Figures 3a and 3b show the sample flow and analytical samples by trial and study arm during which updates during fieldwork (particularly in Kibera where there was a lag between the initial listing and the baseline survey) led to reduced samples excluding ineligible girls. Notably, accurately capturing age in both contexts benefited from multiple visits.
[FIG 3a. Kibera Sample Flow]
[FIG 3b. Wajir Sample Flow]
Baseline data included 2,390 girls in the Kibera site and 2,147 girls in the Wajir site. Baseline survey non-response was 6.2% in Kibera and 6.5% in Wajir and most commonly due to unavailability of the girl for interview (because she was not located, temporarily away or after the rapid listing it was determined that she was not a usual resident) and in some cases (about one-third in Kibera and 10% in Wajir), refusal. Because there was substantial residential mobility for girls in the study sites (particularly in Kibera), to keep attrition at a minimum the two-year follow-up survey tracked girls to their new locations. In Kibera this included following girls to 27 (out of 47) different counties throughout Kenya; about one-fifth of girls re-interviewed in the two-year follow-up were located outside of Nairobi City County. In Wajir mobility was lower and girls were tracked to new locations throughout the county as well as a small number outside the county; approximately 14% were interviewed in a location different from the baseline. At the two-year follow-up in 2017, 91.6% of girls from the Kibera baseline sample were re-interviewed and 88.9% from the Wajir sample. The final analysis sample included 2,190 girls from Kibera and 1,909 from Wajir. Reasons for loss to follow-up are shown in Figs. 3a and 3b. The most common were failure to locate girls who had moved or unavailability of the girl during the interview period. Direct refusal was more common in Kibera (particularly in the V-only arm), but rare in Wajir. In Wajir, the 15 girls in the cluster dropped from the trial in the VE arm due to security concerns were not followed.
Figure 3 also shows that two-year follow-up rates differed by study arm (85–95%); therefore, we tested baseline balance for selected background characteristics using the non-attrited sample. Table 2 (Kibera) and Table 3 (Wajir) present baseline means for all available outcomes for the sample of girls interviewed at the two-year follow-up (outcomes not included were the new measurements introduced in the two-year follow-up). The samples in the two-year follow-up in each study arm were balanced on a range of baseline characteristics and there were no large differences and only one significant difference for the general self-efficacy score in Wajir. Appendix Tables 2 and 3 report means by intervention for the full baseline sample [60].
We estimated linear probability models for each study site predicting re-interview in the two-year follow-up to examine its correlates and characterize potential attrition bias (Appendix Tables 4a and 4b). In addition to binary indicators for the study arms, the models included baseline measures of age, cognitive test score, school enrollment (Wajir only), grade attainment, parental education, co-residence with parents, household wealth and district (Wajir only). Conditional on those covariates, in Kibera the probability of re-interview was nearly 10 percentage points higher in all three arms compared to V-only (Appendix Table 4a, column 1) and higher for the youngest, those with marginally higher cognitive test scores and those who resided with both of their parents. Expanding the model to also include interactions of the baseline controls with an indicator for each study arm (column 2), however, indicates only one individually significant variable (paternal schooling) between the relationship of the baseline controls and attrition, and neither the joint test of all interactions nor joint tests of interactions for each study arm are significant. Conditional on the covariates, in Wajir the probability of re-interview was about 4 percentage points higher in the VEH and VEHW arms compared to V-only (Appendix Table 4b, column 1), and higher for the youngest, those initially enrolled in school or with higher grade attainment, and those living in East Wajir. Expanding the model to also include interactions of the baseline controls with an indicator for each study arm (column 2) indicates only a few significant differences between the relationship of the baseline controls and attrition. However, the joint test of all interactions and joint tests of interactions for the VEH and VEHW study arm are significant. The balance results across study arms and these patterns for attrition suggest that while there are some differences (especially in Wajir), large systematic biases in the ITT estimates threatening internal validity are unlikely. Nevertheless, to assess further the possibility of bias below we also characterize the results controlling for these additional covariates and also accounting for attrition using inverse probability weights.
At baseline in Kibera, girls averaged 12.6 years old and about half lived with their parents, a majority of whom had completed primary school. Nearly one-third of the girls had experienced violence perpetrated by a male in the past year, notably high given such violence is often underreported. Virtually all girls in Kibera were enrolled in school and average grade attainment was 5.7 years, two years below completed primary consistent with their ages. Over 90% were literate in Swahili and English and all but a few girls were expected to complete secondary school. Reproductive health knowledge was low, with fewer than one in ten girls identifying the most fertile period during the menstrual cycle. Girls scored on average 4 out of 6 in general self-efficacy; many of the hypothetical situations posed in the scale proved difficult for younger girls to understand in both sites. On average girls scored nearly 6 out of 10 on financial literacy and about one-quarter had saved any money in the past six months.
