This document is an archive for the final data sets analyzed in:Facilitating behavioral change: A comparative assessment of ASHA efficacy in rural Bihar
by
Oskar Burger, Faiz Hashmi, Maciej J. Dańko, Santosh Akhauri, Indrajit Chaudhuri, Emily Little, Hannah G. Lunkenheimer, Sudipta Mondal, Nachiket Mor, Neela Saldanha, Janine Schooley, Palash Singh, Tracy Johnson, Cristine H. Legare
published in PLOS Global Public Health, 2022
In what follows we provide downloadable datasets in .csv format for datasets used in the figures and tables presented in the published manuscript. Each table has buttons labeled ‘Copy’ and ‘CSV’. Pressing Copy will copy the data set to the user’s clipboard. Pressing CSV will download a copy of the dataset in .csv format.
For questions regarding this data archive contact: oburger@omni.org
Table 2.1: Two aggregate scores are computed and analyzed in the main text, the maternal health score and ASHA interaction score
The definitions of healthscore and ashaintscore are defined in the methods section of the paper.
Table 2.1: The survey data led to a tallying of counts for number of times each influencer was mentioned by mothers. Here we present the percent of times each influencer was mentioned.
| Column | Definition |
|---|---|
| Qvalue | 0 - mother did non-recommended behavior; 1 - mother did recommended behavior |
| influencer | Which influencer category was named in association with the decision to do or not do the recommended behavior. See main text for description of influencer categories. |
| value | count of mentions (number of times an influencer was associated with behavior (Qnames) and decision (Qvalue) |
| Qnames | Each behavior analyzed in the paper. See main text for definitions. |
| type | Is doing the behavior recommended or not recommended. |
| percent | The percent of total mentions for this influencer by behavior association; the column ‘value’ converted to a percent. |
Table 2.2: The survey data with covariates as used in logistic regressions, which are the main analyses of the paper.
| Column | Definition |
|---|---|
| Qvalue | 0 - mother did non-recommended behavior; 1 - mother did recommended behavior |
| Age | Age, categorical |
| Age_married | Age at marriage, categorical |
| Parity | Number of children at time of interview (total for that participant) |
| Edu | Years in education, self-report, categorical |
| Weatlh | Wealth, see main text for details of calculation. |
| Q{behaviorname} | Series of columns starting with Q; dummy variables for each behavior in the analysis |
| Family:no_one | Series of columns for each influencer in the analysis; dummy variables where 1 indicates that influencer was mentioned. |
Table 2.3: Computed model-based probabilities based on the model
This is the model-output described in the main text. The results of making predicted probabilities based on the model selection procedure.
| Column | Definition |
|---|---|
| Q | Each behavior, see main text for details |
| I | The influencer category, see main text for details |
| fit | the model-predicted probability that a woman does Q in the recommended direction given influence from I and the control variables. |
| se | standard error of the estimate of fit |
| N | number of times that I was associated with Q |
Table 2.4: Moderation analysis for relationship between asha interaction score and maternal health score
| Column | Definition |
|---|---|
| healthscore | Maternal health score as defined in main text |
| Parity | Number of births as self reported by participant at time of survey |
| EDU | Years of education, categorical, self report |
| ashaintscore | ASHA interaction score as reported in main text of paper |
| ASHA_int_c | ASHA interaction score, mean centered |
| wealth_diff | ASHA’s wealth score - Mother’s wealth score. Procedure for calculating wealth score described in main text. |
| wealth_diff_c | Wealth difference, mean centered |
| caste_diff | 1 = ASHA and Mother are same caste, 0 = ASHA and Mother are different caste |
| relig_diff | 1 = ASHA and Mother are same religion, 0 = ASHA and Mother are different religion |