Investigating cognitive, non-cognitive and behavioral mediators for building literacy through edutainment in Kenya

How to cite

Ghosh, Anushka; Bottan, NIcolas; Baier, Jasmin. 2024. "Investigating cognitive, non-cognitive and behavioral mediators for building literacy through edutainment in Kenya." AEA RCT Registry. AEARCTR-0014858. https://doi.org/10.1257/rct.14858-2.0

Download: BibTeX · RIS — for reference managers without a browser translator. Zotero and most others pick up this page's citation metadata directly.

Anushka Ghosh* , NIcolas Bottan , Jasmin Baier

* Primary investigator

Registry ID
AEARCTR-0014858
DOI
https://doi.org/10.1257/rct.14858-2.0
First registered
Last updated
Status
Completed
Trial dates
2024-10-01 – 2024-11-30
Countries
Kenya (Nairobi)
Registry record
http://www.socialscienceregistry.org/trials/14858

Abstract

Educational television has shown promise for creating supplemental learning resources in the home and shaping attitudes in developing countries. We conducted a comprehensive evaluation of the effects of watching a new educational television program at home. Besides instructional content, a key innovation of the show is its objective to change children's mindsets about reading, gender attitudes and socio-emotional learning (SEL). Following a large-scale RCT, we have designed a lab study that investigates the learning mechanisms that are crucial in building literacy through educational television. In order to better understand if and how the show succeeds in building literacy skills, this study attempts to look closer at whether there is evidence that this program facilitates cross linguistic transfer of reading skills, and expands the RCTs exploration of potential mechanisms. It also evaluates the role of curiosity, distinguishing between trait curiosity and state curiosity, directed toward books, as well as the influence of executive functions such as inhibitory control, cognitive flexibility and working memory on literacy and SEL outcomes. Additionally, the study assesses the moderating effects of caregiver engagement on literacy and curiosity outcomes and explores whether knowledge retention from the show is essential for developing comprehension skills. By integrating qualitative and quantitative methods, the research seeks to identify active pathways linking the intervention to its observed effects.

Intervention

The Nuzo and Namia show, developed by Ubongo, targeted 6-9-year-old children with the aim of improving their literacy, social-emotional learning, and gender attitudes. The plot revolves around 2 protagonists, Nuzo and Namia, as they meet with different African characters at different moments throughout history. By interacting with the protagonists, who appear just in time to help them solve a problem that they are facing, viewers see and hear the protagonists model key strategies in early reading comprehension, such as predicting, questioning, clarifying, and analyzing information. For example, a character may predict a book's content based on the story's cover and title. The audience can then later see them refine their predictions halfway through the story or when they start reading. They may also compare texts they have read in different parts of the book and then self-question their comprehension.

Furthermore, seeing protagonists who are just like the audience (i.e., relatable role models) reading and using stories/books to practise their reading aims to give children agency, confidence, and motivation to improve and practise their own reading skills. In interactions with the historical figures in the show, along with their family and friends, viewers witness the protagonists identify and model positive social-emotional learning outcomes and gender attitudes. The protagonists are taken on adventures where they can try things that earlier may have seemed unsuitable or impossible beforehand. Through the open-mindedness and perseverance of the protagonists, viewers learn that by trying different things and sticking to them for a while, they can find activities that truly suit them and their skills, even if societal norms or early difficulties don’t encourage it. The series aims to both “show” these character traits (like confidence and curiosity) and explicitly state how these helped the protagonists succeed (e.g., a protagonist saying, “I am really glad I didn’t give up when XY happened!”). The programme consists of 13 22-minute episodes.

Experimental design

Overview:
We design a lab-in-the-field RCT in 3 communities with public schools in Nairobi. We have 3 groups in the study: placebo/ control and 2 treatment groups that receive variations of the intervention. Each group or treatment arm within the group will be seated in a room with a smart TV, where the show will be broadcast. We envision six-episode viewings over six weeks.