At baseline in Wajir, girls averaged 11.9 years old and about three-quarters lived with their parents, very few of whom had ever themselves attended school. Girls were younger in Wajir on average largely because there was no time gap between the household listing and baseline survey unlike in Kibera. In both settings some girls determined to be a little outside the initial age ranges after initial inclusion, for example during the two-year follow-up survey, were retained in the samples (and all analyses control for age). Less than 5% experienced violence perpetrated by a male in the past year, although as in Kibera underreporting is possible. Approximately three-quarters of girls were enrolled in school and average grade attainment was only 2.9 years, about half that of their counterparts in Kibera. Low levels of schooling were reflected in literacy with less than 40% literate in both Swahili and English. Although lower than in Kibera, however, more than 80% of girls were still expected by their household to complete secondary school. Reproductive health knowledge and general self-efficacy were even lower in Wajir than in Kibera: hardly any girls were able to identify the most fertile period during the menstrual cycle, and they scored on average 2 out of 6 in general self-efficacy. On average girls scored 4 out of 10 in financial literacy, and almost none had saved any money in the past six months.
Table 2. Kibera baseline means for key outcomes among analytical sample at two-year follow-up, by study arm
|
(1)
|
(2)
|
(3)
|
(4)
|
(5)
|
|
V-Only
|
VE
|
VEH
|
VEHW
|
p-value
|
Background
|
|
|
|
|
|
Age, mean (sd)
|
12.6 (1.2)
|
12.5 (1.3)
|
12.6 (1.3)
|
12.5 (1.3)
|
0.707
|
Cognitive score (0–16), mean (sd) [n=2,173]
|
8.2 (3.0)
|
8.4 (3.0)
|
8.3 (3.1)
|
8.3 (3.2)
|
0.732
|
Lives with both parents, % [n=2,174]
|
52.1
|
55.9
|
51.0
|
53.3
|
0.384
|
Mother completed primary school, % [n=2,040]
|
62.4
|
63.3
|
62.5
|
64.1
|
0.938
|
Father completed primary school, % [n=1,803]
|
77.7
|
78.9
|
74.7
|
79.6
|
0.305
|
Violence Prevention
|
|
|
|
|
|
Experienced violence by a male in the past year, %
|
29.0
|
29.8
|
30.6
|
32.2
|
0.703
|
Positive gender attitudes score (0-4), mean (sd)
|
3.6 (0.7)
|
3.5 (0.7)
|
3.6 (0.7)
|
3.6 (0.7)
|
0.458
|
Education
|
|
|
|
|
|
Grade attainment, mean (sd)
|
5.7 (1.4)
|
5.7 (1.3)
|
5.7 (1.4)
|
5.7 (1.3)
|
0.990
|
Primary school completion, %
|
7.9
|
5.9
|
6.7
|
6.1
|
0.600
|
Enrolled in current school year, %
|
99.2
|
99.1
|
98.6
|
99.3
|
0.713
|
Literate in Swahili and English, % [n=2,173]
|
92.2
|
93.6
|
92.9
|
94.0
|
0.700
|
Health
|
|
|
|
|
|
Knows most fertile period during menstrual cycle, %
|
8.7
|
8.2
|
7.0
|
5.8
|
0.243
|
General self-efficacy score (0-6), mean (sd)
|
3.8 (1.7)
|
4.0 (1.6)
|
3.9 (1.6)
|
4.0 (1.6)
|
0.493
|
Wealth Creation
|
|
|
|
|
|
Financial literacy score (0-10), mean (sd)
|
5.8 (1.9)
|
5.6 (1.9)
|
5.7 (1.9)
|
5.8 (1.9)
|
0.426
|
Saved money in the past six months, %
|
27.8
|
25.7
|
26.4
|
28.4
|
0.722
|
Household-level
|
|
|
|
|
|
Household expects girl to complete secondary, % [n=2,222]
|
99.8
|
99.6
|
99.8
|
99.8
|
0.934
|
Household wealth quintile (1-5), mean (sd) [n=2,236]
|
3.1 (1.4)
|
3.1 (1.4)
|
3.0 (1.4)
|
3.0 (1.4)
|
0.631
|
|
|
|
|
|
|
Sample by arm when n=2,190
|
507
|
561
|
569
|
553
|
|
Notes: P-values from F tests for joint differences across study arms to test baseline balance for the sample of non-attritors in the two-year follow-up. n = 2,190 unless otherwise indicated.
