Key Message: Tests that have the capacity to measure individual student progress provide the most useful information for improvement. Testing is just one of several meaningful components of measuring school and student success.

The implementation of No Child Left Behind (NCLB) has given unprecedented attention and power to standardized tests. It has also caused tests scores to be used in some ways for which they were not designed: evaluating teachers and schools and using results to determine future funding. The results can trigger rewards and penalties for teachers, schools and districts, without adequate regard for the student variables that are beyond the influence of schools. Also, many of the standardized tests provide very little information about individual student growth and progress.

Standardized tests fall into the following main categories:

Criterion-referenced tests. Criterion-referenced tests (CRT) compare student performance to a specific set of criteria. The Minnesota Comprehensive Assessments (MCA-II), the Minnesota Basic Standards Tests (BST) and the Graduation Required Assessments for Diploma (GRAD) are examples of criterion-referenced tests. That means they are closely aligned to what is taught and may indicate what and how much is learned.

• Criterion-referenced tests cover limited areas of content in-depth.
• Criterion-referenced tests usually have a specific level of performance that is considered “passing” or “proficient.” 
• Criterion-referenced tests measure students more accurately when they score near the passing score than when they score well above or well below the passing score.
• Criterion-referenced tests are usually not appropriate for identification of exceptional students (gifted/talented or Title 1; see previous)
• Criterion-referenced tests generally are poor at measuring individual student growth.

Norm-referenced tests. Norm-referenced tests (NRT) compare student performance to a norm (or average) group across content areas. The Iowa Test of Basic Skills (ITBS), the California Achievement Tests (CAT), The Stanford Achievement Tests (SAT), and the NWEA MAP tests are examples of norm-referenced tests.

• Norm-referenced tests measure students’ mastery of commonly-taught content.
• Norm-referenced tests utilize percentile rank as their main measurement indicator.
• Norm-referenced tests are not aligned to district curriculum; they do not identify how much an individual student has learned.
• Norm-referenced tests were not designed to judge instructional quality.
• Important note: It is possible to define a criterion score on an NRT (for example, students must score below the 40th percentile to be eligible for Title 1), and it is possible for a CRT to have normative scores (the MCA-II reports percentile ranks each year, which describe how a student performed relative to all the other students in the grade for that year.)

Growth–model tests. Growth models are not a test type by themselves, but rather represent a wide range of approaches to assessing schools, all of which focus on student progress over time. There clearly are characteristics of tests that used to measure growth. An example of a test designed to measure and report the growth of all students is an adaptive test such as Northwest Evaluation Association’s Measures of Academic Progress (NWEA MAP). Adaptive tests match the difficulty of the questions to the performance of the student.

Two important characteristics in measuring individual student growth are:

1. Consecutive tests used to measure growth (pre-test and post-test) must measure the same content.
2. Consecutive tests need to be valid and reliable, and they need to have a minimum amount of measurement error. Most measurement error comes from a test that is too difficult or too easy for the student.

“Many educators feel a growth measure is the fairest system of accountability, because it takes into account the well-known fact that some schools have high proportions of low achieving students, and we really need to concern ourselves with whether schools are earning a year’s worth of growth in a year’s time.” – Jim Angermeyr, director of research and evaluation, Bloomington Public Schools

Standardized tests cost the state of Minnesota millions of dollars every year. It is important to ensure the those dollars contribute to the future growth and success of our students. Schools need to be very thoughtful about what they test, how they test, and that the testing used provides information about how our schools can help each individual grow to their greatest potential. “The perfect test that everyone wants is one that is useful, delivers results quickly, and gets a lot of student achievement information in as few questions as possible.” – Dirk Mattson, director of assessment and testing, Minnesota Department of Education.

Most educators agree that standardized tests are a useful part of a comprehensive student and school assessment system, but they are concerned that many current tests do not assess or promote a variety of important aspects of teaching and learning. “When all your time is directed to test preparation, the students lose out on problem-solving skills, creativity, multiple methods for learning, outside resources and the desire to be a life-long learner.” – Minnesota 2020 Principal Survey, as quoted by the ECM Editorial Board.

This information was taken from Minnesota 2020, the ECM Editorial Board and the Minneapolis StarTribune, Jennings and Corcoran,Phi Delta Kappan, May 2009, and with the assistance of Jim Angermeyr, director of research and evaluation for Bloomington Public Schools.