Friday 5 June 2015

Is it worth giving children pre-school education?


It may seem sacrilegious to suggest otherwise, but perhaps we make too much of early childhood, and harbour illusions which are, quite frankly, blank-slate-ist. Blank slate-ists, as you already know, are deluded persons who imagine that children’s minds are a blank slate upon which care-givers can impose their will. Leave aside the fact that from birth onwards it is evident that neonates differ in their reactions and behaviours, notably whether they are calm and easy to care for, or fractious and unable to sleep through the night (easy, difficult or slow to warm, as Alexander Thomas and Stella Chess described them in the The New York Longitudinal Study on temperament and personality, started in 1956). Blank-slate-ists ignore or downplay temperamental differences, and believe that the waking moments of early childhood have sacrosanct status, and are ripe for heroic interventions. I should confess at this juncture that, for most of my professional life, I have been a big fan of early childhood intervention studies, so I am struggling with my prior assumptions at this point. Onwards, rational beings, onwards.

A converse view is that kids will be pretty much OK so long as they are given the basics that most families provide: a reasonable life, though nothing fancy, which is how most humans were raised for most of history. Indeed, prior to 1840, getting too attached to children before the age of 7 did not make much sense, because one third of them died before that age. John Graunt’s 1662 comments on the Bills of Mortality show that in England only 64 in 100 persons survived to age 6. In fact, there was little point getting too attached to teenagers, since another third of them had died by 16. By the way, the 1993 figures are ancient history, because now people live for ever (almost).


(The bottom of the slide shows that in 1662 1 in a 100 made it to 76 years of age, in 1993 70 in a 100 reached that age).

I digress. Even when children have a 99+% chance of surviving into adulthood, it is still an open question whether extra education should be given to all children prior to age 5, or at age 11, or 13, 14, 15 or only at precisely the age when an individual child encounters a problem. Early intervention studies can only follow general risk guidelines, like betting that poor and dull mothers will have children who, from an environmental rather than primarily a genetic point of view, will benefit from an early childhood compensatory education program. Waiting until the child is older, while providing standard subsidised education from 5 or 6 or 7 onwards will give good results for most children, and extra help can be targeted on those who really need it precisely when and how they need it.

In his paper to the London Conference, Andrew Sabisky argued in favour of targeted later interventions. Most striking, to my mind, was a meta-analysis of the early childhood intervention studies, the key figure from which I reproduce below.


Sabisky has gathered a lot of data for his presentation, so you should look at the slides first, and then at his comments which go with the slides, and then at the review paper.

My impression is that he has made a good case for abandoning the view that early childhood is a special period for learning (which the Danes have known for ages) and that you might be better of letting kids enjoy family life and the usual experiences of childhood until their brains are capable of beginning academic learning, in the age range 5 to 7 according to ability and interest.

See what you think.

The slides:

This leads you to the notes on the slides 

The Duncan and Magnusson 2013 paper from which I took Fig 2


  1. I went to Nursery School. I admit my memory is lousy, but I don't think we were taught anything, except to rub along with each other. Primary School was different - we were streamed. I suppose that was consistent with your remark "until their brains are capable of beginning academic learning, in the age range 5 to 7 according to ability and interest".

  2. Is there any evidence that non g-loaded gains mean anything in the real world? To paraphrase Linda Gottfredson, g is the active ingredient in mental ability.

  3. I can't find any reference in the paper you linked to to Engelmann's Direct Instruction preschool programme, or to the large scale study Project Follow Through, which found that it was so much better than Perry Highscope. A discussion just published here gives some details:

  4. at least one rigorously standardized ability test for children in the last 25 years took the trouble & expense to norm the sample on (among other things!) pre-school attendance (yay vs. nay). the norms were the same either way. so it didn't matter. those were heady days back when we thought it might matter:)

  5. Speaking of pre-K, you may be interested in my latest blog post on socio-economic mobility. I analyzed a new data set produced by the US department of education, which includes test scores and GPA (categorical variable), income, SES, and more to assess SES mobility (in the main).

    I tried to make this one a bit more a bit more brief than the last post I shared.... :-)


  6. Err, that should read "a bit more brief than the last post I shared".

  7. What if the child in question doesn't have a family? What if he sees one parent (usually the mother) at intervals, and never sees his father? What if the alternative to preschool is social isolation? Or the lackadaisical care of this month's boyfriend?

    The hearing children of deaf parents still had two parents. Could the decline in the line in the graph be related to the increasing prevalence of single parent families?

    The argument for later, targeted intervention would only work in a world without politicians. Money is fungible. Money not spent on preschools is not guaranteed to be spent on high school interventions. In today's world, I'd wager that it would be much more likely to be spent on caretakers for senior citizens.

    Infants don't vote. Single parents tend not to vote. Seniors vote.

    In our well-to-do community, I have seen the value of early childhood programs, if only to identify children who need more intensive interventions. I have seen friends' children make progress. It often took heroic efforts from the family as well as the school. Such efforts may be more effective in two-parent families that can afford to sacrifice one parent's career.

  8. I'm more or less on board with conclusions regarding the utility of early childhood education. But what a delight my then 3 year old daughter's carefully designed Montessori school was to her! A constructivist model is ideally suited to the highly independent temperaments of the 3 year old set.Paula are has done some interesting work in this area. Time will tell. Thank you for your column.

  9. Pre school education is must for developing creative little minds. Fun activities keeps the tiny tots engaged removing the thought of going back home.

  10. This comment has been removed by the author.

  11. Early education is something that cannot be ignored,A child's first time away from home in the real world, first interaction with people other then family.Schools such as Williamsburg Northside Preschool focus on academic learning incorporating some fun activities for children to love it.