Table 3. Wajir baseline means for key outcomes among analytical sample at two-year follow-up, by study arm
|
(1)
|
(2)
|
(3)
|
(4)
|
(5)
|
|
V-Only
|
VE
|
VEH
|
VEHW
|
p-value
|
Background
|
|
|
|
|
|
Age, mean (sd)
|
11.9 (1.3)
|
12.0 (1.3)
|
11.8 (1.2)
|
11.8 (1.3)
|
0.128
|
Cognitive score (0–16), mean (sd) [n=1,874]
|
5.3 (2.9)
|
5.0 (3.0)
|
5.0 (3.2)
|
5.4 (3.0)
|
0.577
|
Lives with both parents, % [n=1,892]
|
73.6
|
73.9
|
76.2
|
73.7
|
0.846
|
Mother ever attended school, % [n=1,892]
|
1.4
|
1.5
|
0.6
|
0.4
|
0.193
|
Father ever attended school, % [n=1,888]
|
5.7
|
2.8
|
3.3
|
3.2
|
0.314
|
Violence Prevention
|
|
|
|
|
|
Experienced violence by a male in the past year, % [n=1,883]
|
4.1
|
3.9
|
3.1
|
2.2
|
0.444
|
Positive gender attitudes score (0-4), mean (sd) [1,883]
|
3.0 (1.0)
|
2.8 (1.1)
|
2.9 (1.1)
|
2.9 (1.1)
|
0.509
|
Education
|
|
|
|
|
|
Grade attainment, mean (sd)
|
3.1 (2.1)
|
2.8 (2.3)
|
2.8 (2.2)
|
2.6 (2.1)
|
0.467
|
Primary school completion, %
|
1.3
|
0.6
|
0.8
|
0.8
|
0.890
|
Enrolled in current school year, %
|
83.4
|
72.8
|
79.6
|
74.1
|
0.141
|
Literate in Swahili and English, % [n=1,874]
|
41.6
|
34.8
|
40.3
|
34.5
|
0.417
|
Health
|
|
|
|
|
|
Knows most fertile period during menstrual cycle,% [n=1,883]
|
1.6
|
0.4
|
0.8
|
0.8
|
0.298
|
General self-efficacy score (0-6), mean (sd) [n=1,883]
|
2.3 (2.0)
|
2.3 (1.9)
|
2.2 (1.9)
|
1.7 (2.0)
|
0.007
|
Wealth Creation
|
|
|
|
|
|
Financial literacy score (0-10), mean (sd) [n=1,883]
|
4.5 (2.0)
|
4.2 (2.1)
|
4.1 (2.0)
|
4.1 (2.0)
|
0.571
|
Saved money in the past six months, % [n=1,883]
|
0.7
|
0.4
|
0.4
|
0.8
|
0.837
|
Household-level
|
|
|
|
|
|
Household expects girl to complete secondary, % [n=2,008]
|
86.8
|
80.8
|
81.7
|
81.7
|
0.347
|
Household wealth quintile (1-5), mean (sd) [n=2,011]
|
3.0 (1.4)
|
2.7 (1.5)
|
2.9 (1.5)
|
2.9 (1.4)
|
0.791
|
|
|
|
|
|
|
Sample by arm when n=1,909
|
447
|
471
|
490
|
501
|
|
Notes: P-values from F tests for joint differences across study arms accounting for clustering at the village level to test baseline balance for the sample of non-attritors in the two-year follow-up. n = 1,909 unless otherwise indicated.
Table 4 presents results from the ITT analyses for the Kibera site. The first column displays the means of the two-year follow-up outcomes for the V-only study arm. The next three columns present the ITT estimates relative to V-only for the VE, VEH and VEHW study arms allowing assessment of the program effects on the secondary outcomes for each study arm. To explore whether there were incremental effects across study arms with additional interventions, in particular assessment of cash versus cash plus programming, the final three columns 5–7 present differences between pairs of study arms with education intervention, based on the ITT estimates in columns 2 to 4.
The intervention led to reductions in the experience of male-perpetrated violence between 4–9 percentage points compared with an average of 42% in the V-only arm, statistically significant for the VE and VEH arms, but VEH is not significant after consideration of multiple hypothesis testing based on the FDR q-value (Appendix Table 5). There was a modest reduction in gender schooling attitudes of about 0.1 SD in VEHW, but no effect on overall gender equitable attitudes or on the violence prevention summary z-score (Table 5, bottom panel). For schooling, there were small positive effects on grade attainment of about 0.05 grades (only significant for VEHW) and small positive but insignificant effects on primary school completion and enrollment. Examining these three indicators together in a summary variable suggests an increase of about 0.1 SD for the VEHW arm only. Consideration of the conditional schooling variables indicates VEHW increased the probability of completing primary school and, separately, transitioning to secondary school by 7 percentage points. Positive and statistically significant effects are found for the VEH and VEHW arms in SRH knowledge outcomes and condom self-efficacy. There was a greater than 10 percentage point increase in knowledge of at least one method of modern contraception compared with 56% knowledge in V-only, and a 0.2 SD increase in general SRH knowledge and more than 0.1 SD increase in condom self-efficacy. A positive and statistically significant effect of more than 0.1 SD is observed for general self-efficacy in the VE arm. Estimation of the impact on the summary health outcome average z-score yields significant increases of about 0.3 SD for the VEH and VEHW study arms. Large positive and statistically significant effects on both financial literacy and saving behavior are found for the VEHW arm (and correspondingly in the summary measure), but not elsewhere. For example, savings increased by 20 percentage points compared with 45% in V-only. There were no effects of the interventions on the household-level wealth quintile and, unsurprisingly given the near-100 percent levels at baseline, no effect on household expectations the girl would complete secondary school.
