A new paper has been published by Bruce Fuller and his colleagues at Berkeley. It seems to demonstrate the advantages of taking a more academic approach to preschool education. This is no surprise. I recently reviewed a report by the Brookings Institute that came to much the same conclusion.
However, a peer-reviewed paper is always welcome and this one utilises an interesting quasi-experimental design. I don’t know enough about ‘marginal structural models’ to offer a critique but it seems like a reasonable attempt to investigate causal relationships when a true experiment would be pretty hard to conduct.
Fuller and colleagues suggest that they:
“…observe positive benefits on the average child’s cognitive proficiencies after about five to six months of attending a preschool that is academic-oriented, and these effects display stronger magnitudes than prior studies with national samples, where investigators did not focus on academic intensity, as one specific element of classroom quality.”
Their argument relies to some extent on comparing effect sizes. I am far more comfortable with the way that this paper compares effect sizes for similarly designed studies carried out with groups of students of similar ages than with the way that effect sizes are often (ab)used in meta-analysis.
Crucially, the research established no relationship between an increased academic orientation and the level of social development, contradicting the idea that an academic orientation is harmful in this way, and the positive effects on academic achievement persisted into kindergarten; the first year of formal instruction in the U.S.
One point that is often raised about preschool programmes is that the effects don’t tend to persist. Even if they are significant in kindergarten, other studies should lead us to expect them to wash out over time. Is it therefore a waste of time to invest in more effective preschool programmes?
I wonder if we could consider this differently.
Imagine a two-stage race. The first stage is a classic running race over 400 metres. In this event, the difference between the best kind of training and the least effective training generates an average variation of a few seconds among participants. It might not sound like much but in the context of this kind of race it is a real and significant difference.
However, the second stage of the race is a 400 metre egg-and-spoon race. For those of you who are unfamiliar with the concept, in an egg and spoon race contestants must walk as quickly as they can while balancing a raw egg on a spoon in front of them. If they drop the egg, they have to start this part of the race again.
In the case of such a two-stage race, the advantage gained of a few seconds at the end of stage one would likely be completely washed-out by the end of stage two. This is not because there is anything inherently wrong with the stage one training.
If students exit an academically orientated preschool and enter an elementary school that uses a ‘balanced’ literacy approach such as New South Wales’s L3 programme, we cannot expect them to utilise the literacy advantages that they accrued at preschool because the elementary programme is simply not designed to do that. We already know that the most effective literacy programmes involve systematic synthetic phonics and not the analytic, in-context and multi-cuing approaches typical of balanced literacy.
The question we need to consider is not whether it is worth having effective preschools but whether it is worth having elementary schools that can be just as effective at moving students forward.