CWTL CRCT Analysis

Preparation

data<-read.csv("/Users/samanthabouwmeester/Documents/CWTL/merged_data_all.csv")

data$SumAttendance<-rowSums(data[,which(substring(colnames(data),1,4)=="week")],na.rm = TRUE)
data$Attendance=1
data$Attendance[data$SumAttendance<21*5*.8]=0


# Missings values 

#baseline num
bas.num<-data[,c("total_nid","total_qds","total_mis","total_add","total_sub","total_prb","total_nu")]

#Are there any missings values?
any(is.na(rowSums(bas.num)))

## [1] FALSE

# FALSE, so no missing values

#baseline lit
bas.lit<-data[,c("total_lid","total_pho","total_muw","total_non","total_com","total_lis","total_wrt")]

# Are there any missings values?
any(is.na(rowSums(bas.lit)))

## [1] FALSE

# FALSE, so no missing values

#endline num
end.num<-data[,c("total_nid_end","total_qds_end","total_mis_end","total_add_end","total_sub_end","total_prb_end","total_nu_end")]

# Are there any missings values?
any(is.na(rowSums(end.num)))

## [1] TRUE

# TRUE, so missing values: multiple imputation

mis.number=rep(NA,7)
for(i in 1:ncol(end.num))
mis.number[i]=length(which(is.na(end.num[,i])))

mis.number

## [1] 69 69 69 69 69 69 69

# 69 participants miss the endline data for numericy all tasks. This is 69/1507=.046, 4.6%
# These endline scores were imputed.

scores.mice<-mice(end.num,print=FALSE)
end.num.MI<-complete(scores.mice)

data[,c("total_nid_end","total_qds_end","total_mis_end","total_add_end","total_sub_end","total_prb_end","total_nu_end")]=end.num.MI

#check
end.num<-data[,c("total_nid_end","total_qds_end","total_mis_end","total_add_end","total_sub_end","total_prb_end","total_nu_end")]
any(is.na(rowSums(end.num)))

## [1] FALSE

# no missings anymore


#endline lit
end.lit<-data[,c("total_lid_end","total_pho_end","total_muw_end","total_non_end","total_com_end","total_lis_end","total_wrt_end")]


# Are there any missings values?
any(is.na(rowSums(end.lit)))

## [1] TRUE

# TRUE, so missing values: multiple imputation

mis.lit=rep(NA,7)
for(i in 1:ncol(end.lit))
  mis.lit[i]=length(which(is.na(end.lit[,i])))

mis.lit

## [1]  69  69  69  69 416  69  69

# 69 participants miss the endline data for literacy all tasks. This is 69/1507=.046, 4.6%
# Task total_com has 416 missings values, this is 416/1507=.276, 27.6%
# These endline scores were imputed.

scores.mice<-mice(end.lit,print=FALSE)
end.lit.MI<-complete(scores.mice)

data[,c("total_lid_end","total_pho_end","total_muw_end","total_non_end","total_com_end","total_lis_end","total_wrt_end")]=end.lit.MI


#check
end.lit<-data[,c("total_lid_end","total_pho_end","total_muw_end","total_non_end","total_com_end","total_lis_end","total_wrt_end")]
any(is.na(rowSums(end.lit)))

## [1] FALSE

# no missings anymore

Long datafile

### Long datafile ###

baseline<-cbind(data[,c(2:5,616)],data[,c(colnames(bas.num),colnames(bas.lit))])
colnames(baseline)

##  [1] "participantID" "GenderChild"   "School"        "ppt_group"    
##  [5] "Attendance"    "total_nid"     "total_qds"     "total_mis"    
##  [9] "total_add"     "total_sub"     "total_prb"     "total_nu"     
## [13] "total_lid"     "total_pho"     "total_muw"     "total_non"    
## [17] "total_com"     "total_lis"     "total_wrt"

endline<-cbind(baseline[,1:5],data[,c(colnames(end.num),colnames(end.lit))])
colnames(endline)=colnames(baseline)

longdata<-rbind(baseline,endline)
longdata<-longdata[order(longdata$participantID),]
longdata$time<-rep(c("baseline","endline"),nrow(baseline))

longdata[c(1:5),]