All results are robust to the inclusion of additional controls and to weighting for attrition, with significant point estimates changing only marginally and remaining significant at a 0.05 level (Appendix Table 5). The results for summary variables are also robust to these variations (Appendix Table 6). In addition, consideration of the combined education treatments (Appendix Table 6, column 5) indicate for all but the violence prevention domain positive significant effects of nearly 0.1 SD for education and 0.2 SD for health and wealth creation.
Examining columns 5–7 in Table 4, there were few significant differences between the effects of study arms with the education intervention for violence prevention and education outcomes. Several health indicators, including the summary indicator, were significantly larger in the VEH and VEHW arms compared with the VE arm. One notable exception is general self-efficacy which showed improvement only in the VE study arm. The two wealth creation indicators (and the summary measure) were larger in the VEHW compared with the VE and VEH arms.
Table 4: Kibera estimated effects of intent-to-treat
|
(1)
|
(2)
|
(3)
|
(4)
|
(5)
|
(6)
|
(7)
|
|
V-Only two-year follow-up
Mean
|
VE
Estimate
|
VEH
Estimate
|
VEHW
Estimate
|
VEH vs
VE
(3)-(2)
|
VEHW vs
VE
(4)-(2)
|
VEHW vs
VEH
(4)-(3)
|
Violence Prevention
|
|
|
|
|
|
|
|
Experienced violence by a male in the past year (=1)
|
0.422
|
-0.088**
|
-0.059*
|
-0.042
|
0.028
|
0.046
|
0.018
|
95% CI
|
[0.38, 0.47]
|
[-0.14, -0.03]
|
[-0.12, 0.00]
|
[-0.10, 0.02]
|
[-0.03, 0.08]
|
[-0.01, 0.10]
|
[-0.04, 0.07]
|
Gender equitable attitudes z-score1
|
0.000
|
-0.036
|
0.054
|
0.078
|
0.090
|
0.114*
|
0.024
|
95% CI
|
[-0.09, 0.09]
|
[-0.15, 0.08]
|
[-0.07, 0.17]
|
[-0.04, 0.19]
|
[-0.02, 0.20]
|
[0.00, 0.23]
|
[-0.09, 0.14]
|
Positive gender schooling attitudes z-score
|
0.211
|
-0.024
|
-0.091†
|
-0.137*
|
-0.067
|
-0.113*
|
-0.046
|
95% CI
|
[0.14, 0.28]
|
[-0.12, 0.07]
|
[-0.19, 0.01]
|
[-0.24, -0.03]
|
[-0.17, 0.03]
|
[-0.22, -0.01]
|
[-0.15, 0.06]
|
Education
|
|
|
|
|
|
|
|
Grade attainment
|
7.501
|
0.052†
|
0.048†
|
0.067*
|
-0.003
|
0.016
|
0.019
|
95% CI
|
[7.38, 7.62]
|
[0.00, 0.11]
|
[-0.01, 0.10]
|
[0.01, 0.12]
|
[-0.05, 0.05]
|
[-0.03, 0.06]
|
[-0.03, 0.07]
|
Primary school complete (=1)
|
0.517
|
0.020
|
0.017
|
0.032
|
-0.004
|
0.011
|
0.015
|
95% CI
|
[0.47, 0.56]
|
[-0.03, 0.07]
|
[-0.03, 0.07]
|
[-0.02, 0.08]
|
[-0.05, 0.05]
|
[-0.04, 0.06]
|
[-0.03, 0.06]
|
Enrolled in current school year (=1)
|
0.959
|
0.005
|
0.011
|
0.020†
|
0.006
|
0.015
|
0.010
|
95% CI
|
[0.94, 0.98]
|
[-0.02, 0.03]
|
[-0.01, 0.03]
|
[0.00, 0.04]
|
[-0.01, 0.03]
|
[0.00, 0.03]
|
[-0.01, 0.03]
|
Conditional primary school complete (=1)2 [n=1,104]
|
0.887
|
0.037
|
0.020
|
0.070**
|
-0.016
|
0.033
|
0.050*
|
95% CI
|
[0.85, 0.93]
|
[-0.01, 0.09]
|
[-0.03, 0.07]
|
[0.02, 0.12]
|
[-0.06, 0.03]
|
[-0.01, 0.07]
|
[0.01, 0.09]
|
Conditional transition to secondary school (=1)3 [n=1,131]
|
0.836
|
0.036
|
0.035
|
0.071*
|
-0.001
|
0.036
|
0.037
|
95% CI
|
[0.79, 0.88]
|
[-0.02, 0.10]
|
[-0.02, 0.09]
|
[0.01, 0.13]
|
[-0.06, 0.05]
|
[-0.02, 0.09]
|
[-0.01, 0.09]
|
Health
|
|
|
|
|
|
|
|
Knows most fertile period during menstrual cycle (=1)
|
0.087
|
0.016
|
0.019
|
0.022
|
0.003
|
0.006
|
0.003
|
95% CI
|
[0.06, 0.11]
|
[-0.02, 0.05]
|
[-0.02, 0.05]
|
[-0.01, 0.06]
|
[-0.03, 0.04]
|
[-0.03, 0.04]
|
[-0.03, 0.04]
|
Knows method of modern contraception1 (=1) [n=2,175]
|
0.555
|
-0.007
|
0.130***
|
0.119***
|
