One important issue discussed in my new book, From Preschool to Prosperity, is what can be done by government to increase the quality of local pre-K programs. Although there is good evidence that a wide variety of state and local agencies can successfully implement quality pre-K programs, there is no guarantee that any particular local program will be high-quality. Regardless of whether state governments or the federal government finance pre-K, there will be some demand for accountability for results.
For pre-K, federal and state governments should resist the pressure to over-emphasize standardized tests in evaluating individual pre-K providers. I believe that there should be regular monitoring using tests of whether the overall state or federal pre-K program is improving kindergarten readiness. This can be done through comparing test scores at pre-K entrance and kindergarten entrance, and adjusting for the improvements that are due to aging alone, without pre-K, by using the range in test scores with age among each group of children. (Technically, this is referred to as doing a “regression discontinuity” study of pre-K, where we see how test scores vary with age, and then see what “jump” there is at kindergarten entrance.)
Such an evaluation of the overall program answers the important question: is this program improving overall kindergarten readiness? However, state or federal agencies overseeing pre-K funding should resist using measures of test score gains at kindergarten entrance to evaluate individual agencies, for several reasons.
The most important reason for being wary of standardized-test based accountability of individual pre-K agencies is that such a system would put enormous pressure on individual agencies to aim at maximizing the performance of 4-year-olds on standardized tests that examine cognitive skills. While cognitive skills are important, pre-K should also be aiming at improving social skills and character skills, which are hard to measure using standardized tests. Narrowing pre-K to a cognitive skills-only approach is undesirable.
For individual agencies, one approach to accountability is to use measures of classroom quality derived from the observations of trained outside observers, for example the CLASS measure of pre-K classroom quality. These measures have trained observers come in and periodically rate in a reasonably objective fashion various features of how children are interacting with each other and with the teacher. In the case of CLASS, there is some evidence that this measure is correlated with later test scores gain.
However, the CLASS measure has the significant advantage in that in looks at both quality features related to cognitive skills and quality features related to social and character skills. For example, the CLASS measure includes some measures of whether the classroom climate is positive.
The CLASS measure also has the advantage of providing useful feedback to individual programs and classroom teachers. Test-based accountability measures are more of a black box. It is difficult to use standardized test results to figure out how to improve program quality.
Observational classroom measures might be complemented by teacher mentoring/coaching approaches. Experienced, high-quality pre-K teachers might add their observations of individual classrooms, and meet with the class’s teacher to make suggestions for how the class’s quality could be improved.
Accountability is a legitimate demand of taxpayers. Quality is essential for pre-K programs to have high benefit-cost ratios. The challenge is developing a clear-cut accountability and quality improvement system that doesn’t encourage strategic behavior to maximize some test score, and that truly provides assistance that makes a difference to quality. Classroom-based observations coupled with follow-up by teacher mentors seems a reasonable approach to try. As we gain experience, we should be able to improve the usefulness and reliability of these classroom-based observations, and be able to focus the teacher mentoring on the most important elements of quality.
I share your concern about narrowing the focus of pre-K to cognitive skills only. But what if we used a comprehensive assessment (such as Teaching Strategies Gold, for one example) that assesses skills beyond the cognitive domain. Would you still not want those used to evaluate individual agencies?
Thanks for your comment. However, from my understanding of the research literature, there is serious concern that we do not have adequate measures of non-cognitive skills. For example, there is some evidence that when the classroom climate improves, and the general standard of behavior improves, that ratings of non-cognitive skills by teachers and other observers tend to become more stringent. In other words, we may not be able to compare non-cognitive skills ratings across different classrooms, and we might even find skill ratings to go down as classrooms improve.
Pingback: Increasing quality in preschool through accountability and mentoring by timbartik « St. Joseph County Great Start Collaborative