Fig 1.
Implementation of participatory women’s groups.
Vertical lines indicate exposure to an intervention. The total number of women interviewed refers to the number of women for which their household agency was measured at endline.
Fig 2.
Trial flow for the data collection process.
Table 1.
Baseline characteristics disaggregated by availability for interview.
All values except deliveries in Periods B-E (see Fig 1) refer to measurements at baseline.
Table 2.
Descriptive statistics collected from respondents upon follow-up.
n = 4030. †6 respondents had missing values on health-seeking behaviour. ‡19 respondents had missing values on group participation. • Values reported in the format mean (SD). Scores for agency freedom are raw, unstandardized scores.
Table 3.
Impact estimates of the effect of exposure to the original Makwanpur intervention on overall agency measured with the RAI.
Effect sizes in standard deviations. 95% confidence intervals in brackets.
Table 4.
Impact estimates of the effect of exposure to the original Makwanpur intervention on sub-domains of the overall agency score.
Effect sizes in standard deviations. 95% confidence intervals in brackets.
Fig 3.
Estimates from models on the effect of the original Makwanpur intervention and subsequent exposure after varying θ in IMPEXP(θ).
Effect sizes are in units of standard deviations of the overall agency measured with the RAI. Bands around point estimates of effect sizes represent 95% confidence intervals.θ refers to the weight given to exposure to PLA groups while non-pregnant. When θ we control exclusively for later exposure while pregnant. When θ = 1, we weight exposure to PLA groups while pregnant and non-pregnant equally.
Table 5.
Impact estimates of the effect of exposure to the original Makwanpur intervention on alternative measures of agency.
95% confidence intervals in brackets. Effect sizes in logits for financial power. Effect sizes in steps on the Power Ladder (out of 10) for the Power Ladder question.