0.137***
|
0.125***
|
-0.011
|
95% CI
|
[0.51, 0.60]
|
[-0.06, 0.05]
|
[0.07, 0.19]
|
[0.06, 0.18]
|
[0.08, 0.19]
|
[0.07, 0.18]
|
[-0.06, 0.04]
|
SRH myths knowledge z-score1 [n=1,948]
|
0.000
|
0.016
|
0.213**
|
0.158*
|
0.197**
|
0.142*
|
-0.055
|
95% CI
|
[-0.09, 0.09]
|
[-0.11, 0.14]
|
[0.09, 0.34]
|
[0.03, 0.29]
|
[0.08, 0.31]
|
[0.02, 0.26]
|
[-0.17, 0.07]
|
General self-efficacy z-score
|
0.325
|
0.152**
|
0.030
|
0.055
|
-0.122*
|
-0.097*
|
0.025
|
95% CI
|
[0.25, 0.40]
|
[0.06, 0.25]
|
[-0.07, 0.13]
|
[-0.05, 0.16]
|
[-0.22, -0.03]
|
[-0.19, 0.00]
|
[-0.08, 0.12]
|
Condom use self-efficacy z-score1 [n=1,777]
|
0.000
|
0.038
|
0.178**
|
0.124†
|
0.140*
|
0.086
|
-0.054
|
95% CI
|
[-0.10, 0.10]
|
[-0.10, 0.17]
|
[0.05, 0.30]
|
[-0.01, 0.25]
|
[0.02, 0.26]
|
[-0.04, 0.21]
|
[-0.17, 0.06]
|
Wealth creation
|
|
|
|
|
|
|
|
Financial literacy z-score
|
-0.085
|
0.014
|
0.045
|
0.381***
|
0.030
|
0.366***
|
0.336***
|
95% CI
|
[-0.17, 0.00]
|
[-0.11, 0.13]
|
[-0.07, 0.16]
|
[0.26, 0.50]
|
[-0.09, 0.15]
|
[0.25, 0.48]
|
[0.22, 0.45]
|
Saved money in the past six months (=1)
|
0.448
|
0.000
|
0.017
|
0.202***
|
0.017
|
0.202***
|
0.185***
|
95% CI
|
[0.40, 0.49]
|
[-0.06, 0.06]
|
[-0.04, 0.08]
|
[0.14, 0.26]
|
[-0.04, 0.07]
|
[0.15, 0.26]
|
[0.13, 0.24]
|
Household-level outcomes
|
|
|
|
|
|
|
|
Household wealth quintile [n=2,236]
|
2.823
|
0.101
|
0.123
|
0.105
|
0.022
|
0.004
|
-0.019
|
95% CI
|
[2.70, 2.95]
|
[-0.06, 0.26]
|
[-0.03, 0.28]
|
[-0.06, 0.27]
|
[-0.13, 0.18]
|
[-0.15, 0.16]
|
[-0.17, 0.14]
|
Expected girl to complete secondary school (=1) [n=2,193]
|
0.994
|
0.002
|
-0.001
|
0.004
|
-0.003
|
0.002
|
0.005
|
95% CI
|
[0.99, 1.00]
|
[-0.01, 0.01]
|
[-0.01, 0.01]
|
[0.00, 0.01]
|
[-0.01, 0.01]
|
[0.00, 0.01]
|
[0.00, 0.01]
|
Summary index z-scores
|
|
|
|
|
|
|
|
Violence prevention outcomes summary index z-score4
|
0.000
|
0.066
|
0.037
|
-0.004
|
-0.029
|
-0.070
|
-0.040
|
95% CI
|
[-0.09, 0.09]
|
[-0.05, 0.19]
|
[-0.08, 0.16]
|
[-0.13, 0.12]
|
[-0.15, 0.09]
|
[-0.19, 0.05]
|
[-0.16, 0.08]
|
Education outcomes summary index z-score (grade, primary, enroll)
|
0.000
|
0.062
|
0.082
|
0.123*
|
0.020
|
0.061
|
0.041
|
95% CI
|
[-0.09, 0.09]
|
[-0.04, 0.17]
|
[-0.02, 0.18]
|
[0.03, 0.22]
|
[-0.07, 0.12]
|
[-0.03, 0.15]
|
[-0.05, 0.13]
|
Health outcomes summary index z-score
|
0.000
|
0.113†
|
0.306***
|
0.279***
|
0.193**
|
0.167**
|
-0.026
|
95% CI
|
[-0.09, 0.09]
|
[0.00, 0.23]
|
[0.19, 0.42]
|
[0.16, 0.40]
|
[0.08, 0.31]
|
[0.05, 0.28]
|
[-0.14, 0.09]
|
Wealth creation outcomes summary index z-score
|
0.000
|
0.015
|
0.056
|
0.517***
|
0.041
|
0.502***
|
0.461***
|
95% CI
|
[-0.09, 0.09]
|
[-0.10, 0.13]
|
[-0.06, 0.17]
|
[0.40, 0.63]
|
[-0.07, 0.16]
|
[0.39, 0.62]
|
[0.35, 0.57]
|
|
|
|
|
|
|
|
|
Notes: The table reports two-year follow-up means for V-only in column 1, the estimated ITT effect for each study arm relative to V-only in columns 2–4 and differences in the estimated ITT effects across study arms in columns 5–7. Column 5 compares the estimates for VEH to VE, column 6 compares VEHW to VE, and column 7 compares VEHW to VEH. For example, the estimate in column 7 for ‘Experienced violence by a male in the past year’ (0.018) is the difference between the estimate for VEHW (-0.042) in column 4 and the estimate for VEH (-0.059) in column 3. Minor differences in the reported differentials compared to the estimates presented in columns 2–4 are due to rounding. Numbers in square brackets indicate 95% confidence intervals. Regressions were estimated with robust standard errors and included controls for age and the outcome measured at baseline unless otherwise noted. N = 2,190; sample is smaller for some individual outcomes due to missing data as indicated.
