Testing the effectiveness of edutainment videos on improving knowledge, attitudes and outcomes towards Minimum Diet Diversity (MDD) for young children (6-23 months)

Tales of Mazrupur: A series of 5 episodic-style narrative animated videos. The duration of each video is 3.5 to 4 minutes on average. They address the primary outcomes of:
1. The importance of MDD for infants (to build immunity and strength etc.)
2. Food groups/items young children in the age group of 6-24 months should be fed

Khaan Paan Gaan: A series of 4 animated videos of bollywood song adaptations. The duration of each video is 1 to 1.5 minutes on average. They address the primary outcomes of:
1. MDD for young children over 6 months of age
2. Food groups/items young children in the age group of 6-24 months should be fed

Control 1: Two videos developed for Poshan Abhiyan (Module 6 and 9), Government of India were edited to match the length of the Tales of Mazrupur intervention videos. The videos are animated and Informational/Instructional modules in the form of a skit. They address the primary outcomes of:
1. MDD for young children over 6 months of age
2. Importance of CF for young children (for mental and physical development)
As well as other secondary information about MDD commonly covered by government collateral such as food preparation etc.

Control 2: Four separate short unedited musical videos from different sources developed by/distributed by the Ministry of Women and Child Development, Government of India on MDD for infants (compiled to match duration of Khaan Paan Gaan intervention videos). These include three no- animated and one animated informational/entertaining videos tailored to the general audience. They address the primary outcomes of:
1. MDD for young children over 6 months of age
2. Importance of CF for young children
As well as other secondary information about child and maternal nutrition commonly covered by government collateral such as EBF, colostrum, maternal diet diversity etc.

2023-03-13

2023-03-28

MDD Knowledge
Knowledge about consequences of not following MDD
Attitudes about what is healthy feeding
Intention: Giving food made for everyone at home
Intention: Diverse types of food
Intention: Buy diverse fruits and vegetables

1. MDD Knowledge: This variable is constructed using various separate variables regarding respondent knowledge regarding child feeding practices with regard to separate food groups and initiation age for young children ('Knowledge: MDD initiation', 'Knowledge: Grains', 'Knowledge: Seasonal Vegetables', 'Knowledge: Meat', 'Knowledge: Leafy Vegetables', 'Knowledge: Condiments'). Each variable has a binary question. The outcome variable is constructed using responses to each of the separate knowledge questions, where for each correct answer the score increases by 1.

2. Knowledge about consequences of not following MDD: This variable is constructed using various separate variables regarding respondent knowledge regarding the consequences of not adhering to MDD for young children ('Knowledge: No MDD affects growth', 'Knowledge: No MDD affects learning', 'Knowledge: No MDD affects strength', 'Knowledge: Food may cause choking', 'Food may cause: Diarrhoea'). These variables are measured using 5-point likert scale questions in the survey. For their response to each question, respondents are assigned a score between 0 and 1. The constructed variable uses the cumulative score from their response to both questions.

Attitudes about what is healthy feeding: This variable is constructed using various separate variables regarding respondent attitudes regarding healthy feeding practices for young children ( 'Attitude: Variety', 'Attitude: Giving healthy food made for adults'). The two variables related to attitude are measures using 5-point likert scale questions in the survey. For their response to each question, respondents are assigned a score between 0 and 1 (MDD positive attitudes are assigned a score greater than 0.5, MDD negative attitudes are assigned a score lesser than 0.5). The constructed variable uses the cumulative score from their response to both questions.

Intention: Giving food made for everyone at home: This variable is measured using a binary question in the survey. MDD positive intention is assigned a score of 1, MDD negative intention is assigned a score of 0.

Intention: Diverse types of food: This variable is measured using a binary question in the survey. MDD positive intention is assigned a score of 1, MDD negative intention is assigned a score of 0.

Intention: Buy diverse fruits and vegetables: This variable is measured using a binary question in the survey. MDD positive intention is assigned a score of 1, MDD negative intention is assigned a score of 0.

Narrative technique
Relatability

Narrative Technique: This variable is constructed using the 'Novelty', 'Engagement' and 'Transportation' outcome variables measured in the endline survey through 5-point Likert scale questions. Respondents will be assigned a score between 1 and 0 for their responses to each of the three questions (Effective Narrative Techniques are assigned a score over greater than 0.5, Non-effective Narrative Techniques are assigned a score lesser than 0.5) .The cumulative score will be their 'Narrative Technique' score.

Relatability: This variable is measured using a 5-point likert scale question in the survey. Each response is assigned a score between 1 and 0. High relatability is assigned a score greater than 0.5, Low relatability is assigned a score lesser than 0.5.

This evaluation will have the following components:

Baseline: This will be self-administered by respondents using tablets. For certain respondents (grandmothers and other respondents that are not able to self-administer the survey, the survey will be administered by enumerators. Data regarding whether the survey was self-administered or CAPI will be recorded)

Treatment A (Tales of Mazrupur): Watching a series of 5 episodic-style narrative animated videos (approximate duration: 15 mins) regarding minimum diet diversity (MDD) for children between 6 to 23 months of age.

Treatment B (Khaan Paan Gaan): Watching a series of 4 standalone bollywood song adaptation animated videos (approximate duration: 5 mins) regarding minimum diet diversity (MDD) for children between 6 to 23 months of age.

Control A: Watching two animated videos developed for Poshan Abhiyan (edited to match treatment length) (approximate duration: 15 minutes) regarding minimum diet diversity (MDD) for children between 6 to 23 months of age.

Control B: Watching four musical videos (three non-animated and one animated) developed by/distributed by the Ministry of Women and Child Development, (approximate duration: 5 minutes) regarding minimum diet diversity (MDD) for children between 6 to 23 months of age.

Endline: This will be conducted in the same format as the baseline.

The administration of baseline, treatment/control and endline was completed in one study session for each respondent.

The respondents were randomised across treatment/control arms through a random number generator within the programmed survey form (SurveyCTO).

Individual

No

0 clusters

The minimum sample size (across all arms) for the study is 484 respondents (i.e including mothers, fathers and grandmothers)

The minimum required sample size is 121 respondents per arm.

Power calculations were carried out to determine the minimum sample size required per arm to detect a medium effect size (Cohen’s H = 0.5). Data from two unpublished studies conducted by CSBC in Madhya Pradesh, India in 2021 were used to carry out power calculations. Power calculations were carried out separately for two outcome variables of interest 1) MDD knowledge and 2) MDD intention. As per the calculations, the percentage point difference (outcome proportion for the treatment group - outcome proportion for the control group) required is 0.177 for a medium effect size (h = 0.5) for MDD knowledge . For MDD intention, the percentage point difference required for a medium effect size is 0.179. The SD for MDD knowledge was 0.03 while the SD for MDD intention was 0.032. Power calculations were done for one treatment vs one control. We calculated Cohen’s h2 (for one-sample) and converted it to Cohen’s h for comparison. Cohen's h2 was multiplied by the square root of 2 to find Cohen's h. We set h2= 0.354 for medium effect size (h=0.5)and p1=0.33 for MDD knowledge, and h2= 0.0.368 for medium effect size (h=0.5) and p1=0.55 for MDD intention respectively, and solved for p2 in the following equation: h2 = [2*arcsine(√p2)] - [2*arcsine (√p1)]