample sizes. Some appraisers use 20-30 data-points as a rough guideline to how many data-points are required, and tend to be a bit less confident when using sample sizes smaller than that. However, if you see a clear trend tow
