Saturday, February 20, 2016

How High-stakes Assessments Are Impacting Students, Teachers, and Schools

High-stakes testing – the widespread use of standardized tests that can affect, at times, the futures of teachers and students – has greatly increased in this era of policymakers striving for “accountability,” and in some states has been a wave that has crested, leaving everyone looking for an elusive new equilibrium.

The criticism of the testing regime that has grown vary familiar in the United States centers on these points:

•    It narrows the curriculum by excluding subject matter that isn’t tested
•    It reduces learning to the memorization of facts easily recalled for multiple-choice testing
•    It diverts too much classroom time to test preparation rather than learning

Educators report that the testing emphasis has led to a narrowing of the curriculum, focusing more on critical reading and math skills.  Some think this narrowing shortchanges students from learning important subjects while others see it as the recipe to help low-achieving students catch up.
All in all, for all its flaws, high-stakes testing does seem to increase the amount of learning, and “emerging studies suggest that teaching to the test can be good or bad: Good if it means teaching a focused and aligned curriculum; bad if it reduces instruction to the memorization of test items.”

I spoke to two teachers in New York state public schools to get their perspectives on the effects of high-stakes testing in recent years. Here’s what I found.

In New York, as of this year, the connection between high-stakes testing results and teacher evaluations has been greatly reduced, as a result of a parent popular uprising against too much testing. The governor, in essence, declared a four-year moratorium on the practice. But it would be a mistake to suppose that tests are not continuing to be a major presence with a big impact. For one thing, it depends on how the district’s union negotiated the last contract, and in places it can still be between 20 and 50 % of a teacher’s score.

Tavis teaches 9th grade global studies, for which there is no test. As a result, his assessment is based on results of school-wide Math and English Regents exams which are entirely unrelated to the subject matter he teaches – a situation which he finds to be ludicrous. He also teaches AP Psych, for which students spend significant time in test prep, taking 13 exams leading up to the AP test.

Steve considers himself lucky because he mostly teaches in a special STEM program that is unrelated to any standardized testing. 2 of his 10 classes, though, are in 5th grade math, which is tested heavily: a “Star Assessment” 3 times/year; “benchmarking” tests of teachers; Common Core state assessments; and, last year, field testing for a future test, which he finds ‘criminal,’ essentially free market testing for testing behemoth Pierson. He feels there is too much testing.

Both Tavis and Steve affirm that they and other teachers most definitely teach to the test. Tavis reports that students are taking the AP course to have it on their transcript, and for credit, and they need a 4 or 5 for credit they have to get good results. He says that those teachers who don’t deliver results, the admin asks them “what’s going on.” He says the administration “says they want you to teach various ways, but in the end they look at the scores.”

Steve agrees, but he also highlights the other side: “Content knowledge in tests, especially as per Common Core, is not inherently bad. And we need assessments, even if the way they’re doing it is bad. I’m being hired by them, it’s my job, I can’t just say my ‘y’ is better than your ‘x.’ And a lot of teachers are much more comfortable teaching to the test. It gives them a framework, and without it, they flounder a bit.”

Does this testing increase pressure on the students? Both teachers say it does, but Tavis emphasizes that there have been tests for a long time, and students put pressure on themselves. But now, the students have to pass tests in order to graduate. It used to be, the school would “find a way to make them pass.” And now, with test directly tied to teacher, teachers put more pressure on the students to perform.

Steve suggests that the test highlight the value of being a high performer. “Face it,” he says, “it’s status in the classroom. It’s real.” And in tune with the research of Carol Dweck, he sees how test success translates into a desire by student to preserve the appearance of high performer, over intellectual curiosity. High achievement leads them, he says, away from wanting to explore processes more deeply. And the irony, he says, is that this this effect is exactly antithetical to the Common Core goals.

Considering Teacher Evaluation

It’s time to consider teacher evaluation. What practices lead to effective teaching outcomes? On what elements should a teacher be judged?

A most instructive starting point is to examine the broad principles and propositions for teacher evaluation put forward by the National Education Association (NEA), Interstate Teacher Assessment and Support Consortium (InTASC), and the National Board for Professional Teaching Standards’(NBPTS). I am not going to reproduce them all here; they can be read on pages 6-10 (PDF pages 12-16) of the NEA document “Teacher Evaluation: A Resource Guide for National Education Association Leaders and Staff.”

