| Field | Before | After |
|---|---|---|
| Field Last Published | Before December 21, 2020 10:33 AM | After April 29, 2021 01:50 PM |
| Field Intervention (Public) | Before The intervention will be implemented at the village level for 6-months. The financial incentive scheme consists of a coupon, on which participants will receive a stamp for every Rs. 20 (Indian rupees) spent on fruit and vegetables from a participating local vendor. If participants spend Rs. 500 within one week, they will receive a Rs. 100 reimbursement from the APCAPS field team. | After The intervention will be implemented at the village level for 3-months. The financial incentive scheme consists of a coupon, on which participants will receive a stamp for every Rs. 10 (Indian rupees) spent on fruit and vegetables from a participating local vendor. If participants spend Rs. 250 within one week, they will receive a Rs. 50 reimbursement from the APCAPS field team. |
| Field Primary Outcomes (Explanation) | Before We shall conduct a quantitative survey of randomly selected households of the intervention and control villages at baseline, 2, 4, and 6-months of the total 6-month implementation period. The same household member will be surveyed at each time point, and we shall ask for information on the quantity of different fruit and vegetables purchased for all members of household over a one week period (we shall survey them on alternate days of the week, asking about the previous two days for maximum accuracy). For items that are frequently reported in measures other than weight (e.g. bunches or pieces) we have previously collected data on the weight of locally traded fruit and vegetable servings, which we will use to convert the amount of fruit and vegetables purchased to weight in kg. The amount (kg) of fruit and vegetables will be summed per household per week to generate the amount of fruit and vegetable purchased. | After We shall conduct a quantitative survey of randomly selected households of the intervention and control villages at baseline and after 3-months (i.e. at the end of the implementation period). The same household member will be surveyed at each time point, and we shall ask for information on the quantity of different fruit and vegetables purchased for all members of household over a one week period (we shall survey them on alternate days of the week, asking about the previous two days for maximum accuracy). For items that are frequently reported in measures other than weight (e.g. bunches or pieces) we have previously collected data on the weight of locally traded fruit and vegetable servings, which we will use to convert the amount of fruit and vegetables purchased to weight in kg. The amount (kg) of fruit and vegetables will be summed per household per week to generate the amount of fruit and vegetable purchased. |
| Field Randomization Unit | Before Villages and households within selected villages. | After Villages. |
| Field Power calculation: Minimum Detectable Effect Size for Main Outcomes | Before We calculated that a sample of 1,182 participants is required to detect a 10% difference between intervention and control arms in our primary outcome of weekly household fruit and vegetables purchase (in kg), with 80% power and at 5% significance level (assuming mean household weekly fruit and vegetable purchase of 10.4kg with standard deviation 4.5kg, estimated from previous data). This sample size estimate accounts for the fact that each participant will be followed up 3 times (within-person intra-cluster correlation coefficient of 0.4, estimated from previous data) and participants are clustered in villages (within-village intra-cluster correlation coefficient of 0.02, estimated from previous data). To allow for possible loss-to-follow-up during the study, we plan to recruit 1500 participants in total (250 from each village). | After We had originally calculated that a sample of 1,182 participants is required to detect a 10% difference between intervention and control arms in our primary outcome of weekly household fruit and vegetables purchase (in kg), with 80% power and at 5% significance level (assuming mean household weekly fruit and vegetable purchase of 10.4kg with standard deviation 4.5kg, estimated from previous data). This sample size estimate accounts for the fact that each participant will be followed up 3 times (within-person intra-cluster correlation coefficient of 0.4, estimated from previous data) and participants are clustered in villages (within-village intra-cluster correlation coefficient of 0.02, estimated from previous data). To allow for possible loss-to-follow-up during the study, we planned to recruit 1500 participants in total (250 from each village). After the study was delayed and shortened due to the COVID-19 lockdown in India, we revised the above sample size calculation as follows: with 1500 (1200 after allowing for loss to follow-up), we would have 80% power (at 5% significance level) to detect a ~15% difference in primary outcome of household fruit and vegetable purchase (assumptions as above, except without repeat measures per individual). |
| Field Intervention (Hidden) | Before We will implement the village level intervention in the 3 randomly selected intervention villages for 6-months, with another 3 villages acting as controls. The intervention consists of a coupon promising a value of 100 Indian rupees to be redeemed after a purchase of Rs. 500 worth of fruit and vegetables within 1-week. At the start of the 6-months, the study team will distribute the coupon booklets (valid for the total 6-months) to every household in the intervention villages. We will provide an instruction pamphlet which will consist of details of the programme, coupon distribution and particulars on coupon redemption. Each vendor from the intervention villages will be provided with a stamp with a unique serial number. The coupon consists of 25 circles which will be stamped by participating vendors, with each circle amounting to Rs. 20. If a participant purchases multiples of Rs. 20 from the same vendor, they