Posted on 1 Comment

The effects of early childhood education programs don’t look good: a large, randomized pre-kindergarten study

There’s a large, new study out on the “Effects of a statewide pre-kindergarten program on children’s achievement and behavior through sixth grade,” and it’s important and unusual because of its results: it finds that pre-k education doesn’t help later educational or behavioral achievement and, if anything, hurts later student achievement. This study is also significant due to its comprehensiveness; the study follows 3,000 kids, who appear to be randomly assigned to pre-k services or not, and the study follows those kids for a long period of time—at least seven years, it seems, and possibly longer. I’ve not got the full manuscript yet but am seeking a copy. Most education studies are observational, in that they observe two or more cohorts, but they don’t use randomized controls, like this one does, and observational studies are particularly prone to bias. The new study is also pre-registered—that is, the authors say what they’re looking for, what success looks like, and how they’re going to measure success before they get their data. There’s a “replication crisis” in social science and medicine, because it’s possible to torture a positive result out of all sorts of data, and this study avoids most if not all of the common pitfalls.

The study’s abstract says:

Data through sixth grade from state education records showed that the children randomly assigned to attend pre-K had lower state achievement test scores in third through sixth grades than control children, with the strongest negative effects in sixth grade. A negative effect was also found for disciplinary infractions, attendance, and receipt of special education services, with null effects on retention

Wow: that’s counter to the intuition of most people, politicians involved in early childhood education, and “common wisdom.” The study is not the last word—no study is—but it is persuasive. For most practitioners, this won’t be immediately relevant, because Head Start and Pre-K For All aren’t likely to see real changes in the near term. But we may see the political winds change over time.

This post is not a policy recommendation: as grant writers, we don’t do policy recommendations, although I do think a lot of students are in college who’d be better served by alternatives, and yet society as a whole hasn’t yet figured that out or properly grokked it, even as total student loans owed passes $1 trillion. But, if America wants to do some form of daycare for all (“universal daycare”), as is proposed in the stalled Build Back Better legislation, that’s a fine goal and we should call it that, instead of pretending it’s possible to have academic, “educational” experiences for the vast majority of kids under the age of five. Four-year olds are not falling “behind,” because, except in the case of unusual prodigies, there is nowhere to fall behind. If anything, excess regimentation and premature optimization are likely to be bigger problems than “falling behind.”

I’ve long been somewhat suspect of early childhood “education”—not from studies per se, but from being around small children. Most don’t have the executive function to do much in the way of what might be called “education.” Trying to create “education” in the sense that we see with older kids or adults seems improbable for very young children. The veneer of “education” using “curriculums” like “The Creative Curriculum” and “The Creative Curriculum GOLD” that we cite in grant proposals seems faintly ridiculous; whether or not a four-year old can identify different kinds of leaves or songs or animals by name doesn’t seem to indicate how that four-year old will do in middle or high school, or college. But there’s a lot of social and economic anxiety around class, economic achievement, and housing; we’ve collectively adopted policies focused on creating scarcity, not abundance, and that’s resulted in intense, and probably pointlessly intense, competition in many fields.

Trying to indoctrinate small children into social, academic, and economic competition culture seems difficult to me, and yet that’s been one response to scarcity policies. Making early childhood teachers, who are really more like caregivers in the classroom, have degrees or advanced degrees seems like a way of raising the cost of childcare without providing much in benefits; everyday human experience seems to be sufficient for taking care of small kids. Maybe small kids are learning cultural markers and such in the early early childhood education setting that will help them later, but, if so, that later help isn’t showing up in the data. There’s a lot of desire to make education a panacea for various kinds of social and economic inequality, but that desire keeps running up against uncomfortable ideas (I won’t call them “truths,” although some might).

Head Start was launched in 1965 as on the initial programs in President Johnson’s “War on Poverty;” if there’s been a large boost in real educational attainment (which is different from “degrees achieved”), I’ve not seen it. I’ve been teaching college undergrads since 2008, and in that time my anecdotal impression is that smartphones and social media have been net bad for learning, noting however that some people do leverage Internet technologies to learn more and faster than they could without. Anecdotes are not data, but, since the late ’90s and early ’00s, we here at Seliger + Associates Grant Writing have been writing proposals for programs like the 21st Century Community Learning Centers (21st CCLC), and in that time we’ve not seen learning substantially improve from the dissemination of computers and the Internet. In 2013, I wrote a post about a pair of studies finding that computer access appears, if anything, to lower educational attainment. In 2015, I wrote about Kentaro Toyama’s book Geek Heresy: Rescuing Social Change From the Cult of Technology. “Computers in education” is not the same thing, obviously, as early childhood education, but both are attempts at improving education and life outcomes that are popular but may not be efficacious. If you work in the education industry with students ages 10 or higher—ages old enough for smart phones to have penetrated the population—ask those around you to look at their Apple “Screen Time” app or Android “Digital Wellbeing” controls. Those show how many minutes or hours a day a smartphone is being used, and what a person is doing on that phone. From what I’ve observed, very few people are using the book apps, the Duolingo systems for language learning, or Anki for space-repetition learning. Ask around, see what you find. Think about what that might mean.

Real education is hard. I’ve tried to impart some to students. Probably it’s always been hard and always will be. We should collectively try to do better while also understanding what might be limited, what might be futile, and what might be counterproductive. I’m struck by, at the college level, how little time is spent trying to learn how to teach more effectively, and friends who teach in K – 12 often report the same.

1 thought on “The effects of early childhood education programs don’t look good: a large, randomized pre-kindergarten study

  1. PLEASE!! Tell me I am missing something. The correct control group would have been children admitted to the program who did not have the intensive academic instruction. The correct control group is not the kids who stayed home (63%) and some who went to other programs.
    The variable to be tested was the extra academic education. Please tell me I am wrong and demonstrate this research means something. If the correct control group are the children who were admitted to the program but did not receive the extra academic instruction, it’s not too late to go back and look at the test results of the correct control group. If the correct control group was the kids admitted to the program who did not receive the academic education, the paper needs to be withdrawn.

Leave a Reply

Your email address will not be published.