Data for: Mobile Interventions for Upscaling Participatory Action and Videos for Agriculture and Nutrition (m-UPAVAN) – User Guide

Permanent identifier

10.17037/DATA.00003635

Description

These data were collected to assess the feasibility, acceptability and equity of an intervention from the ‘UPAVAN’ trial (see: 10.1016/S2542-5196(21)00001-2) adapted to a mobile intervention (m-UPAVAN) during the COVID-19 pandemic in rural Odisha, India using a convergent parallel mixed-methods design. This includes a dataset and Corresponding Stata DO file used for m-UPAVAN quantitative analysis. The dataset covers monthly monitoring phone surveys assessing coverage, uptake, and acceptability of the m-UPAVAN intervention among women with a child aged <2 years, merged with sociodemographic data collected during a phone survey before the intervention began.

Data collection methods

Quantitative data:

Identifying participants: In Dec 2020-Jan 2021, m-UPAVAN intervention facilitators visited all households in the 133 villages in Keonjhar district, Odisha India where the m-UPAVAN interventions were to take place and collected information on phone ownership and numbers, and identified eligible participants for quantitative feasibility and acceptability assessment (women with a child <2 years of age and with intra-household phone access (i.e., either owned a phone or had access to a phone in their household).

Baseline sociodemographic phone survey: From the list of eligible participants, women and their spouses were randomly selected and invited to participate in phone-based sociodemographic baseline surveys in Jan-Feb 2021. The target sample size was 1,000 mothers and their spouses.  

Monthly monitoring phone surveys: Each month for five months (May-Sept 2021) mothers were randomly drawn from those who participated in the baseline phone survey and invited to participate in phone-based monitoring surveys designed to assess intervention feasibility and acceptability. The target sample size each month was 200 mothers. The recall period for the monitoring surveys is the past month/4 weeks.

All phone surveys were conducted by an independent data collection firm, DCOR Consulting Pvt. Ltd.

Data were captured in SurveyCTO.  

Qualitative data:

At the end of the intervention, qualitative in-depth interviews were conducted to understand women’s experiences of the intervention, factors that affected it’s feasibility and acceptability, and preferences for mobile (m-UPAVAN) vs face-to-face (UPAVAN) nutrition-sensitive agriculture interventions. 38 women with a child aged <2 years were recruited using purposive sampling to reflect the caste diversity of the population and varied phone types and access (33 mothers with intra-household phone access from the quantitative data collection strand, and 5 mothers without intrahousehold phone access). A

All in-depth interviews were semi-structured, with open-ended questions and relevant prompts. Interviews were conducted in Odiya at participants home.

Data collectors sought informed consent from participants orally which was recorded and documented by DCOR. For in-person qualitative interviews, data collectors sought written (signature or thumbprint) informed consent from participants.

Data analysis and preparation

Quantitative data:

Each participant is assigned a unique identifier indicated in the variable ‘hhid’. Data collected during monthly monitoring surveys (five months of monitoring surveys) were appended (long format) and merged with baseline sociodemographic data using hhid.

Appropriate data cleaning was indicators used for feasibility and acceptability assessment were generated. Variables and their values are labelled. Participants who were randomly drawn and contacted to participate at least once in monitoring surveys but never responded/answered the call are labelled accordingly (variable: ‘non-response’), and those who participated at least once in the monitoring surveys are indicated by the variable ‘monthly_response’.

The Stata do file used for the quantitative analysis of m-UPAVAN is included.

Qualitative data: Interview recordings were translated and transcribed verbatim into English language by DCOR before analysis.

Geographic regions

99 (133 villages) of the 148 UPAVAN clusters (village and surrounding hamlets) in all four blocks of the UPAVAN trial (Ghatgaon, Harichandanpur, Patna, and Keonjhar Sadar) in Keonjhar district, Odisha, India.

Quality controls

Quantitative data: Phone calls to collect quantitative data were recorded, with permission for this gained at the consent stage. Supervisors checked 10% of the calls to ensure the electronic data capture is accurate. Furthermore, supervisors joined up to 20% of live phone calls to monitor the conduct and content of the interview. Electronic data capture method ‘SurveyCTO’ included relevant constraints on responses to avoid missing data and outliers/implausible values.

Qualitative interviews were coded by three-research and a forth researcher double-coded on in three transcripts, and reviewed inconsistencies any inconsistencies. If they arose, this was discussed with the whole team and resolved.

Population

Women with a child <2 years of age and their spouses, in Keonjhar district, Odisha, India.

Privacy

All participant identifiable information has been removed from all qualitative and quantitative data.

Ethics

Keywords

Nutrition, nutrition-sensitive agriculture, India, feasibility, intervention, mHealth, behaviour change communication, equity, acceptability, coverage

Language of written material

English language

Project information

Creators

Associated roles

File description