British Educational Research Journal

The statutory ‘phonics screening check’ was introduced in 2012 and reflects the current emphasis in England on teaching early reading through systematic synthetic phonics. The check is intended to assess children's phonic abilities and their knowledge of 85 grapheme–phoneme correspondences (GPCs) through decoding 20 real words and 20 pseudo words. Since the national rollout, little attention has been devoted to the content of the checks. The current paper, therefore, reviews the first three years of the check between 2012 and 2014 to examine how the 85 specified GPCs have been assessed and whether children are only using decoding skills to read the words. The analysis found that out of the 85 GPCs considered testable by the check, just 15 GPCs accounted for 67% of all GPC occurrences, with 27 of the 85 specified GPCs (31.8%) not appearing at all. Where a grapheme represented more than one phoneme, the most frequently occurring pronunciation was assessed in 72.2% of cases, with vocabulary knowledge being required to determine the correct pronunciation within real words where multiple pronunciations were possible. The GPCs assessed, therefore, do not reflect the full range of GPCs that it is expected will be taught within a systematic synthetic phonics approach. Furthermore, children's ability to decode real words is dependent on their vocabulary knowledge, not just their phonic skills. These results question the purpose and validity of the phonics screening check and the role of synthetic phonics for teaching early reading.