Sometimes, users may see that tests have different standard scores/percentile ranks, but the same age/grade equivalent.
This can occur, due to how each of these score types are computed.
Age/Grade Equivalents only take into consideration the average W score of your examinee's peers in the norming sample.
For example, if an examinee has a grade equivalent of 10.0 on Test 1: Oral Vocabulary, this means that the W score your examinee obtained is the same as the average W score for 10th graders in the norming sample.
In comparison, Standard Scores and Percentile Ranks are computed using a transformation of the W Difference Score.
The W Difference Score is computed as follows (Your examinee's W score - Their same age/grade peer's average W score). This score is then transformed statistically to generate standard scores and percentile ranks.
As you can see from the above, Age/Grade Equivalents are computed differently when compared to Standard Scores and Percentile Ranks.
This difference in computation is why we cannot compare age/grade equivalents and standard scores/percentile ranks directly to one another.
The Fifth Chapter of the WJ IV™ Examiner's Manual provides extensive information regarding the different score types. It notes that scores at a different level (Age/Grade Equivalents are at Level 2 of the WJ IV's score hierarchy; SS/PRs are at Level 4) cannot be used interchangeably, as scores at each level report different information about an individual's performance.
Below is further information regarding four of the common score types, with interpretative information:
Age & Grade Equivalents
During the norming process, items serve as a means of helping us estimate an examinee’s location on the W-scale “ruler.”
When we administer the test to a large representative group of people (norming), we obtain information about the average ability of people at different ages. We can then use that information to create norms.
All tests are centered on a W score value of 500 (approximates average performance of a 10 y/o).
The AE and GE are based on the median W-score. Note that when hand scoring on your test records AE’s and GE’s can only be approximated as estimates. For accurate scoring, you must run a report via Riverside Score.
Age Equivalent Example
If the median W-score for Letter Word Identification for 11 year 0-month-old examinees is 600, then an examinee who receives a W score of 600 on that test would receive 11:0 as their age equivalent score.
Grade Equivalent Example
If the median W-score on Applied Problems for students in the sixth month of fifth grade is 500, then an examinee who receives a W score of 500 on that test would receive 5.6 as their grade equivalent score.
Considerations for Age and Grade Equivalents
•They are often misinterpreted.
•They do not reflect the ability of an examinee.
•For example, if a fourth grader earns a grade equivalent of 8.5 on a test meant for fourth graders, it does not mean that the student can perform at a mid-eighth-grade level. For example, the equivalent of 8.5 does not mean that student would be able to achieve under a curriculum designed for eighth graders.
•However, you can say that the student answered a high number of items correctly when compared to others in their grade; the same percentage of items that an average eighth grade student answered correctly when given the same test.
•Reflects their accuracy level rather than their ability to handle higher-grade-level tasks.
Unlike the Age and Grade Equivalents, which are based on median W-scores, the Standard Score is based on a transformation of the W-Difference Score. The W-Difference Score is the difference between an examinee's test or cluster W Score, and the average test or cluster score for the reference group you choose to compare to the examinee to (same-age or same-grade peers). The standard score follows a normal curve (bell curve), and is influenced by the normative distribution of scores in the norming sample.
Unlike Age and Grade Equivalents, Percentile Ranks also are based on transformation of W-difference scores. They describe performance on a scale from 1-99 in relation to the performance of same-age or same-grade peers in a norming sample.
Percentile Rank Example-
An examinee with a percentile rank of 80% would suggest their tested performance is better than or equal to 80% of individuals in the reference group. Note that the percentile rank does not provide an indication of proficiency or accuracy in a given domain. For proficiency information, users may wish to review the Relative Proficiency Index Score.
Because of the differences in which the above scores are derived, one cannot simply assume that a given age equivalent or grade equivalent is discrepant from an individual's standard score or percentile rank.