Standardized Tests

Standardized tests can provide a wealth of information relevant to the field of life-span human development. This is accomplished by comparing measures of behavior or status (e.g., academic aptitude and achievement or height and weight) to population averages to make current judgments relative to a norm or reference group. This entry describes two ways in which tests can be standardized, provides examples of how standardized tests are used in studying life-span human development, discusses reliability and validity of standardized tests, and covers limitations and potential biases of standardized tests.

[Page 2134]The term standardized is actually used in two related ways. First, tests can be standardized according to their administration and scoring procedures, and second, tests can be standardized according to the normative populations against which results are compared. For example, when assessing a cognitive aptitude such as working memory, an individual may be orally presented with a particular list of digits and letters in a predetermined order. The examinee subsequently must orally repeat the letters first, in alphabetical order, followed by the digits in numerical order. In this case, the original order of letters and digits is standardized so that the order remains invariant across examinees, and the way in which the response is evaluated or scored is exactly the same for every examinee. This kind of standardization ensures that all examinees are treated in exactly the same way and therefore permits comparisons among them on the construct being assessed. In this sense, the standardized test is like an experimental task where surrounding conditions are held constant for all examinees.

The other way in which tests can be standardized refers to the populations against which results are compared, that is, the normative sample. Here, a relatively large group of individuals are administered the test under highly controlled conditions, and their results are statistically summarized to permit the comparison of an individual to the group. Thus, the measure obtained for the individual can be compared to the descriptive statistics for the group, and a judgment can be made about the individual’s relative standing. That is, an individual’s measure might be above or below the arithmetic mean for the normative sample by a specific amount. The norms of the standardization sample usually are presented in percentiles, which refer to the percentage of individuals who fall below a particular value or in standard-score units with a specified mean (e.g., 100) and a specified standard deviation (e.g., 15). Usually, normative samples are stratified according to demographic variables that permit the test user to determine whether the normative sample is representative of the individual examinee. Typical stratification variables often include gender, age, ethnicity, and geographic region, among others. Of course, it is incumbent on the test user to ensure that the examinee matches the characteristics of the sample. For example, very different normative groups would be required to assess the working-memory skills of a 70-year-old woman suspected of having Alzheimer’s disease versus a boy in the third grade who is experiencing academic problems.

Standardized tests are used in virtually every sphere of the field of life-span human development, including physical, cognitive, and affective domains. Among the most popular are group administered academic achievement tests that are routinely administered in public and private schools. Other standardized tests are individually administered and used not only in educational settings from preschool through college years but also in clinical (e.g., medical, psychological) and vocational settings. An almost overwhelming list of group and individually administered standardized tests is available in the resource, Mental Measurements Yearbook With Tests in Print, which is available through most academic libraries. Here, tests and associated critiques can be accessed by their name, author, category, acronym, population, and purpose, among other characteristics.

...