Skip to main content

Table 7 Linear regression results for impact of mHealth treatment on sexual and reproductive health knowledge z-score, stratified by target subgroups, at 3-month and 15-month follow-ups

From: Assessing the reach and effectiveness of mHealth: evidence from a reproductive health program for adolescent girls in Ghana

 

Parental education

Baseline knowledge

Sexually active at baseline

Parental Support

 

Both parents low

One parent high

Both parents high

Missing/Other

No knowledge deficit

Knowledge deficit

Not sexually active

Sexually active

Refused

Low

High

Panel A: 3-month follow-up

Treatment

2.148***

1.679***

1.481***

1.982***

1.388***

1.700***

1.722***

1.836***

1.661***

1.748***

1.477***

 

(0.301)

(0.271)

(0.262)

(0.235)

(0.132)

(0.257)

(0.215)

(0.332)

(0.364)

(0.186)

(0.310)

Home econ. Class

0.446

0.461**

0.482**

0.903**

0.590*

0.634**

0.772***

−0.353

0.661

0.718***

0.440

 

(0.747)

(0.217)

(0.219)

(0.390)

(0.301)

(0.226)

(0.182)

(0.586)

(0.412)

(0.174)

(0.285)

High Quality public school

0.989

−0.286

0.384

−0.337

0.285

−0.0810

−0.0568

0.563

0.0339

0.165

−0.0607

 

(0.682)

(0.331)

(0.309)

(0.280)

(0.238)

(0.276)

(0.228)

(0.600)

(0.279)

(0.208)

(0.352)

Low Quality public school

0.876

0.00474

0.250

−0.854***

0.223

−0.217

−0.0882

0.448

−0.166

−0.000159

0.0725

 

(0.616)

(0.208)

(0.329)

(0.224)

(0.214)

(0.210)

(0.171)

(0.676)

(0.265)

(0.195)

(0.268)

Constant

−1.055**

0.0161

−0.394*

−0.281

0.0442

−0.664***

−0.459**

0.468

−0.348

−0.538***

−0.0570

 

(0.441)

(0.141)

(0.199)

(0.495)

(0.255)

(0.168)

(0.198)

(0.533)

(0.408)

(0.154)

(0.306)

Observations

37

108

236

97

242

236

336

62

61

324

153

R-squared

0.647

0.359

0.307

0.498

0.339

0.411

0.401

0.415

0.369

0.416

0.263

p-valueinteraction

0.12

0.10

0.78

0.15

Panel B: 15-month follow-up

Treatment

0.881***

0.812***

0.832***

1.181***

0.923***

0.598***

0.829***

0.980**

0.929**

0.887***

0.895***

 

(0.305)

(0.191)

(0.246)

(0.173)

(0.116)

(0.159)

(0.127)

(0.410)

(0.433)

(0.157)

(0.171)

Home econ. Class

0.616

−0.0862

−0.0355

0.277

0.106

−0.0325

0.0843

−1.189*

0.810*

0.0363

0.159

 

(0.784)

(0.286)

(0.252)

(0.306)

(0.263)

(0.245)

(0.235)

(0.602)

(0.447)

(0.219)

(0.268)

High Quality public school

1.475***

0.246

−0.0608

0.0939

0.284

0.139

0.274

0.673

−0.235

0.261

−0.0683

 

(0.190)

(0.461)

(0.341)

(0.214)

(0.235)

(0.277)

(0.253)

(0.577)

(0.392)

(0.236)

(0.300)

Low Quality public school

0.950***

0.449

0.139

−0.329

0.369

0.0488

0.351

0.704

−0.355

0.327

0.0116

 

(0.240)

(0.487)

(0.352)

(0.217)

(0.231)

(0.266)

(0.249)

(0.651)

(0.332)

(0.238)

(0.297)

Constant

−0.651

0.676***

0.753**

0.442

0.669***

0.503

0.382

1.646***

0.0955

0.446*

0.678***

 

(0.708)

(0.217)

(0.281)

(0.396)

(0.225)

(0.302)

(0.282)

(0.489)

(0.409)

(0.256)

(0.231)

Observations

36

113

225

100

241

233

345

65

63

326

147

R-squared

0.434

0.138

0.111

0.277

0.210

0.077

0.146

0.282

0.193

0.149

0.144

p-valueinteraction

0.51

0.11

0.37

0.99

  1. Notes: Each column is separate regression model, stratified by variable in column header. Clustered standard errors at school level in parentheses. *** p < 0.01, ** p < 0.05, * p < 0.1. The p-valueinteraction is from linear hypothesis test of equality of coefficients for treatment variable, calculated after model interacting treatment with indicator for subgroup. Home econ. = Home economics class in school. Reference category for category of schools is Private school