In this approach, each item is classified according to the process, concept, or skill the subtest assesses. These classifications constitute error categories. For example, items on Reading Comprehension and Listening Comprehension are classified as measuring either literal or inferential comprehension and either narrative or expository passage comprehension. Written Expression error analysis is referred to as item-level, even though there are several error categories associated with each item. The error classification for Phonological Processing is based on item (or section) type rather than skill.
On the Letter & Word Recognition, Nonsense Word Decoding, Spelling, and Oral Expression subtests, error analysis requires qualitative analysis (within-item classification) of the response to determine which error categories are involved. The Math Computation subtest includes two sets of error categories: one requires within-item error classification and the other allows errors to be classified automatically at the item level. For example, a Math Computation error on an item involving whole-number subtraction with regrouping would automatically be counted as an error in the Subtraction category (within-item), depending on the specific reasons for the student's incorrect response.
Development of Error Analysis Norms
The error analysis methodology in the KTEA-3 resembles other criterion-referenced assessment procedures used to provide specific information on the acquisition of selected skills; however, key differences set it apart. Traditionally, criterion-related tests indicated mastery with a specific cut-off score. Grade appropriate content and cut-off scores for criterion-referenced tests are usually based on the judgments of curriculum experts. Instead of using arbitrary cut-off scores to represent mastery of a skill, the KTEA-3 compares an examinee's total errors in a category with the average number of errors made by the reference group to indicate whether an examinee's performance is above average, average, or below average.
The entire KTEA-3 grade norm sample (n = 2600) was used to develop the error norms for the six subtests with item-level error analysis. For the subtests that use within-item error analysis, a subset of the grade norm sample was included in a stratified error analysis norm sample to establish cut-score norms.
Error analysis is available on Q-global or for hand-scoring
Using Q-global®, each of the 10 subtests that offer error analysis is listed in the Error Analysis tab, if you enter raw scores for those subtests. Refer to the Error Analysis Summary section in the score report for a summary of the total errors made, the normative comparisons (average number of errors), and the descriptive categorization (weakness, average, strength). For hand scoring, the flash drive in the KTEA-3 kit offers Error Analysis Worksheets and the Error Analysis Summary, plus the Technical & Interpretive Manual.