The big common themes among all three are:
  • Understands how learners grow and develop
  • Demonstrates in-depth content and professional knowledge
  • Understands and uses multiple and varied forms of assessment
  • Establishes environments conducive to effective teaching and learning
  • Integrates cultural competence
  • Develops collaborative relationships and partnerships
  • Provides leadership
  • Participates in ongoing professional learning
I see no reason to argue with those guidelines or attempt to innovate. They are justifiable professional standards that are clearly oriented to creating the conditions for effective teaching and learning, and they can hold teachers accountable in meaningful ways. They provide a cohesive and coherent professional context and sense of direction.

Undoubtedly, though, how assessors arrive at judgments about individual teachers is a different matter, needing a strong effort to achieve validity, minimizing bias and variability.  Clear, rigorous expectations, multiple measures, meaningful ratings, regular feedback and meaningful, actionable implications, all as defined by the New Teacher Project document “Teacher Evaluation 2.0,” are essential to the process. (I’ve been evaluated in other professional contexts and found processes that pretty well lack all these qualities!)

I was really impressed with the 2012 article in The Atlantic, “Why Kids Should Grade Teachers,” by Amanda Ripley. It shows that students’ evaluations of their teachers, collectively, are very accurately correlated with the results of other measures. I would be more than happy, as a teacher, to have my evaluators put significant weight on student evaluations. Of course, other data sources are necessary, to continue the validation of correlation.

I looked at the Tennessee Value-Added Assessment System, a measure of what value individual educators add to their students’ educational growth. For grades 4-8, the system measures the growth, gr 4-8, of a single student from one year to the next. It uses state test assessment data, with various factors to adjust and create “scale scores” that show a student’s performance in relation to other students across the state. The starting line is different for each child – it’s based on their previous year’s scores. Then teacher is evaluated based on results for all his/her students.

It’s all very rational. But I would be concerned about the amount of testing it requires, and how much such a system drives teachers towards teaching to the test, and pressuring students on that basis. Also, in my conversations with New York state public school teachers, I found a lot of frustration with teacher assessment based on student test scores. Two teachers spoke of being evaluated on testing that everyone acknowledges doesn’t even assess the subject matter the teachers teach. I can’t imagine Tennessee having solved this except at the cost of testing the students way too much, which has caused a great outcry in New York. I would want it to be put in a bucket along with in-person assessment by administrators and teacher peers, and student assessments.

I also looked at the Ohio Department of Education’s State Board of Education Approved Framework, and I liked how it offers an “alternative” assessment component that can include student surveys and student portfolios. If that alternative option is chose, the final summative evaluation draws on 50% teacher performance (as evaluated by school staff), 35% student growth measures (similar to Tennessee’s), and 15% alternative components. As for me, I’d be willing to up the percentage of the alternative measures.

Saturday, February 06, 2016

On pre-assessment and innovative differentiation strategies

I have developed a unit for 8th grade social studies, addressing the New York Social Studies standard 8.1f, “Muckrakers and Progressive Era reformers sought to address political and social issues at the local, state, and federal levels of government between 1890 and 1920. These efforts brought renewed attention to women’s rights and the suffrage movement and spurred the creation of government policies to enact reform.”

In a nutshell, the activity is for students to select any aspect of the era that interests them beyond the material for which they are all responsible, research the topic, and figure out how to present it to the class in an interesting and engaging way that also relates in some way to present times.

What follows is a hypothetical case study.


To get a sense of what the student know, and especially of what gaps exist in the students' knowledge around this topic, and who is ahead on this topic and who knows little about it, I began with a simple pre-assessment, a Quizlet with a number of terms that embody the era and a task to match them up with their definitions. Quizlet is a really neat tool for quick assessment; I could have used a more standard quiz format such as matching, multiple choice, and fill in the blank, but this time I chose to use the scatter function, in which the student must match up terms and their definitions against a timer.

Differentiation strategies

Next, I sought ways to provide innovative differentiation strategies to work with the fact that some students already seemed to have the topic down, and others were discovering it for the first time. The methods are shown in this lucidchart mind map

Top performers

These students will take a closer look at the question of contemporary relevance of the issues being investigated by the whole class, and will provide resources that others can use. They will create a "Then and Now" mapping of parallels issues and events, drawing a strong distinction between issues that seem to have been put to rest, those that haven't, and those that have successor/related issues today.

Some Awareness (the majority)

As these students work through the required concepts in an open classroom setting using varied resources (texts and online, we will have a running competition for those who get there first to speak up and share what they’ve found, provide sources, and generate visual aids for digital display in a group resource of findings.

Those with Limited Knowledge

These students will be paired them up with others who are tracking down the same concepts, to share, explain, and provide peer assessment. For students who would have significant difficulty presenting their topic solo, there will be opportunities for practice sessions with peer feedback, preferably from volunteer 'advanced' students.