I'm attempting to perform a Hinkin and Tracey content validation on potential scale items and have the following dataset (sample) with 74 unique columns:
cleandata <- structure(list(Condition = c("RS", "AS", "BGPS", "APCS", "OP", "TS"
), alt_energy = c(2, 5, 3, 3, 2, 2), animal_product = c(5, 3,
4, 4, 3, 1), deforest = c(5, 1, 4, 1, 2, 1)), row.names = c(NA,
6L), class = c("tbl_df", "tbl", "data.frame"))
Only one column (Condition) is the "dimension," and I want to see if the mean responses for the remaining 73 columns significantly differ between conditions. Basically, I want to see if the scale item successfully reflects only 1 dimension.
I want to run an anova and Tukey HSD on all the columns at once to get everything in one neat output:
test <- aov(as.matrix(cleandata[, -1]) ~ as.factor(Condition), data = cleandata)
summary(test, effect.size = "both", detailed = TRUE, observed = NULL)
But I'm unable to run the HSD follow-up
tukeyHSD(test)
Getting the following error: Error in model.tables.aov(x, "means") : 'model.tables' is not implemented for multiple responses
Is there anyway to loop the HSD so I get one clean output of anova results and follow-up pairwise comparisons?