Treatment Groups:
Group 1: Placebo Control
This group will watch a placebo show content and not be exposed to the Nuzo and Namia treatment. At this moment, we propose showing the children a wildlife documentary to ensure no similarities in programming in literacy, SEL, or gender attitudes.
This group represents a control group against which we can compare results from the treatment groups to extract the impact on the sub-skills developed by the show.
Group 2: Receives the N&N show in Kiswahili without caregiver intervention
No iterations. This group watches the Kiswahili version of the Nuzo and Namia show.
This group helps us decode the cross-linguistic transfer component compared to the other treatment arms. It would be interesting to see if these children demonstrate skills transfer from Kiswahili to English or their mother tongue and whether we should expect that to happen to the original study sample.
Group 3: Receives N&N in Kiswahili) with salient caregiver engagement nudges
This group watches in Kiswahili and receives some variations in the treatment, with caregiver engagement as the primary component.
Narration, a recap at the end of the episode, and a short demonstration of building conversational literacy in the home while watching the show can drive caregiver engagement.
Any other mechanism-related design will be incorporated during the research conceptualization phase.
We can decode the aspect of cross-linguistic transfer from this group by using the relative literacy scores between languages but controlling for absolute level differences between groups. We can also compare the results of having received caregiver intervention (absolute level differences).

Design details

Randomization method
Randomized in office by a computer using R as the statistical software.
Randomization unit
Randomized at the household level
Clusters
150 households.
Observations
300 observations with 150 children and 150 caregivers.
Arms
50 households in control, 50 households in treatment arm 1 with exposure only to the Nuzo and Namia show, 50 households in treatment arm 2 with exposure to Nuzo and Namia and caregiver intervention.
Primary outcomes
We explore the following primary research questions: 1. Does the show facilitate cross linguistic transfer (CLT) of reading skills? Since the program exists in a language agnostic space, are we on track to build literacy in the long-term by anticipating a transfer of skills across languages? 2. How does caregiver engagement through targeted dual-generation programming modulate variations in programme impact? 3. How closely do curiosity and executive functioning skills relate to literacy development, following the Active View of Reading Model (Duke and Cartwright, 2021)? 1. Reading ability To assess one primary summary measure for latent reading and listening comprehension skills, which are the primary teaching objective of Nuzo & Namia, we perform Item Response Theory (IRT) analysis on all items included in: i. EGRA listening and reading comprehension sections ii. EGRA adapted subtasks on object word association and spatial recognition IRT models the relationship between individuals' latent trait (e.g., ability) and their performance on observed items. We compute this separately for each of the three languages tested in the survey: English, Kiswahili, and Luyo. 2. Confidence and Curiosity score (developed by RTI)- Combined score: 8 items coded from 0 to 2. We follow Anderson (2008) to create an index which has a mean of zero and standard deviation of one, and is standardized relative to the control group. Given that the combined score mixes two constructs, we will furthermore conduct a principal component analysis as a more data-driven approach and as a robustness check. Confidence score: 4 sub-items coded from 0 to 2. Similar to the combined score we will create an index following Anderson (2008). Curiosity score: 4 sub-items coded from 0 to 2. Similar to the combined score we will create an index following Anderson (2008). 3. Summary measure: Behavioral index As primary measure in this family, we will conduct principal component analysis (PCA) on the following likert-scale measures, after having standardized each item. IB1: How often do you read outside of school, i.e. for enjoyment? IB2: How often do you look for or ask for books for enjoyment and not school work ? IB3: How often does your mum, dad or carer read books or stories to you? 4. Information and Skill (Book descriptives and Executive Function) In this family, we construct 6 primary measures. Primary measures: I. Executive Function (comprising of): a. Inhibitory Control b. Working Memory c. Cognitive Flexibility & Creativity- Alternative uses task measure: Sum of eligible alternative uses the child could think of Functions of a book II. Book and Print Recognition measure: Sum of correctly identified book elements and functions 5. Social Interactions I. Social interaction index: We construct an index using principal component analysis (PCA) on the following items, measured on a 4-point likert scale. We standardize all items before conducting PCA: a. "How often do you meet with friends to read (for enjoyment, not school books)? b. "How often do you ask friends for help if you have a problem to solve (problem related to studies or school)? c. "How often do you ask your parents for help if you have a problem to solve (related to studies or school)? d. "How often do you talk to your friends about your feelings? e. "How often do you talk to your parents about your feelings? II. Parent conversational interaction: We construct an index using principal component analysis (PCA) on the following items a. Talk to you while doing an activity w/ you around b. Describe things to you when walking c. Pointed, named object and asked you to repeat d. Read/ looked at a book with you e. Told you a story III. General adult conversational interaction: We construct an index using principal component analysis (PCA) on the following items a. Sing to you b. Play with you c. Read/ looked at a book with you d. Told you a story e. Described things to you

Cited by

No citation data available for this registration yet.

Keywords