Appletcollection Vertical Menu java applet, Copyright 2003 GD

• Memory for Names [Goodbye, Jawf, whichever way your sash should go.]
• Separate Picture Vocabulary, Oral Vocabulary, and Verbal Analogies in the Cognitive Battery, now Verbal Comprehension subtests. There is now a 44 item Picture Vocabulary in the ACH Battery.
• Listening Comprehension. Now Oral Comprehension in the ACH Battery.
• Sound Patterns (e.g., "doolooWHEEP DOOloowheep")
• Cross Out (there goes a nice assessment of one possible aspect of ADHD). It was too redundant with Visual Matching and has been replaced by Decision Speed, which may be more valuable.
• Memory for Sentences. It was too redundant with Memory for Words. I will miss the comparison, though, which often highlighted a strength or weakness in oral language when Sentences was much higher or lower than Words. Also, Memory for Sentences is a mixed measure, assessing mostly Gc (LD) in younger kids and Gsm (MS) in older kids and adults – handy for clinicians, but messy for the McGrew, Flanagan, and Ortiz Integrated Cattell-Horn-Carroll Gf-Gc Cross-Battery Approach.
• Visual Closure can still be found in the K-ABC and children's magazines.

• Retrieval Fluency: naming as many things as possible in each of three categories as fast as possible for one minute each. This is similar to several old neuropsychological tests. It is part of Glr.
• Delayed Recall of Visual-Auditory Learning can now be done as little as half-hour after the initial test (although we still like it as the only test on the market we have seen with a recall interval of a week). Delayed scores are obtainable only as z scores.   A negative z indicates that the delayed score was less than predicted and a positive z indicates that the delayed score was higher than predicted from the subject's initial recall score. Scores between -1.00 (PR=16) and +1.00 (PR=84) may be considered "within normal limits." To show z scores in the Score Report, select the Options tab in the software menu and then the Report Options tab. The Additional Score Column allows selection of the type of score to be reported in the last column of the Score Report. Choose "z score" and click OK

Clarification from Dr. McGrew regarding the Delayed Recall Scores on the WJ III®:  

All WJ III® DR test scores are based on a new set of procedures not previously used in the WJ-R DR tests. A person's DR performance is compared to others with (a) the same age or grade, (b) the same delay interval, AND, (c) others who had the same INITIAL score. Part (c) is new. A person's DR performance takes into account their initial level of performance (a feature that is just one more reason for computer scoring).

Conceptually you need to think of the WJ III® DR scores as discrepancy scores...much like ability-achievement discrepancy scores. An "expected" level of performance on a DR test is generated (which is based on their initial score, age or grade, and delay interval). The person's actual DR performance is then compared to the "expected" DR performance. This is much like an ability-achievement discrepancy.

The reporting of the WJ III® DR scores only in the z-metric makes it clear to the user that these are indeed a different form of score. They are a discrepancy score....between the DR performance expected and the DR performance obtained. The z-scores can be interpreted just like the SD scores for the other types of WJ III® discrepancies.

• Decision Speed asks you to scan each row of simple pictures and circle the two that are most similar conceptually, e.g., aardvark and anthill or fork and spoon. Most of the decisions are easy – it is a speed test.
• General Information has two subtests. Where asks, e.g.: "Where would you find a WISC kit?" "Where would you find a crayon?" What asks, e.g., "What would you do with a WISC kit?" "What would you do with a pencil?" No score can be obtained for the separate parts, only for the total.
• Planning asks you to trace all the lines in each figure, starting wherever you mark an 8 , without skipping any lines and without retracing any of your previous tracing. The test is not timed, which makes it pretty easy for anyone with good attention and adequate Gv.  The easel provides a color-coded scoring guide. Each color constitutes a complete segment. The colors do NOT specify an order, only which lines are part of the same segment. It is essential that evaluators refer to the guide since some segments may include only 1 line (worth 1 point) while another segment may include 2 or 3 lines yet be equal to only 1 point. Subjects may trace the designs with their fingers but ONLY if this is the subject's idea. The examiner must not give any tips or guidance about strategies to use in solving the planning items. 
• Pair Cancellation shows rows of many little pictures. You are told to circle each instance of a certain picture followed by a certain other picture and only those instances. Score is the number of correctly circled pairs within the time limit. Popular on neuropsychological batteries.
• Auditory Working Memory asks you to repeat dictated words and digits (e.g., cow 7 5 up 9 tree 3) with the words first and the digits next, both in the order they were dictated. Partial credit is given for either the words or the digits in correct order as long as the words are attempted first.
• Auditory Attention. As Kevin McGrew says, the 1973 Goldman-Fristoe-Woodcock Auditory Skills Test Battery was ahead of its time. This is the test of recognizing spoken words (e.g., "Point to 'map.'") on four-picture, multiple-choice pages with interference from increasingly loud background noise (a loud cafeteria tape recorded and played backwards). As on the GFW, the student is first taught the names used for the pictures.
  Despite the fact that all tape recorded tests were normed with the child hearing the tape through headphones, some examiners will be tempted to administer the taped tests without this device.  This particular test, requiring the child to distinguish spoken words with "taped" interference noise, seems highly susceptible to other, extraneous noises.  This, if not all taped tests, SHOULD  be given using headphones.
• Rapid Picture Naming is just what you would expect. Another neuropsychological measure.

