Last active
August 29, 2015 14:27
-
-
Save wetherc/46c70e78b53b15e5e01e to your computer and use it in GitHub Desktop.
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
######################################## | |
######################################## | |
# | |
# Repeated Measures ANOVAs in R | |
# | |
######################################## | |
######################################## | |
######################################## | |
# Construct a sample dataset | |
######################################## | |
set.seed(5250) | |
# Say we're measuring changes in stress on | |
# a numerical measure (1 = low stress; 100 = | |
# high stress) in response to viewing | |
# a happy or angry image accompanied by | |
# either Disney or horror music | |
myData <- data.frame(PID = rep(seq(from = 1, | |
to = 50, by = 1), 20), | |
stress = sample(x = 1:100, | |
size = 1000, | |
replace = TRUE), | |
image = sample(c("Happy", "Angry"), | |
size = 1000, | |
replace = TRUE), | |
music = sample(c("Disney", "Horror"), | |
size = 1000, | |
replace = TRUE) | |
) | |
# Convert to factors | |
myData <- within(myData, { | |
PID <- factor(PID) | |
image <- factor(image) | |
music <- factor(music) | |
}) | |
# Order for tidy viewing | |
myData <- myData[order(myData$PID), ] | |
######################################## | |
# Extract condition means | |
######################################## | |
# Group outcomes by participant, music | |
# condition, and image condition and | |
# compute the mean stress score | |
myData.mean <- aggregate(myData$stress, | |
by = list(myData$PID, myData$music, | |
myData$image), | |
FUN = 'mean') | |
# Rename the columns for ease of use | |
colnames(myData.mean) <- c("PID","music","image","stress") | |
# Order data for tidy viewing | |
myData.mean <- myData.mean[order(myData.mean$PID), ] | |
######################################## | |
# Construct the ANOVA | |
######################################## | |
# Include error term for between-participant | |
# error across each within-subjects independent | |
# variable. Asterisks represent interaction | |
# effects | |
stress.aov <- with(myData.mean, | |
aov(stress ~ music * image + | |
Error(PID / (music * image))) | |
) | |
# Summarize results of ANOVA | |
summary(stress.aov) | |
######################################## | |
# Include between-subjects effect | |
######################################## | |
set.seed(5250) | |
# Basically our original data set, but with | |
# time of day added in as a between-subjects | |
# variable | |
myData <- data.frame(PID = rep(seq(from = 1, | |
to = 50, by = 1), 20), | |
stress = sample(x = 1:100, | |
size = 1000, | |
replace = TRUE), | |
image = sample(c("Happy", "Angry"), | |
size = 1000, | |
replace = TRUE), | |
music = sample(c("Disney", "Horror"), | |
size = 1000, | |
replace = TRUE), | |
time = rep(sample(c("Day", "Night"), | |
size = 50, | |
replace = TRUE), 2)) | |
# Convert to factors | |
myData <- within(myData, { | |
PID <- factor(PID) | |
image <- factor(image) | |
music <- factor(music) | |
time <- factor(time) | |
}) | |
# Order data | |
myData <- myData[order(myData$PID), ] | |
# Aggregate mean stress values by participant | |
# by condition | |
myData.mean <- aggregate(myData$stress, | |
by = list(myData$PID, myData$music, | |
myData$image, myData$time), | |
FUN = 'mean') | |
# Name the columns for ease of use | |
colnames(myData.mean) <- c("PID", "music", "image", | |
"time", "stress") | |
# Order data | |
myData.mean <- myData.mean[order(myData.mean$PID), ] | |
# Construct aov object, including between-subjects | |
# time factor. We do not include this in our error | |
# term -- only the within-subjects factors | |
stress.aov <- with(myData.mean, | |
aov(stress ~ time * music * | |
image + Error(PID / (music * image)))) | |
# Summarize findings | |
summary(stress.aov) | |
######################################## | |
# What does significance look like? | |
######################################## | |
set.seed(982) | |
# Basically our original data set, but with | |
# time of day added in as a between-subjects | |
# variable | |
myData <- data.frame(PID = rep(seq(from = 1, | |
to = 50, by = 1), 20), | |
stress = sample(x = 1:100, | |
size = 1000, | |
replace = TRUE), | |
image = sample(c("Happy", "Angry"), | |
size = 1000, | |
replace = TRUE), | |
music = sample(c("Disney", "Horror"), | |
size = 1000, | |
replace = TRUE), | |
time = rep(sample(c("Day", "Night"), | |
size = 50, | |
replace = TRUE), 2)) | |
# Convert to factors | |
myData <- within(myData, { | |
PID <- factor(PID) | |
image <- factor(image) | |
music <- factor(music) | |
time <- factor(time) | |
}) | |
myData$music[myData$stress <= 25] <- "Disney" | |
myData.mean <- aggregate(myData$stress, | |
by = list(myData$PID, myData$music, | |
myData$image), | |
FUN = 'mean') | |
# Name the columns for ease of use | |
colnames(myData.mean) <- c("PID", "music", "image", "stress") | |
# Remove a participant with missing data | |
myData.mean <- myData.mean[myData.mean$PID != 18, ] | |
# Construct aov object, including between-subjects | |
# time factor. We do not include this in our error | |
# term -- only the within-subjects factors | |
stress.aov <- with(myData.mean, | |
aov(stress ~ music * image + | |
Error(PID / (music * image)))) | |
# Summarize findings | |
summary(stress.aov) |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment