Sample size estimation is fundamental to traffic engineering analysis. An iterative procedure using standard deviation estimates is the most reliable method of sample size estimation, but practitioners find it cumbersome. In the past, attempts have been made to simplify this process and render it a one-step exercise. To this end, the Institute of Transportation Engineers (ITE) manuals provide standard deviation values for spot speeds for roadways classified by annual average daily traffic (AADT) volumes. This study obtained new estimates of standard deviation using a more granular, operational-level data collected by traffic detectors. It validated the earlier ITE estimates for moderate traffic conditions on freeways, but found the ITE estimates to be inadequate for arterials and for very low and heavy traffic conditions on freeways. This study also found a consistent U-shaped relationship between the standard deviation of speed and traffic volume, which alludes to an inherent speed distribution profile of traffic data. These speed distribution profiles are robust since they exist across different a) locations, b) roadway types (freeway and arterial), and c) aggregation intervals (5-, 15-, and 60-minutes). The U-shaped curves are also validated through existing traffic safety literature.