• Even more explicit than in the WJ-R® is the combination of WJ III® COG and ACH tests to do a complete McGrew, Flanagan, and Ortiz Integrated Cattell-Horn-Carroll (CHC) Gf-Gc Cross-Battery Approach or to explore completely a CHC factor.
• Usage score from Dictation was voted off. Spelling and Punctuation & Capitalization are now two separate tests, replacing Dictation so we need no longer write "writing from Dictation" in our reports.
• Science, Social Studies and Humanities are now one test, Academic Knowledge, also treated as a cluster.
• Reading Fluency, potentially very valuable, has been added, but may be too complex a task for the purpose because of the comprehension and yes/no decision-making demands (Suanne Bickum 10/24/00).
• Math Fluency has been added, completing the new Academic Fluency cluster.
• Story Recall and Story Recall-Delayed for you Babcock and WRAML fans. See note above on Delayed Recall scores.
• Spelling of Sounds. Prof. McGrew is again correct: the Goldman-Fristoe-Woodcock was ahead of its time.
• Understanding Directions At last a test to measure exactly what we expect children to be able to do in school and an improvement over the subtest dropped when the DTLA-2 was revised into the DTLA-3.. 
• Writing Evaluation Scale is a one-page, checklist procedure for evaluating other samples of the student's writing.
• Picture Vocabulary is now part of the Achievement Battery (see above).

Other Changes and Notes

• There is no hand-scoring option. You can look up "estimated" age- and grade-equivalent scores right on the record form, as on WJ-R, but these are just that - "estimates."  The exact equivalents are obtained only by using the scoring software. First-printing protocols contain errors on 6 tests, which are corrected in a separate sheet sent out by Riverside and in later printings.
• The total Cognitive Score (General Intellectual Ability or GIA) is now based on the g loadings of the various tests at each age level, rather than a simple total.
• The WJ III® Cognitive also allows for a Brief Intellectual Ability score (BIA) composed of only 3 tests.
• RPI (Relative Proficiency Index) replaces the RMI (Relative Mastery Index). RPI is a score that predicts a person's degree of proficiency in comparison to age or grade peers. It describes the quality of the performance or functionality. A subject's performance is compared to the point at which average students in the comparison group (either age or grade) would perform similar tasks with 90% proficiency - something similar to the Snellen Eye Chart that produces results comparing the person to the average proficiency of a 20/20 sighted person.  A person with 20/60 eyesight sees at 20 feet what a normally sighted person sees at 60 feet.  An RPI of 60/90 suggests that the person performs a particular task with 60% proficiency while persons from a similar age- or grade-comparison would perform at 90% efficiency.
• CALP [Cognitive-Academic Language Proficiency (Cummins, 1984)] can now be reported.  It, like the Delayed Recall scores, must be specifically chosen from the option menu. Since the computer software allows you to choose only one additional score at a time (e.g., z, T, NCE, CALP), if you wish to see more than one score you must use at least the print preview twice, a task that is very easy to do. 
• Predicted Achievement measures have been refined.
• Interpretive Options, including many kinds of discrepancy, abound and are explained in some detail in the WJ III® Manuals.
• The WJ III® Manuals contain genuinely useful information about testing students with various disabilities and disadvantages.
• Richard Woodcock's Diagnostic Worksheet is essentially unchanged and still useful.  There have been some additions made to the worksheets by Drs. Kevin McGrew and Randy Floyd.  The WJ III® Excel Template has incorporated those additions.
• The "Supplemental" tests are now "Extended," but still essentially diagnostic.
• There are now a lot of tests related to current research on reading, including measures of phonological awareness and Rapid Automatized Naming (RAN). Spelling of Sounds and Selective Attention (cafeteria noise) have been borrowed and updated from the Goldman-Fristoe-Woodcock.
• There are now more post-high-school grade norms.

