ANOVA, with main effects only, was used to compare the effects of temperature, cyclic movement, RH, and UV on the modulus ratio. Statistical significance was set at the conventional 5 % level. Table 1 summarizes the results of ANOVA. F-tests show a significant effect of UV, cyclic movement, and RH on the modu­lus decrease; however, ANOVA fails to reveal a significant impact of tempera­ture on the modulus decrease at the 5 % level. We find this result unbelievable after looking at the dex plots, and it might be due to at least one of the following assumptions’ being faulty: normality, homogeneity of variance, independence, or no interactions.

With the above ANOVA caveats in mind, we may form, based on the signifi­cance levels, the following ranked list of factors: (1) cyclic movement, (2) UV, (3) RH, and (4) temperature. Because of the caveats, we use the second statisti­cal method (block plots) to assess factor significance and form a ranked list of importance.