stage, and concluded ‘Just over three-quarters of the children expressing an intention
in year 7 carried out that intention post-16’ (p.406). This sounds impressive evidence
for the accuracy of early intentions, and indeed this apparently high level of accuracy
forms the basis for his whole paper.
Table 1 - Reported intentions for post-16 participation, year 7 pupils
|
Stay on |
Leave |
Don’t know |
Total |
|
725 |
ЇЇ8 |
238 |
1081 |
Source: Croll (2010, p.406)
Croll does not say exactly how many of the 1,081 pupils in Table 1 did indeed stay
on, and this is an important piece of data missing from the paper, and from his
consideration of the meaning of the 75.9% ‘accuracy’ figure. However, the overall
staying on rate for the various panels he considers (totalling 1,715 individual pupils)
is 71.6% (p.406), so let us assume that this is something like the true figure for the
1,081 pupils questioned in year 7. In fact, 72% could be an underestimate of the
population figure for Britain, which has been over 70% participation in full-time
education post-16 for many years now, topped up by others participating in part-time
education, plus employer and government-funded training schemes (Gorard et al.
2007). By ignoring those replying ‘don’t know’ in Table 1, like Croll, we end up with
843 pupils in year 7 (1081-238), of whom 86% (725/843) reported an intention to stay
on in education. So now we have the three numeric parameters we need to make a
judgement about the claim to accuracy. In this study, 86% of year 7 pupils intended to
stay on, around 72% did so and, added to those intending not to stay on and not
staying on in fact, these intentions are 76% accurate. The first thing to notice, of
course, is the imbalance in the two prior figures. A clear majority of pupils intend to
stay on and a clear majority do so. Table 2 summarises this imbalance. It also divides
pupils into a notional 72% ‘observed’ to stay on and 28% observed to leave for both
those pupils intending to stay on and those intending to leave. In other words, Table 2
provides the baseline of frequencies that would be expected if intention in Year 7 and
participation after Year 12 were completely unrelated.
Table 2 - Assessing the accuracy of intentions if participation is unrelated