Alternate Assessment: A substitute approach used in gathering information on the performance and progress of students who do not participate in typical state assessments. Under the reauthorized Individuals with Disabilities Education Act, alternate assessments are to be used to measure the performance of a relatively small population of students who are unable to participate in the regular assessment system, even with accommodations.

Alternative Assessment: A generic term that typically is applied to a variety of different assessment activities. These assessments provide an alternative to multiple-choice tests that require students to select one response. Writing samples, portfolios, and performance-based assessments may all be considered to be forms of alternative assessment.

Assessment: The process of collecting data for the purpose of making decisions about individuals, groups, or systems.

Assessment Literacy: Obtaining knowledge of the basics of assessment, terminology and philosophy.

Authentic Assessment: Often used synonymously with performance assessment, this term can also mean an assessment that only uses real-world tasks as the basis for information about how well an individual can perform certain tasks.

Confidence Interval (CI): A numerical range that shows the interval around a score that one would expect a person or group of persons to obtain if they were to take the same test again. A CI of 95% indicates that one can be 95% confident that if the person or group was retested their average score would fall into the same range.

Criterion-referenced test (CRT): Criterion-referenced tests are measures used to examine student performance relative to state and/or district criteria or standards. Instead of comparing students' scores to a national normative standard, scores are interpreted in terms of various performance standards, usually at the district or state level (e.g. mastery vs non-mastery; low proficiency, moderate proficiency, and high proficiency within a particular subject area).

Norm-referenced Test (NRT): Norm-referenced tests are those that provide a comparison of individual performance to that of a state or national comparison (standardization) sample. A norm-referenced test measures the performance of a student against the performance of other individuals. Use of the norming sample allows for raw scores to be converted to grade-equivalent scores, percentile scores, and standard scores.

Normal Curve Equivalents (NCEs): Normal Curve Equivalents are standard scores that are generated from a normal distribution. One begins with a set of raw scores, converts them to percentile ranks, and then, using a z-score table, converts the percentile rank to a z-score. The resulting z-scores are usually transformed using a linear transformation to a new scale, (e.g. such as the SAT, where the mean equals 500 and the standard deviation equals 100.) Normal Curve Equivalent scores can only be used for students who are similar in age or grade to those in the norming sample.

Percentile Scores: These scores tell the percent of people in the normative sample that scored at or below a student's score (e.g., a percentile rank of 80 means that 80% of the normative group earned a score at or below that student's score).

Performance Assessment: A form of testing that requires the creation of an answer or a project, rather than the selection of an answer as in many traditional, multiple-choice tests. In many cases, such assessments are intended to represent or simulate real-life situations that require problem-solving. The term often is used synonymously with authentic assessment.

Portfolio Assessment: A collection of student-generated or student-focused products that provide the basis for judging student accomplishment. In school settings, portfolios may contain extended projects, drafts of student work, teacher comments and evaluations, assessment results, and self-evaluations. The products typically depict the range of skills the student has, or reveal the improvement in a student's skill level over time.

Raw Scores: These are simply the scores obtained when one sums the score on each item. If items are scored dichotomously (1 or 0), then a raw score represents the total number of items answered correctly.

Reliability: Reliability is the extent to which a test measures what it purports to measure time after time. It is the accuracy, precision, or stability of a measuring instrument.

Rubric: A scoring guide that facilitates rater consensus on students' performance of assessment tasks. A rubric provides criteria from which those assessed can learn to improve their performance.

Standard Error of Measurement (SEM) : An index of reliability that essentially converts reliability data from a test into a confidence interval around a given score. Knowing the standard deviation, the reliability, and a person's score, one can estimate a confidence band within which you would expect the individual to score (in typical cases, 95% of the time) if that individual repeatedly took a parallel version of the test.

Standard Scores: These scores are linear transformations of raw scores, and are considered easiest to interpret. With standard scores, the mean and standard deviation of any distribution can be placed onto a similar scale. Common examples of standard scores are the SAT with a mean of 500, and a standard deviation of 100, or a typical IQ test with a mean of 100 and a standard deviation of 15.

Standards-based Assessment: An assessment instrument, battery or system that has been constructed to measure the achievement of individual students or student populations in attaining certain standards, generally established by local districts or state educational agencies. Most state-level standards-based assessment programs currently in place measure student performance against articulated standards in core academic content areas, such as reading, mathematics, writing, science and social studies.

Test Matrix: Nearly every test developed has a test matrix that describes the types of items and nature of content included in the test. Often a test matrix is provided in the test's administration manual.

Validity: Validity means truth or correctness. It is the correspondence to a proposition describing how things work and how things actually work. Test validity refers to the degree to which a test actually measures what it intends to (or is supposed to) measure.