We have just supported a number of schools in analysing the outcomes of the recent, controversial, Year 1 Phonics Check. The Phonics Check measures a 5 or 6 year olds’ reading level in terms of their ability to “decode phonics” (including fake words such as yop and groiks). The outcomes for schools will not be publically available, and so schools will not be directly held to account for their pupil’s achievement in the check; however, what can our analysis tell us about how important teachers consider their pupil’s scores?
Well, the expected level for the check is 32 points (out of 40). When we see the distribution of scores for a random sample of pupils, we see that, generally speaking, more pupils achieve each score the nearer to 40 we get (with 40 out of 40 being the most common score). However, there are a couple of clear anomalies shown in the graph below; the number of pupils achieving 31 points (just below the pass cut off) is well below the trend line (less than 2% of all pupils, where the trend line expects nearly 4%), while the number achieving exactly 32 is way above the trend line (over 7%, instead of the expected 4%).
Of course, this may be a quirk of the data… or it could be that we are subconsciously keen for our pupils to hit this 32 mark threshold (not least because they must be rechecked the following year if they do not reach the pass mark).
One other interesting finding from the question level analysis we carried out; there were 20 real words, and 20 fake (or “pseudo”) words in the test. Pupils were no more likely to correctly read “real” words correctly than they were to read the pseudo words.