[R-sig-ME] longitudinal analysis when one group switched from control to treatment

[Simon Harmel]

Hello All,

I have a 3-year longitudinal dataset (*see link below the table*). Up to
year 2 (coded "1"), 8 schools (4 in Treatment, 4 in Control) cooperated
with the study. But in year 3 (coded "2"), one of the Treatment schools
(named "good") dropped out.

Also in year 3 (coded "2"), we were made to move one of the *Control *schools
(named "*orange*") to the *Treatment *group. The full design of the study
is shown in the Table below.

I want to regress "year" and "group" on "y" (a continuous response) in lme4
package in R. But is there a way to capture the switch of one of the
control schools to the treatment group?

Thank you very much, Simon

·       *Switched from control to treatment*

·       *Out as of year coded 2*

*SCHOOL NAMES*

*Year*

*Codes*

*Control*

*Treatment*

0

har

john

bus

orange

caro

good

bla

carm

1

har

john

bus

*orange*

caro

good

bla

carm

2

har

john

bus

X

caro

*orange*

bla

carm

*library(lme4)*
*dat <- read.csv('https://raw.githubusercontent.com/hkil/m/master/z.csv
<https://raw.githubusercontent.com/hkil/m/master/z.csv>')*

*m1 <- lmer(y~ year*group + (1|stid), data = dat)      #### 'stid' =
student id                m2 <- lmer(y~ year*group + (1|scid/stid), data =
dat) #### 'scid' = school id*

	[[alternative HTML version deleted]]