Study: the relationship between DoodleMaths usage and academic progression




A summary of a study conducted by Bill Smith, a PhD student, investigating the link between DoodleMaths usage and academic progression.

 

Background


Baseline assessments (BLAs) are used to measure a child’s level of ability. They’re completed when a child first starts using Doodle and every 3-6 months after this.


The results of the BLA are used by Proxima™, Doodle’s algorithm, to personalise each child’s work programme, ensuring they’re learning at the right level and pace. They also provide a numerical measure of each child’s ability, making them a helpful tool for analysis.


By using the BLA results of our users, I wanted to see if using DoodleMaths regularly over a period of time leads to an increase in DoodleAge (a measure of a child’s level of understanding compared to learners of a similar age).


My findings showed that there is a positive correlation between effort (using DoodleMaths regularly) and progression (DoodleAge), with those who regularly use DoodleMaths making more progress in maths than those who don’t.


Key findings:

  • Earning 7 stars a day resulted in an average increase of 1.3 DoodleAge a year

  • Earning 24 stars a day resulted in an average increase of 2.6 DoodleAge a year

  • The more stars a child earns in DoodleMaths a day, the faster they progress

 

Methodology


As children naturally progress in their maths due to their education at school or home, a natural level of progress needed to be found in order to study the effects of using Doodle.


To do this, the results of the first BLAs taken by 60,282 children when they started to use DoodleMaths were sorted into year groups.


This revealed that each year group had an average DoodleAge (DA) of 0.84 higher than the year group below it. This number was used as the natural level of progression made by children not using any Doodle products.


In order to study the effect that using DoodleMaths has on progression, all children who had completed two or more BLAs were then identified.


When a child completes a BLA, their results are converted into a DA. Each possible pair of their BLA results were then listed (e.g. a child who had completed 3 BLAs had three pairs of results: DA from BLA 1, DA from BLA 2; DA from BLA 1, DA from BLA 3; DA from BLA 2, DA from BLA 3), giving 24,468 pairs in total.


By subtracting the first DA from the second DA, we were then able to see how much each child had progressed in between taking their BLAs.


And by using this information, we were able to calculate the rate of progress across all year groups for children regularly using DoodleMaths.

 

Results


The rate of progression across all year groups is shown below (the different year groups are in different colours). It shows the results of each year's group average first BLA result (as a DA).


The findings show a clear gradient and a consistent gap between most year groups.

Further insight can be gained from comparing this gap to childrens’ rate of progression when using DoodleMaths.


Each dot on the graph below represents a group of children who earn a similar number of stars per week and their typical rate of progression (increase in DA) per year:



This graph shows a positive correlation between using DoodleMaths regularly and an increasing DA.


When earning between 6 and 25 stars a day, there is a linear trend with a gradient of 0.08. This means that a group of children who earn just one more star a day than another group will achieve an average of 0.08 DA a year more.


Other key figures are outlined below:

Stars / day

​Progress, DA / year

5

1.17