1 No baseline control for outcome variable available.
2 Among girls who had completed Class 6 but had not yet completed Class 8 at baseline.
3 Among girls who had completed Class 6 but had not yet enrolled in secondary school at baseline.
4 Violence prevention indicator reverse coded prior to inclusion in summary variable.
*** p<0.001, ** p<0.01, * p<0.05, † p<0.1
Table 5 presents parallel results from the ITT analyses for the Wajir site. There were few effects on outcomes related to violence prevention and after accounting for multiple hypothesis testing no significant effects (Appendix Table 7)., unsurprising given the research design. There were large significant effects on schooling, approximately 0.2 additional grades and increased enrollment of 7–14 percentage points compared to an enrollment rate of 80% in V-only. This improvement is on par with some of the most effective CCT programs for schooling elsewhere [53]. There was no increase, however, in completed primary school, likely reflecting the low starting levels of schooling for most girls at the start of the intervention. The point estimates for effects on health outcomes suggest mixed results, although none of the negative estimates for knowing the most fertile period or a modern method of contraception are significant after accounting for multiple hypothesis testing. There were substantial significant improvements for sexual and reproductive knowledge of 0.3 SD or more for VE and VEH. There were no effects on financial literacy but positive and statistically significant effects for the VEH (not significant after accounting for multiple hypothesis testing) and VEHW arms in savings behavior, and the effect for VEHW was a substantial 40 percentage point increase on a base of under 2% in V-only. At the household level, there was an apparent unexpected decline in wealth quintile for the VE study arm, not significant after accounting for multiple hypothesis testing. A robust significant increase of about 5 percentage points (compared with 86% in V-only) was seen for household expectations the girl would complete secondary school for the VE study arm.
Significant results are robust to the inclusion of additional controls and to weighting for attrition, with point estimates and significance levels only changing marginally for the majority of outcomes (Appendix Table 7). The results for summary variables are also robust to these variations (Appendix Table 8). In addition, consideration of the combined education treatments (Appendix Table 8, column 5) indicates positive significant effects of 0.2 SD for education and more than 1 SD for wealth creation.
There are only a few significantly different effects between study arms with the education intervention in Wajir in columns 5–7 of Table 5. There was a significant difference between VEHW and VEH in violence experienced by a male, mirrored by a significant difference between the study arms in the violence prevention summary measure. Reflecting impacts on enrollment in the VE arm twice the size of the others, both VEH and VEHW had significantly lower impacts on enrollment (6–7 percentage points). Effects on financial literacy and savings behavior were substantially larger in the VEHW arm compared to the other two study arms, and most significant.