• The Pronunciation Key does not distinguish between voiced and unvoiced /th/ (the and thin).
• Correct answers in languages other than English are acceptable, which is theoretically a grand idea, but practically complicated, e.g., cheval, caballo, cavalo, cavallo, Pferd, and лóшадь would all be acceptable alternatives to "horse," but you could fool us.
• Scoring oral responses requires, we think, as much clinical judgment as on the WISC, and there is, we think, less guidance. The Manual tells you, in all else fails, to balance 1 and 0 ratings on successive, questionable responses, but we'd still like to see a little more explanation and a few more examples for some items.
• We are, of course, terrified of the exclusive reliance on computer scoring without printed norms tables. We understand and appreciate the excellent reasons for computer scoring, and the scoring sophistication made possible by the computer, but John Henry probably could have recited the virtues of the steam drill. Stay in warm, personal contact with the person who actually orders tests for you, as that is probably the person who will receive revised scoring disks and notifications of scoring-program changes.
  Because of the inability to "check" the scores you obtain in a manual, be sure to look at all computer generated scores carefully.  Computer software has been known to become corrupted.  If you have any questions or concerns, be sure to err on the side of caution. Be sure to select the "raw score" option so you can make certain you have entered all raw scores correctly.
• There are tons of new scoring and interpretation options. We think the WJ III® ACH is the most sophisticated achievement test now on the market. The sophistication of the test requires equal sophistication in the selection of tests, selection of scoring options, and interpretation.  Do not play "interpretation roulette," particularly when it comes to discrepancy procedures.  Do not go "data mining" (running analysis after analysis until you get a result you want). Understand the different procedures used and generated by the software and use the one that best matches your, or your district's, policy, procedures, and understanding and the individual referral questions.
• Basal and ceiling rules are essentially optional. If your clinical judgment (not just your respect for the laws of chance) makes you think there are potential misses below the basal or hits above the ceiling, you are supposed to go for them and take the lowest basal and highest ceiling. This procedure is reminiscent of the much-maligned rule on the K-ABC Faces and Places (remember the Martin Luther King, Jr. item?). With our reality bias, we think this is fundamentally a sound idea, but it will introduce some inter-examiner variability.
• A necessary trade-off for the comprehensiveness of the achievement battery is the brevity of each individual test. The manual recommends interpretation at the cluster, rather than the test level. As diagnosticians, we would still prefer longer tests so we had enough items for more analysis. We think we will need to go, for example, to other, longer word lists to find enough oral reading errors to allow us to determine what's up with those reading errors.
• The powerful prediction procedures will lead the thoughtless to under-identify learning disabilities. The specific achievement aptitude measures practically serve as measures of a "disorder in one or more of the basic psychological processes involved in understanding or in using language, spoken or written, that may manifest itself in an imperfect ability to listen, think, speak, read, write, or to do mathematical calculations." Therefore, the lack of a significant difference between aptitude and achievement not only does NOT rule out a learning disability, it actually increases the chances that there is a learning disability. Please see the excellent explanation by Mather & Schrank (2001). The lack of a significant discrepancy between predicted and actual achievement might point to apedagogia, dyspedagogia, or one of the rule-out circumstances in the federal definition of a specific learning disability [Sec. 300.7(c)(10(ii); Sec. 300.541(b)(2)].
• We still do not think that the easy-difficult range is based on the student's actual range of easy and difficult items, but simply on the student's obtained score. We would prefer to use the grade equivalent for the highest item passed before the first error and the highest item passed on the test to mark off the student's functional range.
• Otherwise, don't use age- and grade-equivalent scores at all. http://alpha.fdu.edu/psychology/oat_cereal.htm.
• Keep posted at www.riverpub.com/support/updates.htm , http://home.att.net/~gfgc/index.htm , and http://alpha.fdu.edu/psychology/

References

Cummins, J. (1984). Bilingualism and special education: Issues in assessment and pedagogy. Austin, TX: Pro-Ed.

Mather, N., & Schrank, F. A. (2001). Use of the WJ III discrepancy procedures for learning disabilities identification and diagnosis (Assessment Service Bulletin No. 3). Itasca, IL: Riverside Publishing.

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

Shrank, F. A., & Woodcock, R. W. (2001). WJ III Compuscore and Profiles program [computer software]. Woodcock-Johnson III. Itasca, IL: Riverside Publishing.

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

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

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

Content on these pages is copyrighted  by Dumont/Willis © (2001) unless otherwise noted.