A follow-up alpha emissivity study was conducted to examine the wide variability observed in previous work that was hypothesized to be due to differences in the pulse height discrimination threshold among participant's equipment. Two samples, one mixed energy and one monoenergetic, were prepared and sequentially circulated to all participants for counting. Analysis of the data demonstrates that only a small portion of the variability is explained by this mechanism. The role of the sample to entrance window gap for some counters was analyzed post hoc using the same data set and may be responsible for a large amount of the variability. The results of this large scale study demonstrate the continuing uncertainty for these measurements and the importance of interpreting their results appropriately when estimating soft error rates.