Table 5: Wajir estimated effects of intent-to-treat
|
(1)
|
(2)
|
(3)
|
(4)
|
(5)
|
(6)
|
(7)
|
|
V-Only two-year follow-up
Mean
|
VE
Estimate
|
VEH
Estimate
|
VEHW
Estimate
|
VEH vs
VE
(3)-(2)
|
VEHW vs
VE
(4)-(2)
|
VEHW vs
VEH
(4)-(3)
|
Violence Prevention
|
|
|
|
|
|
|
|
Experienced violence by a male in the past year (=1) [n=1,878]
|
0.038
|
-0.006
|
0.015
|
-0.022
|
0.021
|
-0.017
|
-0.037*
|
95% CI
|
[0.01, 0.06]
|
[-0.04, 0.03]
|
[-0.02, 0.05]
|
[-0.05, 0.01]
|
[-0.02, 0.06]
|
[-0.05, 0.01]
|
[-0.07, -0.01]
|
Gender equitable attitudes z-score1 [n=1,903]
|
0.000
|
-0.171
|
-0.208*
|
-0.018
|
-0.037
|
0.153
|
0.190†
|
95% CI
|
[-0.17, 0.17]
|
[-0.39, 0.05]
|
[-0.39, -0.02]
|
[-0.26, 0.22]
|
[-0.23, 0.15]
|
[-0.09, 0.39]
|
[-0.02, 0.40]
|
Positive gender schooling attitudes z-score [n=1,878]
|
0.160
|
0.156*
|
0.099
|
0.068
|
-0.057
|
-0.088
|
-0.031
|
95% CI
|
[0.05, 0.27]
|
[0.00, 0.31]
|
[-0.05, 0.25]
|
[-0.08, 0.21]
|
[-0.22, 0.10]
|
[-0.23, 0.06]
|
[-0.18, 0.12]
|
Education
|
|
|
|
|
|
|
|
Grade attainment
|
4.492
|
0.259**
|
0.194*
|
0.181*
|
-0.065
|
-0.079
|
-0.013
|
95% CI
|
[3.98, 5.01]
|
[0.10, 0.42]
|
[0.02, 0.36]
|
[0.01, 0.35]
|
[-0.23, 0.10]
|
[-0.24, 0.09]
|
[-0.19, 0.17]
|
Primary school complete (=1)
|
0.130
|
0.010
|
0.021
|
-0.029
|
0.011
|
-0.039
|
-0.050
|
95% CI
|
[0.08, 0.18]
|
[-0.05, 0.07]
|
[-0.05, 0.09]
|
[-0.08, 0.03]
|
[-0.06, 0.08]
|
[-0.10, 0.02]
|
[-0.11, 0.01]
|
Enrolled in current school year (=1)
|
0.808
|
0.144***
|
0.070*
|
0.084**
|
-0.074*
|
-0.060*
|
0.014
|
95% CI
|
[0.74, 0.87]
|
[0.09, 0.20]
|
[0.01, 0.13]
|
[0.02, 0.14]
|
[-0.13, -0.02]
|
[-0.11, -0.01]
|
[-0.05, 0.08]
|
Health
|
|
|
|
|
|
|
|
Knows most fertile period during menstrual cycle (=1) [n=1,878]
|
0.049
|
-0.031†
|
-0.021
|
-0.013
|
0.010
|
0.018
|
0.008
|
95% CI
|
[0.02, 0.08]
|
[-0.06, 0.00]
|
[-0.05, 0.01]
|
[-0.04, 0.02]
|
[-0.01, 0.03]
|
[-0.01, 0.04]
|
[-0.01, 0.03]
|
Knows method of modern contraception1 (=1) [n=1,848]
|
0.390
|
-0.097†
|
-0.103*
|
-0.096†
|
-0.007
|
0.001
|
0.007
|
95% CI
|
[0.31, 0.47]
|
[-0.21, 0.02]
|
[-0.19, -0.01]
|
[-0.20, 0.01]
|
[-0.10, 0.09]
|
[-0.11, 0.11]
|
[-0.08, 0.10]
|
SRH myths knowledge z-score1,2 [n=1,400]
|
0.000
|
0.344*
|
0.433**
|
0.234†
|
0.090
|
-0.110
|
-0.199
|
95% CI
|
[-0.20, 0.20]
|
[0.08, 0.60]
|
[0.17, 0.70]
|
[-0.03, 0.50]
|
[-0.17, 0.35]
|
[-0.37, 0.15]
|
[-0.47, 0.07]
|
General self-efficacy z-score [n=1,878]
|
0.990
|
-0.054
|
-0.014
|
-0.030
|
0.040
|
0.024
|
-0.016
|
95% CI
|
[0.86, 1.12]
|
[-0.25, 0.14]
|
[-0.22, 0.19]
|
[-0.22, 0.16]
|
[-0.17, 0.25]
|
[-0.17, 0.21]
|
[-0.21, 0.18]
|
Wealth creation
|
|
|
|
|
|
|
|
Financial literacy z-score [n=1,878]
|
0.439
|
-0.069
|
0.030
|
0.185
|
0.099
|
0.254*
|
0.155
|
95% CI
|
[0.25, 0.63]
|
[-0.31, 0.17]
|
[-0.19, 0.25]
|
[-0.06, 0.43]
|
[-0.08, 0.28]
|
[0.04, 0.46]
|
[-0.04, 0.35]
|
Saved money in the past six months (=1) [n=1,878]
|
0.011
|
0.030
|
0.047*
|
0.409***
|
0.017
|
0.379***
|
0.362***
|