##      participantID GenderChild                    School ppt_group Attendance
## 1209      16000650        Male Kyamusooni Primary School         B          0
## 2716      16000650        Male Kyamusooni Primary School         B          0
## 1163      16000674        Male Kyamusooni Primary School         B          1
## 2670      16000674        Male Kyamusooni Primary School         B          1
## 1115      16000681        Male Kyamusooni Primary School         B          0
##      total_nid total_qds total_mis total_add total_sub total_prb total_nu
## 1209        11         6         4         3         4         2       30
## 2716        14         9         5         4         4         4       40
## 1163        13        10         6         7         3         6       45
## 2670        15        10         7         7         6         6       51
## 1115        15        10        10         7         4         6       52
##      total_lid total_pho total_muw total_non total_com total_lis total_wrt
## 1209        21         1         3         1         0         1         0
## 2716        22         3         5         0         0         1         0
## 1163        23         0        13         3         0         0         1
## 2670        26         6        17         8         2         1         7
## 1115        24         4        19         9         2         1         3
##          time
## 1209 baseline
## 2716  endline
## 1163 baseline
## 2670  endline
## 1115 baseline

Descriptives

data.table(dl[-c(1,2),],keep.rownames=TRUE)

##                 rn Baseline A Endline A Baseline B Endline B
##  1: mean total_lid      41.47     53.03      38.93     48.05
##  2:   sd total_lid      12.61     10.77      12.32     11.69
##  3: mean total_pho      19.39     23.70      20.41     23.08
##  4:   sd total_pho       8.14      4.37       7.19      4.99
##  5: mean total_muw       3.71      6.73       2.16      2.88
##  6:   sd total_muw       3.51      3.05       2.83      3.13
##  7: mean total_non       8.37     13.77       7.46     12.07
##  8:   sd total_non       6.91      6.29       6.54      6.99
##  9: mean total_com       3.73      7.92       2.81      5.67
## 10:   sd total_com       4.72      5.31       4.15      5.26
## 11: mean total_lis       0.80      1.78       0.60      1.59
## 12:   sd total_lis       1.27      1.49       1.09      1.43
## 13: mean total_wrt       0.86      1.48       0.64      1.20
## 14:   sd total_wrt       1.26      1.42       1.04      1.32

dm<-t(descriptives_Num)
colnames(dm)<-c("Baseline A","Endline A","Baseline B","Endline B")
data.table(dm[-c(1,2),],keep.rownames=TRUE)

##                 rn Baseline A Endline A Baseline B Endline B
##  1: mean total_nid       0.18      0.18       0.25      0.25
##  2:   sd total_nid       0.39      0.39       0.43      0.43
##  3: mean total_qds      12.25     13.99      12.17     13.60
##  4:   sd total_qds       2.90      2.00       2.95      2.38
##  5: mean total_mis       8.29      9.40       8.25      9.01
##  6:   sd total_mis       2.05      1.23       2.14      1.61
##  7: mean total_add       5.82      7.72       5.27      7.05
##  8:   sd total_add       2.81      2.17       2.94      2.68
##  9: mean total_sub       6.97      9.92       5.83      8.47
## 10:   sd total_sub       4.10      4.29       3.70      4.40
## 11: mean total_prb       4.63      7.27       3.97      5.52
## 12:   sd total_prb       3.66      4.05       2.91      3.49
## 13:  mean total_nu       3.51      4.73       3.44      4.40
## 14:    sd total_nu       1.92      1.38       1.89      1.63

Analysis

nid

tab_model(fit_nid)

	total_nid
Predictors	Estimates	CI	p
(Intercept)	12.04	11.74 – 12.33	<0.001
GenderChild [Male]	0.28	0.02 – 0.54	0.037
time [endline]	1.79	1.54 – 2.03	<0.001
Attendance	0.45	0.10 – 0.79	0.011
ppt group [B]	-0.12	-0.49 – 0.25	0.515
GenderChild [Male] * time [endline]	-0.15	-0.42 – 0.13	0.304
time [endline] * Attendance	0.13	-0.20 – 0.47	0.441
time [endline] * ppt group [B]	-0.31	-0.59 – -0.04	0.027
Random Effects
σ2	3.75
τ00 participantID:School	2.81
τ00 School	0.13
ICC	0.44
N participantID	1507
N School	30
Observations	3014
Marginal R2 / Conditional R2	0.095 / 0.493

plot(colnames(dat)[9])

