Trends in home schooling strategies in Nigeria
As schools close around this world, different households are leaning towards different coping strategies for home schooling.This article walks through using a heatmap to visualise the most popular strategies used across Nigeria.
In March of 2020, Nigeria declared a state of emergency and went into national lockdown for 4 weeks. The Nigerian Bureau of Statistics in collaboration with the World Bank conducted phone surveys starting April 2020 to get a sense of the impact of COVID-19 on households in Nigeria. This article walks through a series of heatmaps to get an overview of typical home schooling strategies that households from different parts of Nigeria are utilising.
The data used for the visualisations were collected during the first 5 rounds of phone surveys conducted in 2020. The households interviewed were from those in the living standards measurement survey. The table below summarises the periods when the data for each round of survey was collected.
Start End Cycle 20-04-2020 11-05-2020 Round 1 (Baseline) 02-06-2020 16-06-2020 Round 2 06-07-2020 20-07-2020 Round 3 09-08-2020 24-08-2020 Round 4 07-09-2020 21-09-2020 Round 5
Except for in round 1 (the baseline round), the households that participated in the survey were roughly evenly distributed across geopolitical zones and approximately 60% of the households interviewed were from rural areas.
Amongst several other key indicators, the interviews explored the impact of COVID-19 on education. Specifically, it explored the strategies adopted by different household to cope with the school closures. These strategies included: the use of assignments, the use of educational mobile applications, the use of educational television, the use of educational radio, studying, tutoring by parents or other household members and tutoring by a teacher.
Each household could respond affirmatively to using a combination of these strategies. The heatmaps presented below serve at least two purposes:
Are there home schooling strategies that are particularly popular across all states?
Are there any clear changes in these strategies over time , i.e., across the different rounds?
Interpreting the heatmap
Each cell in the heatmap is the number of households in the given state that reported using that strategy as a coping strategy for managing education while schools were closed. The heatmap is rendered for each of Round 1 through 5 and the for each state.
The benefit of a heatmap is the quick perception of hotspots in the data. The value encoded in the coloring (i.e., the heat) in this case is the number of households in a given state who responded affirmatively to using the corresponding strategy. Hence we can see at a glance, the relative difference between the adoption of different education strategies across states and over time (i.e., the rounds). Each column represents the households that reported using the strategy for homeschooling. Each cell in the column represents a state and the color of the cell represent the proportion of households who use that strategy and are from that state. A red cell would simply mean more people reported using the strategy than a yellow cell. Vertical streaks show the strategies that are popular across states. The chart shows that nationally, and perhaps across all rounds, tutoring by parents/other household members, as well as studying were popular strategies. This visualisation merges responses from both rural and urban sectors. The cluster of strategies we see are by chance and we are still not clear on how the popularity of the strategies change over the rounds.
The charts below show how the popularity of the different strategy change over the rounds. Both axes are now sorted. On the vertical axis, the state with the highest response rate is at the top. On the horizontal axis, the strategy with the highest response rate is on the left. This is why we see the darker red cells on the top left corner. While studying and parent tutoring are consistently popular across rounds, engagement with tutors were not as popular earlier in the year. Now let's look at the data split by sector.
Effect of sector on home schooling strategies
The heatmap below shows households from urban areas. The reader might notice more holes in the visualisation. These holes represent states who did not report using the corresponding strategy. Where the state is missing entirely, then there were no households from that state who reported continued engagement with education for that round.
If we plotted the same for households in the rural parts of the country, we see states with sparse representations in the rural sector. For this visualisation these were states such as Borno, FCT, Lagos or Ekiti.
We could also just focus on a single strategy to see how different states compare over different rounds for that strategy. That would drive discussions such as "Which states are more likely to engage in learning by mobile apps?" The image below shows that Lagos is where people are most likely to report having access to mobile apps for education. Curiously, Benue, Kwara and Enugu states started off stronger with households reporting use of mobile apps for education purposes. At later rounds, we see households reporting less use of this strategy. Adamawa on the other hand has seen a steady increase in households reported to have used this strategy to cope with the closure of schools.
We could take the visualisation a step further and render the heatmap just for a single state over the 5 round of interviews. This would allow us to see how different strategies compare across different rounds for a single state. The possibilities are endless.
These visualisations were generated using vega-lite in observablehq. There is an accompanying interactive notebook with the dataset used to generate the pictures and all the code. Do check it out.
Acknowledgements
This article was written based on data collected by National Bureau of Statistics. Nigeria COVID-19 National Longitudinal Phone Survey (COVID-19 NLPS) 2020. Dataset downloaded from https://microdata.worldbank.org/index.php/catalog/3712/get-microdata on 31 December 2020.
Jeremiah Iyamabo provided valuable comments on draft versions of this document.