95% CI
|
[0.00, 0.02]
|
[-0.01, 0.07]
|
[0.01, 0.09]
|
[0.31, 0.51]
|
[-0.03, 0.06]
|
[0.27, 0.48]
|
[0.26, 0.47]
|
Household-level outcomes
|
|
|
|
|
|
|
|
Household wealth quintile [n=2,011]
|
3.226
|
-0.434*
|
-0.247
|
-0.211
|
0.188
|
0.223
|
0.035
|
95% CI
|
[2.93, 3.52]
|
[-0.83, -0.04]
|
[-0.64, 0.14]
|
[-0.54, 0.12]
|
[-0.25, 0.62]
|
[-0.16, 0.60]
|
[-0.34, 0.41]
|
Expected girl to complete secondary school (=1) [n=2,007]
|
0.861
|
0.058*
|
0.050†
|
0.043
|
-0.008
|
-0.015
|
-0.007
|
95% CI
|
[0.82, 0.91]
|
[0.01, 0.10]
|
[0.00, 0.10]
|
[-0.01, 0.10]
|
[-0.05, 0.04]
|
[-0.06, 0.03]
|
[-0.06, 0.05]
|
Summary index z-scores
|
|
|
|
|
|
|
|
Violence prevention outcomes summary index z-score3 [n=1,878]
|
0.000
|
0.030
|
-0.101
|
0.104
|
-0.131†
|
0.073
|
0.205**
|
95% CI
|
[-0.15, 0.15]
|
[-0.15, 0.21]
|
[-0.29, 0.08]
|
[-0.07, 0.27]
|
[-0.29, 0.02]
|
[-0.06, 0.21]
|
[0.06, 0.35]
|
Education outcomes summary index z-score
|
0.000
|
0.302***
|
0.189*
|
0.122
|
-0.114
|
-0.180*
|
-0.067
|
95% CI
|
[-0.16, 0.16]
|
[0.16, 0.44]
|
[0.02, 0.36]
|
[-0.03, 0.27]
|
[-0.29, 0.07]
|
[-0.33, -0.03]
|
[-0.25, 0.12]
|
Health outcomes summary index z-score [n=1,878]
|
0.000
|
-0.026
|
0.066
|
-0.014
|
0.092
|
0.012
|
-0.080
|
95% CI
|
[-0.17, 0.17]
|
[-0.22, 0.17]
|
[-0.13, 0.26]
|
[-0.22, 0.19]
|
[-0.06, 0.24]
|
[-0.15, 0.17]
|
[-0.24, 0.08]
|
Wealth creation outcomes summary index z-score [n=1,878]
|
0.000
|
0.167
|
0.349*
|
2.852***
|
0.182
|
2.686***
|
2.504***
|
95% CI
|
[-0.15, 0.15]
|
[-0.14, 0.47]
|
[0.03, 0.66]
|
[2.14, 3.56]
|
[-0.15, 0.51]
|
[1.97, 3.40]
|
[1.79, 3.22]
|
|
|
|
|
|
|
|
|
Notes: The table reports two-year follow-up means for V-only in column 1, the estimated ITT effect for each study arm relative to V-only in columns 2–4 and differences in the estimated ITT effects across study arms in columns 5–7. Column 5 compares the estimates for VEH to VE, column 6 compares VEHW to VE, and column 7 compares VEHW to VEH. For example, the estimate in column 7 for ‘Experienced violence by a male in the past year’ (-0.037) is the difference between the estimate for VEHW (-0.022) in column 4 and the estimate for VEH (0.015) in column 3. Minor differences in the reported differentials compared to the estimates presented in columns 2–4 are due to rounding. Numbers in square brackets indicate 95% confidence intervals. Regressions were estimated with standard errors clustered at the village level and included controls for 2009 district per the stratified randomization, age and the outcome measured at baseline unless otherwise noted. N = 1,909; sample is smaller for some individual outcomes due to missing data as indicated.
1 No baseline control for outcome variable available.
2 Non-response for one or more items on the scale ranged from 24–30% across study arms and was higher for younger girls.
3 Violence prevention indicator reverse coded prior to inclusion in summary variable.
*** p<0.001, ** p<0.01, * p<0.05, † p<0.1