Issues in Early Childhood Education Assessments

Neither authentic evaluation methods nor standardized tests are perfect. The field has a long way to go in increasing the reliability and validity of authentic assessment techniques. The objectivity of authentic assessment is also a problem. No matter how objective teachers attempt to be in selecting work samples to place in a portfolio, their prior experiences, values, and attitudes continue to influence their judgment. Decisions about what behaviors to observe, what the work in the portfolio actually means, or what questions to ask in an interview are not considered fully valid because they are affected by the teacher’s personal belief systems, attitudes, and values. Whether the samples of work or observations really do measure or assess what the teacher thinks they do can be questioned.

Sabrina’s parents were told by the teacher that Sabrina would not be ready for first grade. The teacher said that she had observed Sabrina, who never sat still, never listened, and talked all the time. Further, she said, she asked her to perform several tasks, and Sabrina failed each. She did not know the ABCs when asked to recite them and failed to correctly complete assigned worksheets. Her parents, thinking Sabrina was an especially verbal child, sought the help of the school counselor, who gave Sabrina the Peabody Picture Vocabulary Test (Dunn & Dunn, 2000). Sabrina’s score indicated that she had the receptive vocabulary of a typical seven-year-old rather than a five-year-old. Sabrina did progress to the first grade, where she excelled in reading.

Authentic assessment techniques, in addition to being subjective, lack reliability. They often fail to reproduce the same results if used again with the same child. What a teacher observes one day may not be present the next. Moreover, the findings obtained from authentic assessment techniques conducted on individual children cannot be generalized to other children. This means that the results or findings cannot be applied to the total group of children, nor can comparisons between children or groups of children be made.

Standardized tests are considered another good source of information that can be used diagnostically or to provide insights into children’s development and learning. Although standardized tests purport to be objective, reliable (producing the same score each time given to a specific child), and valid (measuring what they purport to measure), they too are influenced by the culture in which they were constructed and by their administration. Because every test is developed within some culture, no test is culture free. Each reflects the values and attitudes of the culture. Tests that purport to be culture fair, like the Goodenough-Harris Draw a Person Test (Goodenough & Harris, 1969), have been developed in multiple cultures and require no or little language for administration.

Teachers’ administration of tests affects their validity and reliability. Teachers interpret test items—“Look at the picture again, you know, it’s the fluffy dog”—or fail to follow standardized procedures because of a lack of training and the failure of school systems to monitor consistent procedures. Children’s performance on standardized tests depends on their background of experiences. Designers of standardized tests recognize this and have tried to base the selection of test items on experiences common to the group for whom the test will be used. But sometimes this view is narrow and limited and, many critics claim, based on an urban middle-class American culture. It is also true that standardized tests of readiness, achievement, and intelligence, reflecting middle-class culture, are highly verbal and abstract and emphasize speed, competition, and doing “one’s best.”

Standardized tests, unlike authentic assessment (which is ongoing and related to the curriculum), provide information only about what a given child can do at a given moment in time. Just as most of us can recall a “bad hair day,” each of us remembers taking a test when we were not feeling our best, were upset about a personal problem, or were just tired. Some college professors, recognizing this, permit students to remove one test score, grade, or an assignment by replacing it with some other task as a means of accommodating these bad test days.

Standardized tests also are limited in informing teachers about individual children’s needs. They give teachers a score on a test and a percentile ranking that tells them only how this individual child stands in relation to the children in the norming group. This is not very useful in helping teachers plan to meet the needs of an individual child.

Thus, no one form of assessment can stand alone in making decisions about a child’s learning, strengths, or weaknesses (Jones, 2003). No one standardized test score or one set of observations offers sufficient information for teachers, their principals, or the school system to determine the following:

• Children’s access to kindergarten
• Placement in remedial, developmental, transitional, or special classes
• Retention in kindergarten
• Placement in a homogeneous group on the basis of a single test score or any other form of assessment

Denying children access to kindergarten or placing them in developmental or other transitional classrooms before or following kindergarten to ready them for the next grade placement (Bredekamp & Shepard, 1989; Shepard, 2000) has a major impact on children’s lives and can never be made on the basis of any one assessment or evaluation alone (Bredekamp & Copple, 1997, p. 133).

For this reason and for general ethical concerns, parents must be informed before their children are assessed or evaluated in any way. When authentic evaluation and assessment is linked to the curriculum, informed consent is not necessary. Because they are tied to the curriculum, the findings from authentic evaluation are usually shared with parents. However, if the findings of authentic evaluation are to be placed in children’s permanent records or used in making life-altering decisions, then the parents should be informed.

Claiming that standardized tests and other assessments are necessary tools of accountability, school systems usually do not require informed consent in order to use the findings of assessments. Regardless, parents must be informed of the type of evaluation or assessment, notified when it is to take place, and given copies of the test results.

Most school systems accommodate the individual needs of young children, ensuring that all children succeed at their own pace. Thus, it is rare that placement decisions will take place. Even so, there are special cases when decisions must be made about children’s grade placement. When this occurs, multiple sources of information must be considered. This includes, but is not limited to, information obtained from observations by teachers, standardized test results, school specialists, other authorities, and most of all children’s parents and other family members.