Table 2.

Statistical tests of normality: simulated data

Observations
ProcedureFishMeat
Lilliefors0.850.02
Shapiro-Francia0.810.03
Shapiro-Wilk0.650.06
  • Values represent P values associated with procedures designed to assess the separate null hypotheses that the 7 observations in the fish group and the 5 observations in the meat group (see Ref. 10, Table 2) are consistent with having come from a normal distribution. In their nortest algorithms, the Anderson-Darling and Cramér-von Mises procedures require at least 8 observations. The commands in lines 239–267 of Advances_Statistics_Code_Normal.R analyze the data in the fish and meat groups. To obtain these P values, highlight and submit the lines of code from Table 2: first line to Table 2: last line.