# The thin lines are the means of the schools, thick lines are the overall means.

qds

tab_model(fit_qds)

	total_qds
Predictors	Estimates	CI	p
(Intercept)	8.25	8.04 – 8.45	<0.001
GenderChild [Male]	-0.01	-0.19 – 0.17	0.940
time [endline]	1.13	0.94 – 1.31	<0.001
Attendance	0.22	-0.02 – 0.46	0.070
ppt group [B]	-0.05	-0.30 – 0.21	0.727
GenderChild [Male] * time [endline]	-0.08	-0.29 – 0.12	0.428
time [endline] * Attendance	0.15	-0.11 – 0.40	0.253
time [endline] * ppt group [B]	-0.36	-0.57 – -0.15	0.001
Random Effects
σ2	2.11
τ00 participantID:School	1.06
τ00 School	0.06
ICC	0.35
N participantID	1507
N School	30
Observations	3014
Marginal R2 / Conditional R2	0.073 / 0.396

plot(colnames(dat)[10])

# The thin lines are the means of the schools, thick lines are the overall means.

mis

tab_model(fit_mis)

	total_mis
Predictors	Estimates	CI	p
(Intercept)	5.63	5.30 – 5.96	<0.001
GenderChild [Male]	0.26	-0.01 – 0.53	0.056
time [endline]	2.04	1.80 – 2.28	<0.001
Attendance	0.35	-0.01 – 0.70	0.057
ppt group [B]	-0.59	-1.00 – -0.17	0.006
GenderChild [Male] * time [endline]	-0.34	-0.61 – -0.07	0.013
time [endline] * Attendance	0.14	-0.19 – 0.46	0.415
time [endline] * ppt group [B]	-0.12	-0.39 – 0.15	0.399
Random Effects
σ2	3.56
τ00 participantID:School	3.36
τ00 School	0.20
ICC	0.50
N participantID	1507
N School	30
Observations	3014
Marginal R2 / Conditional R2	0.121 / 0.561

plot(colnames(dat)[11])

# The thin lines are the means of the schools, thick lines are the overall means.

add

tab_model(fit_add)

	total_add
Predictors	Estimates	CI	p
(Intercept)	6.73	6.31 – 7.15	<0.001
GenderChild [Male]	0.23	-0.19 – 0.65	0.280
time [endline]	3.10	2.67 – 3.52	<0.001
Attendance	0.69	0.16 – 1.23	0.010
ppt group [B]	-1.20	-1.70 – -0.70	<0.001
GenderChild [Male] * time [endline]	-0.43	-0.91 – 0.05	0.082
time [endline] * Attendance	0.31	-0.27 – 0.89	0.298
time [endline] * ppt group [B]	-0.30	-0.79 – 0.18	0.215
Random Effects
σ2	11.24
τ00 participantID:School	5.57
τ00 School	0.15
ICC	0.34
N participantID	1507
N School	30
Observations	3014
Marginal R2 / Conditional R2	0.129 / 0.423

plot(colnames(dat)[12])

# The thin lines are the means of the schools, thick lines are the overall means.

sub

tab_model(fit_sub)

	total_sub
Predictors	Estimates	CI	p
(Intercept)	4.43	4.07 – 4.79	<0.001
GenderChild [Male]	0.24	-0.12 – 0.60	0.184
time [endline]	2.55	2.17 – 2.93	<0.001
Attendance	0.47	0.02 – 0.93	0.042
ppt group [B]	-0.71	-1.14 – -0.28	0.001
GenderChild [Male] * time [endline]	0.14	-0.29 – 0.56	0.521
time [endline] * Attendance	0.13	-0.39 – 0.64	0.634
time [endline] * ppt group [B]	-1.10	-1.53 – -0.67	<0.001
Random Effects
σ2	8.82
τ00 participantID:School	3.65
τ00 School	0.11
ICC	0.30
N participantID	1507
N School	30
Observations	3014
Marginal R2 / Conditional R2	0.114 / 0.379

plot(colnames(dat)[13])