will receive that many stamps, each worth Rs. 20 (a purchase of Rs. 20 or higher but less than Rs. 40 will equate to one stamp). The APCAPS field staff will visit the intervention villages once a week and reimburse participants Rs. 100 once they provide their completed coupon. The study team will be accompanied by promotional material (for instance jingles, posters, videos) which will advertise the coupon scheme. | After We will implement the village level intervention in the 3 randomly selected intervention villages for 3-months, with another 3 villages acting as controls. The intervention consists of a coupon promising a value of 50 Indian rupees to be redeemed after a purchase of Rs. 250 worth of fruit and vegetables within 1-week. At the start of the 3-months, the study team will distribute the coupon booklets (valid for the total 3-months) to every household in the intervention villages. We will provide an instruction pamphlet which will consist of details of the programme, coupon distribution and particulars on coupon redemption. Each vendor from the intervention villages will be provided with a stamp with a unique serial number. The coupon consists of 25 circles which will be stamped by participating vendors, with each circle amounting to Rs. 10. If a participant purchases multiples of Rs. 10 from the same vendor, they will receive that many stamps, each worth Rs. 10 (a purchase of Rs. 10 or higher but less than Rs. 20 will equate to one stamp). The APCAPS field staff will visit the intervention villages once a week and reimburse participants Rs. 50 once they provide their completed coupon. The study team will be accompanied by promotional material (for instance jingles, posters, videos) which will advertise the coupon scheme. |
| Field Public analysis plan | Before No | After Yes |
| Field Secondary Outcomes (End Points) | Before (1) Amount spent (Indian rupees) on fruit and vegetables per week per household (2) Process evaluation of intervention (following the Medical Research Council Framework) a. Intervention delivery: Fidelity e.g. % of vendors not stamping on a given day/not wearing a badge; dose e.g. % of total stamped circles out of 600 for each participant; reach e.g. % of customers who have used the coupon for >12 weeks with at least 12 stamps b. Causal mechanisms e.g. % of customers who stated that they could not afford to purchase Rs. 500 of fruit and vegetables, % of customers who stated that they were doubtful of being reimbursed Rs. 100 by the programme c. Contextual factors: e.g. variety and quality of fruit and vegetables available | After (1) Amount spent (Indian rupees) on fruit and vegetables per week per household (2) Quantity of fruit and vegetables obtained (in kg) from all sources by households in a week (i.e. sum of purchase, own production, gifts, wild harvest and others). (3) For the quantity (in kg) and value (in rupees) of fruit and vegetables sold by vendors in a week (from vendor survey)/ (4) Process evaluation of intervention (following the Medical Research Council Framework) a. Intervention delivery: Fidelity e.g. % of vendors not stamping on a given day/not wearing a badge; dose e.g. % of total stamped circles out of 600 for each participant; reach e.g. % of customers who have used the coupon for >12 weeks with at least 12 stamps b. Causal mechanisms e.g. % of customers who stated that they could not afford to purchase Rs. 500 of fruit and vegetables, % of customers who stated that they were doubtful of being reimbursed Rs. 100 by the programme c. Contextual factors: e.g. variety and quality of fruit and vegetables available |
| Field Secondary Outcomes (Explanation) | Before The amount spent (in Indian rupees) on fruit and vegetables in the previous week will also be collected during the quantitative survey of the randomly selected households (alongside the primary outcome). A process evaluation will combine routinely collected data (the redeemed coupons), a monthly quantitative vendor survey (to estimate the volume and cost of purchase and sales), a weekly vendor audit survey (conducted through observations by the field team), an intervention evaluation survey with participating households and vendors (conducted in the final month of the intervention), and qualitative interviews with local stakeholders (e.g. government personnel, Anganwadi workers). The process evaluation will evaluate how the intervention was implemented, the potential causal mechanisms (for instance perception and understanding of the intervention), unexpected outcomes, and the contextual factors that may have affected intervention implementation and consequences. | After The amount spent (in Indian rupees) on fruit and vegetables and quantity of fruit and vegetables obtained from other sources in the previous week will also be collected during the quantitative survey of the randomly selected households (alongside the primary outcome). Vendors will be surveyed once per month during the follow-up period, for quantity and value of fruit and vegetables sold. A process evaluation will combine routinely collected data (the redeemed coupons), a monthly quantitative vendor survey (to estimate the volume and cost of purchase and sales), a weekly vendor audit survey (conducted through observations by the field team), an intervention evaluation survey with participating households and vendors (conducted in the final month of the intervention), and qualitative interviews with local stakeholders (e.g. government personnel, Anganwadi workers). The process evaluation will evaluate how the intervention was implemented, the potential causal mechanisms (for instance perception and understanding of the intervention), unexpected outcomes, and the contextual factors that may have affected intervention implementation and consequences. |
| Field | Before | After |
|---|---|---|
| Field Document | Before |
After
Statistical+Analysis+Plan_FV_29042021.pdf
MD5:
13863641b677da8ceea0b9eef3761d50
SHA1:
ce04ee01c793bf8f5a39818130bb006c76aacb84
|
| Field Title | Before | After Statistical Analysis Plan (V-29042021) |