"Bill Curtis presented an alternate dataset in 1981 that substantiated very large variability in programmer performance."
Curtis explicitly responded to criticism of the earlier Sackman work and concluded that an alternate dataset still showed enormous variation between programmers.
View Source