As with the Multiplication Screening Check and KS2 Maths, here at Insight Inform we are often asked whether there is a connection between the raw scores recorded in the Phonics Screening Check (PSC) and scaled scores recorded five years later in the key stage 2 Reading SATs. One regular question is whether the PSC has any predictive power for KS2 results.
To answer the question, we have used data from 93,198 pupils whose 2019 PSC and 2024 KS2 Reading scores are recorded in Insight.
The distributions of the two sets of scores give a good indication as to why there is no simple connection between the two scores; for any individual child, the PSC’s assessment of decoding skills in Year 1 does not appear to have a strong link with the KS2 assessment of comprehension skills in Year 6.
PSC scores
The ‘pass’ mark for the PSC (the number of marks pupils need to have attained the expected standard) has been 32 since the PSC’s introduction in 2012. As can be seen above, the vast majority of pupils score between 32 and 40 (82% of scores are in this range in our sample), with the modal score being 40.
This data is clearly heavily skewed, with a strong ceiling effect; the PSC is not able to distinguish between the more proficient decoders scoring full or close to full marks.
KS2 Reading scaled scores
The scaled scores above are created using a look-up table for each of the potential raw scores (between 0 and 50); some scores have no raw score equivalent and some raw scores have the same scaled score. The data is slightly skewed to the left. Whilst scaled scores are created in a somewhat opaque manner, the expected standard is set at 100, with roughly 75% of pupils scoring above this score.
In our sample, the median score was 106 and the mean score was 105.4.
Connections between the two scores
Once again, we start by looking at the ‘correlation’ to assess patterns in the data. Correlations range from -1 (perfect negative correlation) to +1 (perfect positive correlation). A strong correlation (closer to +1 or -1) suggests that most pairs of observations are relatively close to a line of best fit for the two sets of data; a weaker correlation (closer to 0) suggests the pattern is more random.
For our sample of pupils who took the PSC in 2019 and the KS2 Reading SAT in 2024, the correlation is 0.48. This indicates a moderate relationship. Those who score well in their KS2 Reading are likely to have recorded high scores on their PSC; those who scored below 32 on the PSC are likely to have scored below 115 on their KS2 Reading.
A scatter plot with a line of best fit based on regression looks like this:
This indicates a moderate positive relationship; the higher the PSC scores, the higher the KS2 Reading outcomes, but with wide variation for any given PSC score. When considering the vast majority of PSC scores are in the range 32-40, and that candidates with any PSC score record almost any outcome in KS2 Reading – with the minor exception that those scoring 39 or 40 are unlikely to record KS2 Reading scores below 88, and those scoring below the pass mark of 32 are less likely to score above 110 at KS2 – the relationship between the two scores is relatively weak.
It is worth noting that there will be some degree of measurement error in both assessments as well, which further weakens any connection between the scores. The nature of the PSC means that it has almost no predictive power for the majority of pupils. Those who struggle with decoding in Year 1 do not necessarily struggle with comprehension in Year 6 and vice versa.
Leave a Reply