# The thin lines are the means of the schools, thick lines are the overall means.

prb

tab_model(fit_prb)

	total_prb
Predictors	Estimates	CI	p
(Intercept)	3.33	3.08 – 3.57	<0.001
GenderChild [Male]	0.19	0.02 – 0.36	0.028
time [endline]	1.40	1.22 – 1.58	<0.001
Attendance	0.49	0.27 – 0.72	<0.001
ppt group [B]	-0.11	-0.43 – 0.20	0.484
GenderChild [Male] * time [endline]	-0.24	-0.44 – -0.04	0.020
time [endline] * Attendance	-0.35	-0.59 – -0.10	0.005
time [endline] * ppt group [B]	-0.22	-0.42 – -0.02	0.032
Random Effects
σ2	1.99
τ00 participantID:School	0.81
τ00 School	0.14
ICC	0.32
N participantID	1507
N School	30
Observations	3014
Marginal R2 / Conditional R2	0.103 / 0.394

plot(colnames(dat)[14])

# The thin lines are the means of the schools, thick lines are the overall means.

nu

tab_model(fit_nu)

	total_nu
Predictors	Estimates	CI	p
(Intercept)	40.38	39.02 – 41.75	<0.001
GenderChild [Male]	1.21	0.02 – 2.40	0.046
time [endline]	12.00	11.04 – 12.95	<0.001
Attendance	2.73	1.16 – 4.31	0.001
ppt group [B]	-2.78	-4.49 – -1.08	0.001
GenderChild [Male] * time [endline]	-1.10	-2.17 – -0.02	0.045
time [endline] * Attendance	0.50	-0.80 – 1.80	0.450
time [endline] * ppt group [B]	-2.42	-3.49 – -1.34	<0.001
Random Effects
σ2	56.14
τ00 participantID:School	81.10
τ00 School	2.91
ICC	0.60
N participantID	1507
N School	30
Observations	3014
Marginal R2 / Conditional R2	0.187 / 0.674

plot(colnames(dat)[15])

# The thin lines are the means of the schools, thick lines are the overall means.

lid

tab_model(fit_lid)

	total_lid
Predictors	Estimates	CI	p
(Intercept)	19.27	18.24 – 20.31	<0.001
GenderChild [Male]	-0.18	-0.80 – 0.45	0.581
time [endline]	4.46	3.83 – 5.10	<0.001
Attendance	1.12	0.27 – 1.97	0.010
ppt group [B]	0.95	-0.43 – 2.33	0.177
GenderChild [Male] * time [endline]	-0.17	-0.89 – 0.54	0.635
time [endline] * Attendance	-0.38	-1.25 – 0.49	0.394
time [endline] * ppt group [B]	-1.60	-2.32 – -0.88	<0.001
Random Effects
σ2	25.21
τ00 participantID:School	12.28
τ00 School	2.97
ICC	0.38
N participantID	1507
N School	30
Observations	3014
Marginal R2 / Conditional R2	0.077 / 0.425

plot(colnames(dat)[16])

# The thin lines are the means of the schools, thick lines are the overall means.

pho

tab_model(fit_pho)

	total_pho
Predictors	Estimates	CI	p
(Intercept)	3.86	3.20 – 4.51	<0.001
GenderChild [Male]	-0.33	-0.63 – -0.04	0.027
time [endline]	2.81	2.46 – 3.17	<0.001
Attendance	0.07	-0.33 – 0.47	0.721
ppt group [B]	-1.54	-2.44 – -0.64	0.001
GenderChild [Male] * time [endline]	0.09	-0.30 – 0.49	0.647
time [endline] * Attendance	0.90	0.42 – 1.38	<0.001
time [endline] * ppt group [B]	-2.37	-2.77 – -1.97	<0.001
Random Effects
σ2	7.69
τ00 participantID:School	0.76
τ00 School	1.41
ICC	0.22
N participantID	1507
N School	30
Observations	3014
Marginal R2 / Conditional R2	0.241 / 0.408

plot(colnames(dat)[17])

# The thin lines are the means of the schools, thick lines are the overall means.

muw

tab_model(fit_muw)

	total_muw
Predictors	Estimates	CI	p
(Intercept)	8.21	7.18 – 9.23	<0.001
GenderChild [Male]	-0.75	-1.40 – -0.11	0.023
time [endline]	5.25	4.83 – 5.67	<0.001
Attendance	2.93	2.03 – 3.83	<0.001
ppt group [B]	-1.08	-2.43 – 0.28	0.121
GenderChild [Male] * time [endline]	0.16	-0.31 – 0.64	0.498
time [endline] * Attendance	0.36	-0.22 – 0.93	0.223
time [endline] * ppt group [B]	-0.82	-1.29 – -0.34	0.001
Random Effects
σ2	10.90
τ00 participantID:School	29.36
τ00 School	2.79
ICC	0.75
N participantID	1507
N School	30
Observations	3014
Marginal R2 / Conditional R2	0.165 / 0.789

plot(colnames(dat)[18])

