Once again, the topic of money and student outcomes has been put squarely on the front burner. In response to lower NAEP scores the Kansas Association of School Boards has presented written and verbal evidence that a correlation exists between the volume of money spent on education and test scores. But simultaneously, they are careful to say more money doesn’t actually cause higher scores.
But the unspoken message from KASB is loud and clear: Give us more money and test scores will go up. The latest installment is provided by KASB researcher Ted Carter in this blog. Mr. Carter is careful and clear to say that there is no statistical evidence that a causal relationship exists between money and outcomes. However, his concluding remarks contrarily imply that there is one. KASB’s approach is much like that of the Publisher’s Clearing House sweepstakes. PCH is obliged to disclose that there is no purchase required to win the big prize, but they still want you to believe otherwise.
Correlation and causation: two terms that kind of sound the same and are often confused as interchangeable. But they mean two very different things.
When I was just a pup in college, cutting my teeth on research methodology, one of my professors liked to use a particular example to distinguish between correlation and causation. Back in those days, as we were told, there was a strong correlation between red automobiles and car accidents. Could a car’s color, in this case red, actually cause accidents? Could they not be seen as well by other drivers? Did they blend into the highway landscapes? Did they not have steering wheels? Brakes? Of course those were not the reasons.
That’s a cut and dried example of correlation without causation.
In graduate school, I was fortunate enough to study under two professors who wrote what was at the time considered the bible of social science research, Political and Social Inquiry. The authors, Dickinson McGaw and George Watson specified three defining characteristics of causation:
- Temporal precedence – a fancy term which simply means the cause must precede the effect.
- Constant conjunction – another fancy term which means that whenever the cause is applied the effect occurs.
- Non-spuriousness – that mouthful means that there is no other factor that could cause the observed effect.
It’s that last one that is so tricky when doing research in the social sciences. In the case of money and education outcomes even KASB allowed that other factors could influence outcomes “like median household income, education level of adults, population per square mile, percent of students with disabilities, percent of students eligible for free and reduced-price lunch, percent of students eligible for ELL services.” Although I agree these do represent factors that could potentially influence student outcome, they only cited factors external to the provision of education. After spending 20 years in the public school system, I have witnessed many internal factors of the education system that could also impact outcome. Things like quality of teachers, curriculum, school choice, scheduling, district organization, school culture and climate, and, of course, how schools choose to spend their money.
Here’s what I mean. Let’s say two school districts, districts Aand B are given x number of increased dollars. District Achooses to hire new elementary teachers who directly target lower achieving students in core areas, such as math, reading and language. They also hire additional secondary teachers to help students who are at-risk of not graduating or need assistance with technical education or college prep. District Bchooses to use the money to give an across the board pay raise to teachers (regardless of effectiveness), hire more administrators and provide extra retirement benefits for certain employees (in addition to KPERS).
Which district would you predict more likely to have a stronger correlation between money and outcomes?
It’s time we reevaluate money in education, recognizing it as a means to an end and not an end in itself.