Table 4 Intermediate outcomes

From: Behavioral savings sessions increase the pursuit of solar products among refugees in Uganda

 

(1)

(2)

(3)

(4)

(5)

 

Knowledge: % correct

Trust in solar providers: index

VSLA Support: index

Aspire to purchase solar this year

Self-efficacy: index

Treatment VSLA

0.27***

0.20***

1.03***

0.08**

−0.08

 

(0.02)

(0.06)

(0.09)

(0.04)

(0.06)

Replacement

−0.08***

0.02

−0.16

0.01

0.07

 

(0.02)

(0.07)

(0.11)

(0.04)

(0.08)

Constant

0.15***

3.79***

2.74***

0.76***

2.51***

 

(0.04)

(0.12)

(0.19)

(0.07)

(0.12)

VSLA controls

Yes

Yes

Yes

Yes

Yes

Individual control

Yes

Yes

Yes

Yes

Yes

SEs clustered VSLA

Yes

Yes

Yes

Yes

Yes

R-squared

0.27

0.04

0.25

0.03

0.07

Observations

1186

1137

1164

1186

1184

  1. The table presents the results of OLS regressions to estimate the impact of treatment assignment on the five intermediate outcome variables used in the mediation analysis. All the intermediate outcome variables are indices (see Table 1) except for the aspirations variable, which is dichotomous. Results are ITT estimates due to imperfect intervention compliance. All regressions control for VSLA-level controls (a settlement dummy, a dummy for VSLAs that meet at least weekly, the number of members in the VSLA, and the VSLA’s share price (normalized to weekly)) and individual-level controls (dummy for at least primary education). Robust standard errors clustered at the VSLA level are displayed below coefficients in parentheses. Asterisks denote a statistically significant difference at the 1% ***, 5% **, or 10% * levels.
Back to article page