# The thin lines are the means of the schools, thick lines are the overall means.

non

tab_model(fit_non)

	total_non
Predictors	Estimates	CI	p
(Intercept)	3.56	2.97 – 4.15	<0.001
GenderChild [Male]	-0.00	-0.49 – 0.48	0.985
time [endline]	4.07	3.65 – 4.49	<0.001
Attendance	0.95	0.30 – 1.60	0.004
ppt group [B]	-0.99	-1.73 – -0.25	0.009
GenderChild [Male] * time [endline]	-0.17	-0.64 – 0.30	0.485
time [endline] * Attendance	1.08	0.51 – 1.66	<0.001
time [endline] * ppt group [B]	-1.38	-1.86 – -0.91	<0.001
Random Effects
σ2	10.91
τ00 participantID:School	12.09
τ00 School	0.61
ICC	0.54
N participantID	1507
N School	30
Observations	3014
Marginal R2 / Conditional R2	0.153 / 0.609

plot(colnames(dat)[19])

# The thin lines are the means of the schools, thick lines are the overall means.

com

tab_model(fit_com)

	total_com
Predictors	Estimates	CI	p
(Intercept)	0.76	0.57 – 0.94	<0.001
GenderChild [Male]	-0.03	-0.16 – 0.10	0.679
time [endline]	0.95	0.82 – 1.08	<0.001
Attendance	0.32	0.15 – 0.50	<0.001
ppt group [B]	-0.22	-0.46 – 0.02	0.072
GenderChild [Male] * time [endline]	-0.11	-0.25 – 0.03	0.133
time [endline] * Attendance	0.45	0.27 – 0.62	<0.001
time [endline] * ppt group [B]	-0.02	-0.16 – 0.12	0.799
Random Effects
σ2	0.99
τ00 participantID:School	0.64
τ00 School	0.08
ICC	0.42
N participantID	1507
N School	30
Observations	3014
Marginal R2 / Conditional R2	0.154 / 0.510

plot(colnames(dat)[20])

# The thin lines are the means of the schools, thick lines are the overall means.

lis

tab_model(fit_lis)

	total_lis
Predictors	Estimates	CI	p
(Intercept)	0.87	0.68 – 1.07	<0.001
GenderChild [Male]	-0.15	-0.27 – -0.02	0.019
time [endline]	0.56	0.44 – 0.68	<0.001
Attendance	0.31	0.14 – 0.48	<0.001
ppt group [B]	-0.23	-0.49 – 0.03	0.082
GenderChild [Male] * time [endline]	0.09	-0.05 – 0.22	0.216
time [endline] * Attendance	0.13	-0.04 – 0.30	0.126
time [endline] * ppt group [B]	-0.08	-0.21 – 0.06	0.282
Random Effects
σ2	0.92
τ00 participantID:School	0.56
τ00 School	0.10
ICC	0.42
N participantID	1507
N School	30
Observations	3014
Marginal R2 / Conditional R2	0.077 / 0.462

plot(colnames(dat)[21])

# The thin lines are the means of the schools, thick lines are the overall means.

wrt

tab_model(fit_wrt)

	total_wrt
Predictors	Estimates	CI	p
(Intercept)	3.06	2.60 – 3.52	<0.001
GenderChild [Male]	-0.47	-0.77 – -0.17	0.002
time [endline]	1.61	1.41 – 1.82	<0.001
Attendance	1.26	0.85 – 1.67	<0.001
ppt group [B]	-0.62	-1.23 – -0.01	0.045
GenderChild [Male] * time [endline]	-0.03	-0.26 – 0.20	0.800
time [endline] * Attendance	0.56	0.28 – 0.85	<0.001
time [endline] * ppt group [B]	0.27	0.04 – 0.51	0.022
Random Effects
σ2	2.67
τ00 participantID:School	5.74
τ00 School	0.56
ICC	0.70
N participantID	1507
N School	30
Observations	3014
Marginal R2 / Conditional R2	0.134 / 0.742

plot(colnames(dat)[22])

# The thin lines are the means of the schools, thick lines are the overall means.

Wellbeing