The Woodcock-Johnson III Tests of Achievement

The Woodcock-Johnson III Tests of Achievement (WJ III ACH) contains a set of norm-referenced tests that are administered individually. The WJ III ACH is used to measure academic achievement. It is composed of 22 tests that make up five clusters: reading, oral language, math, written language, and academic knowledge. Since the tests are very comprehensive in terms of the material that they cover, the WJ III can be used for educational, clinical, or research purposes (Mather, 2001).

The WJ III ACH can be administered to almost anyone. Its items range from preschool to above average adult level. The norming sample ranges in age from 2 to over 90 years old (Mather, 2001). Even though the assessment can be used on a wide variety of ages, its primary application is for the school-aged population (McGraw, 2001).

The WJ III ACH can be used to determine a person's current academic strengths and areas that need improvement. It takes approximately 60 to 70 minutes to administer the entire achievement test since each subtest takes about five minutes. Age and grade rank profiles allow educators to interpret whether or not the student is performing consistently with others their age. The assessment can be used in schools to help determine where students should be placed and if any accommodations would be helpful to them. Individual Education Plans can be designed partially from the information received from the WJ III ACH. Since the WJ III ACH can be used on such a wide age range, the results can be used to track growth throughout a person's entire lifetime (Mather, 2001). The WJ III ACH was normed on the same sample as the Woodcock-Johnson III Tests of Cognitive Abilities so the scores between the two can be directly compared with a higher degree of accuracy than would be possible if the two tests were normed on different groups. This allows examiners to be able to evaluate achievement skills jointly with their related cognitive abilities (McGrew, 2001).

The WJ III Tests of Achievement can also be used to determine if students are eligible for special services as defined by the Individuals with Disabilities Education Act (IDEA). IDEA considers a learning disability to be a "discrepancy between intellectual ability and achievement in one or more areas," which makes the WJ III ACH an ideal assessment for this purpose (Nelson Education, 2009). The WJ III ACH has cluster areas such as oral expression, listening comprehension, written expression, and basic reading skills that parallel the areas in IDEA with the same names (Nelson Education, 2009).

A large, nationally representative sample of 8,818 people was used to collect data for the WJ III ACH norms. Subjects came from over 100 geographically diverse communities in the United States. A stratified sampling design was used to control for 10 variables: census region, community size, sex, race, Hispanic, type of school, type of college, education of adults, occupational status of adults, and occupation of adults in the labor force. There were 1,143 subjects in the preschool sample. 4,783 subjects made up the kindergarten through 12^th grade sample. The college sample had 1,165 participants and 1,843 adults were tested for the adult sample. The majority of the subjects were tested during their school aged years because that is when their ability in the tested subjects is changing at the fastest rate (McGraw, 2001).

The test-makers used three stages of sampling to make sure they obtained a representative sample. The three stages were: sampling of communities, sampling of schools, and sampling of subjects. The second stage was only used when sampling school-age subjects. In the sampling of communities, they sampled from the nine census divisions of the country. Communities were chosen based on geographic distribution, size of community, and SES characteristics. Schools were chosen to be sampled if they provided a representative sample of their community. Home schooled students and private schools were included as well. A quota-by-grade-level criterion was used to sample school subjects. Students were chosen at random from the schools to be tested (McGraw, 2001).

The WJ III ACH meets standards for reliability. On the standard battery, the median reliability coefficient alphas range from .81 to .94. The median coefficients on the extended battery range from .76 to .91. These high reliabilities mean that the assessment can be used for decision making purposes.

Several techniques were used to estimate the reliability of the WJ III. The reliabilities for most of the tests were calculated using the split-half procedure which split the tests into two by odd and even numbers. The correlations between to the two halves of the test are used to derive a split-half reliability coefficient. Test-Retest reliabilities were used for the Speeded Tests. The retest interval was one day and the tests were administered in a counter-balanced order to 165 subjects. The reliabilities here ranged from .70 to .96. The Writing Samples, Writing Fluency, and Handwriting tests require subjective evaluation so interrater reliability studies were used where six raters independently scored the same set of tests. The median intercorrelation among the raters was .93. These results show that trained examiners working alone can reliably score these tests (Mather, 2001).

According to the Technical Manual, the scores produced by the test have validity. The content in the WJ III ACH is similar to the material taught in schools and other achievement tests. The Manual gives validity evidence based on test content, developmental patterns of scores, internal structure, and relationships to other external variables. The test content was developed partially by outside experts who made up test items in their fields. Items were reviewed for bias and sensitivity to women, individuals with disabilities, and cultural or linguistic minorities. Sensitive items were modified or removed. All of the tests and clusters show average score changes that follow with the developmental growth and decline of achievement abilities across the ages for which the test is used. The WJ III ACH also has internal structure validity. It correlates highly with other tests that are measuring the same abilities. For example, it has a correlation of .65 with the Wechsler Individual Achievement Test and .79 with the Kaufman Test of Education Achievement. Overall, the WJ III ACH shows validity because its test scores can be used for their proposed purpose (Mather, 2001).

References

Mather, N.. & Woodcock, R.W. (2001). Examiner's manual. Woodcock-Johnson III Tests of Achievement. Itasca, IL: Riverside Publishing.

McGrew, K.S., & Woodcock, R. W. (2001). Technical Manual. Woodcock-Johnson III. Itasca, IL: Riverside Publishing.

Nelson Education (2009, January 2). WJ III Tests of Achievement, Form B, with Normative Update . Retrieved October 12, 2009, from Canada's Learning Advantage: http://www.assess.nelson.com/test-ind/wj-3-ach.html