dfEffects: Degrees of Freedom and Effects

Description Usage Arguments Value References Examples

Description

Calculate the effects and degrees of freedom for a set of fully crossed effects.

Usage

1

Arguments

vars

(character vector) Independent Variables

levels

(integer vector) Levels for each of those indpendent variables respectively

Value

data.frame with two columns, effect (the name of the effect) and df (the df associated with that effect)

References

Rosenthal & Rosnow. (2008). Essentials of Behavioral Research. Boston: Mc Graw Hill.

Examples

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
#Table 18.24
#Calculate the within effects
within <- dfEffects(c("Expectancy","Pupil Sex"),c(2,2))
#Specify the between effects (these can't be calculated using dfEffects because they aren't fully crossed)
btwn <- data.frame(effect=c("Sex of Teacher","Teachers (within sex)"),df=c(1,8))
#Now we rbind these two and use grid.expand to explore all possible combinations

table.18.24 <- rbind(btwn,within,data.frame(effect=apply(expand.grid(btwn$effect,within$effect),1,paste,collapse=":"),df=apply(expand.grid(btwn$df,within$df),1,prod)))
# Table 18.30 uses the same between subjects effects but has a more complex within structure that we create using this function

within <- dfEffects(c("Expectancy","Pupil Sex","Diagnosis"),c(2,2,2))
rbind(btwn,within,data.frame(effect=apply(expand.grid(btwn$effect,within$effect),1,paste,collapse=":"),df=apply(expand.grid(btwn$df,within$df),1,prod)))
# Note: Obviously the order of effects is different here than in R&R.

drknexus/repsych documentation built on May 15, 2019, 2:19 p.m.