Monday, November 29, 2010

Education reforms can't ignore family influences

(Previously published in the Terre Haute Tribune Star, 11/28/2010)

The newspapers are full of coverage on education reform. In last Sunday’s New York Times, Thomas Friedman went so far as to say the Department of Education was the “epicenter of national security.” His op-ed paints a dour picture of the US education system. He endorses the Obama Administration efforts to transform the US education system thereby indirectly endorsing the current efforts in Indiana. Friedmann points favorably to the educational systems in Denmark and Finland (Finland especially has a wonderful educational system). However, pointing to such “socialistic” countries isn’t much of a selling point.

Gov. Daniels’ penned an article in the November 7th Indianapolis Star. He wrote: “If there is one fact that every expert and all the data confirm, it is that the single most important predictor of a child’s academic success is the quality of the teachers he or she encounters.”

I’ll forgive Gov. Daniels for favoring politics over the known science on academic performance. The best predictor of a child’s academic performance is their parents. It is that inconvenient fact that “No Child Left Behind” treats as an excuse instead of explanation and thus ignores parental contribution (or subtraction) to a child’s academic performance.

Research on family type (a different measure than socio-economic status) shows consistent effects on child academic performance (and not just in the US). Children from larger families tend to underperform relative to children from smaller families. Children from single headed families are the most likely to underperform and that type of family is increasing across every category of socio-economic status in the US.

Friedmann’s article was 12 paragraphs long. He waited until the last one to mention parents. Other than the use of proxies, such as parent’s income and education, I couldn’t find any research on the “qualities” of parents that contribute to their children’s academic success. There is research on how parent’s educational goals for their children affect their children’s educational goals. Unfortunately nothing that connects any parent quality to test scores, other than the proxies of socio-economic status and family type. Talk to teachers, however, and one gets enough anecdotal evidence to suggest the need for more systematic research on the subject.

What do you think? Who is likely to perform better on standardized tests? Children who are or not read to at home? Children whose parents monitor their children’s homework and academics or not? How about the simple act of asking kids what did they learn today and not accept “nuthin” as an answer? Kids who eat breakfast in the morning or those who could eat at school, but whose parents can’t get them to school on time? Kids of parents who meet with and work with teachers when a child is struggling or those who refuse to meet with school officials, even when school officials are willing to meet outside the normal school hours? This list could go on and on, but I think the point is made.
Our leaders’ answer to this social problem is merit pay for teachers and school choice. To date, the largest experiment with merit pay for teachers is Tennessee, the results released this year, indicate merit pay (bonuses were significant averaging $9600 to $11,300) had little to no impact on children’s academic performance and what positive effects were found, diminished over time.

Just as Americans (and American children) are getting fatter and as a result diabetes is on the rise, do we conclude then that there is a problem with US doctors?

Those who are interested should find Hanushek and Rivkin’s article in the May issue of the American Economic Review. In short, they lay out very well the problems of tying value added measures of student performance to teacher pay, and examine the relationship between teacher quality and academic outcomes. And contrary to Andrea Neal’s claim in Thursday’s newspaper, Hanushek and Rivkin examine observed differences in teacher quality (GPA, their own tests scores, etc) and find no difference on student outcomes. This is not to say there are no bad teachers. According to Hanushek and Rivkin, just replacing 6-10 percent of the worst teachers and replacing them with average ones (in terms of student outcomes), would make a significant impact on our education system.
Reformers commonly say the US has a good education system for 1950. Arguably, the US had better family situations and fewer distractions for academic performance then than we do today. The education problem can’t be solved by ignoring the contributions, good and bad, of parents and families.

Note: Steiger is married to a public school teacher.

Sunday, November 21, 2010

Merit pay may not bring positive results

Previously published in the Terre Haute Tribune Star, 21 November 2010

TERRE HAUTE — Gov. Daniels gets what he wants. He wanted to change the time in Indiana; no one talks about repeal. He wanted to lease Indiana roads for badly needed cash and he did it. He has set more land aside for conservation than probably any Indiana governor and no one criticized him. When he came into office six years ago, he faced a budget shortfall and fixed it without a tax increase (although he did suggest one, remember?). He pushed for a constitutional amendment on property tax caps and got it. And now, perhaps his final accomplishment before he sets his sights on a presidential run, is to fix Indiana schools.

His plan is simple. Use money to motivate teachers to do a better job and tie student academic performance to teacher pay and continued employment. What could be simpler?

I have seen “merit” pay up and close. I am not sure it changes anyone’s behavior. Why? Among many reasons, the amount of money is key. Gov. Daniels isn’t looking to add money to education, quite the opposite. That is why he pursues unproven educational reforms but doesn’t make full-day kindergarten, a proven means to improve student academic performance, a priority.

Pay for performance works when individuals can alter behavior that directly influences the measured outcomes. For example, sales professionals make more phone calls, contact more people, and result in more sales and more commissions. What behaviors do teachers need to do more of, or change, to increase student academic performance? Are there evidence-based practices that point the direction?

Learning theory is pretty straightforward: time on task impacts learning. So, maybe teachers will assign more homework. If the students don’t do the work (and parents complain about the amount), then student performance doesn’t improve and teachers won’t get paid for their performance.

Teachers are professionals, like lawyers, doctors, nurses, clergy, and accountants. Professional norms control their behavior. And like other professional groups, talent and skill differ across individuals. The “best” lawyers, however, are not necessarily the highest paid. Why would a highly paid lawyer accept a much lower paid position as a judge? Indeed, a judge is a civil servant while most lawyers are in private practice; they do it, in part, because of the honor.

Everyone cannot have the “best” doctor, lawyer, nurse, priest or accountant. Through training and professional norms, a minimum performance standard is established. No, such systems are not perfect and people in such professions can change, they can “burn out.” Are there bad doctors, nurses, lawyers, priests and accountants? Yes and there are “bad” teachers and there has to be a way to deal with those low performers.

When I directed the Sociology Research Lab, I compensated callers two different ways. One paid them for performance, the other was to pay by the hour. When callers were compensated based solely on performance, I found I had a lot of “cheating.” Callers would pretend to call and just answer the questions themselves. It was easy to spot these and remove those interviews from the data set. Others, who were paid by the hour, didn’t cheat. Some were phenomenal. They completed interviews at two to three times the rate of others. I established a reasonable quota of completed interviews per hour. Those who couldn’t make that quota usually quit. A few I had to let go because they were “too expensive.”

I used the first system to train and to gauge the work ethics and habits of callers. The “good” ones were invited to move to an hourly pay system. This system worked because I could establish and easily measure outcomes and establish standards.

How many points on standardized tests should be adequate to keep a new teacher employed? If those benchmarks are not reasonable or if they are vague or unstated, those who are subject to them will not have confidence in them, “cheating” will occur, and cynicism will grow.

I applaud Gov. Daniels for trying to fix education. I suggest he change the teacher retirement system so that burned-out teachers don’t feel handcuffed to a system that only rewards extremely long service. Push for full-day kindergarten. And quit bashing teachers or the most idealistic will instead choose sales, instead of teaching. Continued bashing of teachers will stigmatize the profession and eventually those who make the best teachers will turn away.

Editor’s note: Steiger is married to a public school teacher.
Blog Directory - Blogged The Steiger Counter at Blogged