compare repeated measures model with random effects with DID

ex DID vs repeated measures.R

# *-----------------------------------------------------------------
# | PROGRAM NAME: ex DID vs repeated measures 
# | DATE: 9/9/20 
# | CREATED BY: MATT BOGARD 
# | PROJECT FILE: https://gist.github.com/BioSciEconomist           
# *----------------------------------------------------------------
# | PURPOSE: compare repeated measures model with random effects with DID
# *----------------------------------------------------------------

# see also: http://econometricsense.blogspot.com/2018/02/intuition-for-random-effects.html

library(dplyr)
library(reshape)

# ref: https://stats.stackexchange.com/questions/194888/difference-in-difference-vs-repeated-measures 
cards <- "subj  group t1  t2 diff
1   A   5   6   1
2   A   5   7   2
3   A   6   9   3
4   A   6   10  4
5   A   5   7   2
6   A   5   7   2
7   A   6   9   3
8   A   6   9   3
9   A   2   2   0
10  A   2   3   1
11  B   10  11  1
12  B   10  11  1
13  B   9   11  2
14  B   9   11  2
15  B   10  18  8
16  B   10  21  11
17  B   12  18  6
18  B   12  19  7
19  B   5   20  15
20  B   5   20  15"

df <- read.table(textConnection(cards), header = TRUE) 
closeAllConnections()

df1 <- melt(data=df, id.vars =c('subj', 'group'), measure.vars = c('t1', 't2'))

names(df1) = c("subj","group","time","value")

# DID model

summary(glm(value ~ group +  time + group*time, data=df1))

# repeated measures model with random effect for subject

library(nlme)

mix<-lme(value ~ group*time, random=(~1 | subj), data = df1)

summary(mix)

#---------------------------------
# additional examples repeated measures
#---------------------------------

cards <- "id group pulse time
1     1    35    1
1     1    25    2
1     1    16    3
2     1    32    1
2     1    23    2
2     1    12    3
3     1    36    1
3     1    22    2
3     1    14    3
4     1    34    1
4     1    21    2
4     1    13    3
5     2    57    1
5     2    43    2
5     2    22    3
6     2    54    1
6     2    46    2
6     2    26    3
7     2    55    1
7     2    46    2
7     2    23    3
8     2    60    1
8     2    47    2
8     2    25    3"


df <- read.table(textConnection(cards), header = TRUE) 
closeAllConnections()

## Convert variables to factor
df2 <- within(df, {
  group <- factor(group)
  time <- factor(time)
  id <- factor(id)
})


#
# visualize
#

with(df2, interaction.plot(time, group, pulse,
                             ylim = c(10, 60), lty = c(1, 12), lwd = 3,
                             ylab = "mean of pulse", xlab = "time", trace.label = "group"))


#
# explore
#

# group differences
df2%>% 
  group_by(group) %>%
  summarize(pulse = mean(pulse))

# differences across time
df2%>% 
  group_by(time) %>%
  summarize(pulse = mean(pulse))

# group and time interactions
df2%>% 
  group_by(group,time) %>%
  summarize(pulse = round(mean(pulse),2))

#
# standard regression
#

summary(glm(pulse ~ group +  time + group*time, data = df2))

#
# repeated measures anova
#

summary(aov(pulse ~ group * time + Error(id), data = df2))


#
# linear mixed model with random effects
#

library(nlme)

summary(lme(pulse ~ group*time, random = (~1 | id), data = df2))

library(lme4)

summary(lmer(pulse ~ group*time + (1 | id), data = df2))