highly significantp < .001
Considering just the main correlations, shown in bold, they are highly significant (all p < .001) between the two parts of MRCGP and the three parts of MRCP(UK).
Considering just the main correlations, shown in bold, they are highly significant (all p < .001) between the two parts of MRCGP and the three parts of MRCP(UK).
The pattern is not completely clear, but is certainly not compatible with the new CSA being less valid than the old CSA, nor with the CSA and AKT correlating less in BME than white candidates, and in both cases the pattern may be significant in the opposite direction.