By Stephen Krashen
1. Response to Arne Duncan’s claim that his policies succeed in overcoming poverty:
Duncan states that schools and their “local partners” are “overcoming poverty” by “investing in teachers, rebuilding school staff, lengthening the school day and changing curricula.”
I know of no evidence that this is so. Rather, the research indicates that there are very few high-performing schools in high poverty conditions. Also, to my knowledge, no detailed studies have emerged with descriptions of rebuilt schools with longer days showing consistent, startling progress.
There have been occasional media reports (e.g. Felch, Song and Poindexter, 2010), but these cases of improvement are sketchy. It is not clear whether scores are being pumped up by test prep or are the result of genuine teaching and learning.
The lack of comparison groups makes it impossible to dismiss the possibility that all students in the district are getting better, possibly due to the introduction of new tests and “test inflation,” improvement due to greater familiarity with the test. Gerald Bracey (2009) reported that one highly publicized “success story” published in The New York Times about the Harvard Promise Academy, was true only for one grade, one subject and for one year.
Duncan gives the impression that “overcoming poverty” happens all the time under his administration. There is no real evidence that it happens at all.
2. Response to Arne Duncan claim that there is widespread support for new tests.
Yes, we all want accurate ways of measuring student growth. But does this mean we must have new tests and more testing than has ever been done before? I think we already have a wonderful and accurate way of “accurately measuring what children know.” It also “helps inform and improve instruction.” It’s called teacher evaluation.
There is no evidence that extensive testing does a better job than teacher evaluation done by professionals who deal with children daily.
The plan presented in the Department of Education’s Blueprint for Reform calls for an astonishing amount of testing, far more than we have now with No Child Left Behind. The only people I know who support the testing plan have spent very little time in schools, haven’t read the Blueprint, or just aren’t listening to real education professions or students. Or all three.
We are about to make a mistake that will cost billions and make school life (even more) miserable for millions of teachers and students. The only ones who will profit are the testing companies. We should be talking about reducing testing, not increasing it.
3. Response to Arne Duncan’s claim that more and more people want “a real definition of teacher effectiveness” and multiple measures.
Duncan wrote: “More and more, teachers, parents, and union and business leaders want a real definition of teacher effectiveness based on multiple measures, including student growth, principal observation and peer review.”
No: More and more, Arne Duncan, Bill Gates and companies in the testing business want value-added standardized test scores (widely acknowledged to be inaccurate in evaluating teachers), and want to video-tape teachers to make sure they are focused on test/standards-related items in class. There are no teachers, union members, or parents marching in the streets and writing angry letters demanding new and more rigorous measures for teacher evaluation.
Most important: There is no evidence that there is a crisis in teacher quality, no evidence that teacher quality has declined. When we control for poverty, American students score at the top of the world on international comparisons.
The problem is poverty.