Download CELF-5 Test Features and Interpretation Q & A and more Study notes Communication in PDF only on Docsity!
CELF‐ 5 WEBINAR ON NOVEMBER 14, 2017
QUESTIONS AND ANSWERS
Standard Score Ranges:
Q: Is a standard score of 86 considered within 1 SD of the mean or is 85 within 1 SD of the mean?
A: A standard score of 85 is at the lower limit of the normal range (‐ 1SD).
Q: Would a student with a standard score of 85 be considered to have a language impairment?
A: We would consider this student ‘at risk,’ but standards for service may be different from state to state.
We also recommend in CELF‐5 that you should not make a decision about a student based on a single score. If you use the 90% confidence interval around the single score, you can indicate the range of the student’s potentially best scores in the range above – as ‘in a best case scenario’ ‐‐and the range of the potentially worst scores in the range below – as ‘in a worst case scenario.’
Q: If a student’s standard scores fall 2 SDs below the mean, we should not use the CELF results?
A: No, that’s not what I meant to say. Any standard score at or below 2 SDs of the mean indicates a severe disorder. Item error analysis is recommended and you should also take a look at any evidence of intra‐personal strengths.
Q: In Indiana Psych uses cross battery, but SLPs test differently and work with students below 85. If I understand we should service 77 and below?
A: Good for students in Indiana. As I stated earlier, you should follow State guidelines in determining who is eligible to receive services.
Q: When the Index Score is “rare and unusual” how should that be interpreted” How does that affect the student? It is meaningful, but what does it mean?
A: I talked about the case, when the Receptive v. Expressive or the Content v. Structure Index Scores differ in an amount that occurs rarely (e.g., in less than 5% in the standardization sample) and is therefore considered “rare and unusual.” When this degree of discrepancy occurs between Index Scores, you can identify the nature of the relatively more severe aspects of the disorder.
ORS – Observational Rating Scale:
Q: What is ORS? What does ORS refer to?
A: ORS is the acronym for the Observational Rating Scale. This scale features 40 questions that cover the areas of (a) Listening – 9 items, (b) Speaking ‐ 18 items, (c) Reading – 6 items, and (d) Writing. Each items contains a statement that describes communication problems that are often shared by students with language disorders. The answer to each I expressed as the percentage of time this problem occurs (e.g., Never; Almost never; Sometimes; Often).
Q: Would it be possible to discuss CELF ‐ 5 growth scores in more depth?
Growth Scale Values provide an objective score for measuring changes in CELF– performance over time. Growth Scale Values were developed using the performance of all examinees included in the normative sample. The Growth Scale Value is an IRT‐based ability score with an equal‐interval scale that can be used to compare changes in an individual’s score across multiple administrations. It is a transformation of the raw score and is superior to raw scores for making comparisons for clinical evaluation, in that raw score totals do not account for differences in item difficulty. A student could have gotten three more items correct, but those three items could have been easy items or hard items. Increases in Growth Scale Values are adjusted such that an increase of 3 points represents the same amount of progress anywhere on the growth scale for the respective test. Growth Scale Values corresponding to test raw score totals are presented in Appendix G.
You can use Growth Scale Values to quantify small improvements in the language skills of a student with a moderate to severe language impairment. CELF–5 provides Growth Scale Values for the tests for which a test‐age equivalent can be derived. You can use Growth Scale Values to:
Track a student’s skill development on specific tests (e.g., Recalling Sentences, Linguistic Concepts, Word Classes)
Determine if the student has gained additional language skills since a previous administration of CELF–
Measure the efficacy of an intervention protocol that has been implemented for the student
The advantage of using Growth Scale Values rather than scaled scores to assess improvement in a student’s language ability is that the Growth Scale Value provides a quantifiable measure of a student’s changes in ability, even if the amount of change is not sufficient to narrow the gap between the student’s language skills and those of same‐age peers. That is, Growth Scale Values provide an estimate of language ability based on the range of performance of the entire normative sample rather than that
The items for which a student can guess the correct answer tend to be receptive language tests that provide multiple‐choice response items (e.g., Sentence Comprehension, Semantic Relationships). If there is a question of Growth Value Score increases coming primarily from guessing, compare scores on related tests that are not multiple choice to determine if the improved scores also occurred there. Increases in Growth Score Values that only occur on multiple choice tests suggest chance contributions instead of increases in skills or abilities.
Scores Stay About the Same Possible reasons a student’s Growth Scale Value changes very little, include:
The student may have been tested again, before he or she mastered additional language skills. That is, a younger student (ages 3–5) may reasonably be tested every six months because children develop many language skills quickly at this age. However, a student older than age 5 may or may not be expected to show a difference in skills in six months’ time. It is not recommended that CELF–5 be administered if less than six months have elapsed, unless you have reason to believe that the student has made measurable progress in that time. Alternative forms of assessment (e.g., language sampling, dynamic assessment) are preferable to frequent additional administrations of the CELF–5 tests. Frequently repeated administrations of most tests may result in inflated scores due to practice effects. See Chapter 3 for a discussion of retesting time considerations. At some points along the language developmental continuum, students who are developing language typically plateau for certain types of language skills. Acquisition of morphological markers, for example, does not necessarily proceed at a continuous pace. When this is the case, you may not see progress for certain types of skills. Students who have language disorders, like their typically‐developing peers, may experience plateaus in their language development for certain types of language skills. The student may not have been in therapy for a sufficient length of time for change to occur as a result of intervention (e.g., the student is only in the fourth week of an eight‐week intensive language therapy program).
Scores Decrease
Possible reasons that Growth Scale Values may decrease from previous testing, include:
Growth Scale Values may decrease for a student who is sick, tired, frustrated, or distracted during the second test session. In this situation, the student is not demonstrating a best performance, and it would be erroneous to interpret a lower score as evidence that the student is losing language skills. In some cases, a student may have a progressive or degenerative condition in which the student loses previously acquired language skills. A student who has suffered a traumatic event (e.g., head injury) or illness (e.g., meningitis or sudden onset of a seizure disorder) may also lose previously acquired language skills.
Q: When are GSV scores considered significant: In other words how many GSV score points should a child improve/’grow’ to indicate significant growth/change?
As noted above, GSVs are not normative scores because they do not involve comparison with a norm group. Standard scores, percentile ranks, stanines, and normal curve equivalents (NCEs) compare a student’s performance with that of a reference group representing others of the same age (the normative sample). In contrast, the Growth Scale Value measures a student’s skills with respect to an absolute scale. Because that is the case, there is not a calculation that is done to determine if a score is statistically significant.
As the student’s skill level grows, the Growth Scale Value will increase. You will want to review the Item Analysis to see if the increase in test scores represents acquisition of a new skill (e.g., the student now consistently demonstrates understanding or use of a linguistic form that was not in his or her repertoire before) or if the increase may simply be due to chance (e.g., earning one or two more raw score points by randomly pointing to a correct answer in a multiple choice task).
Q: Where on the Record Sheet (front page) do you record the growth scores so you can easily compare between administrations?
Record the Growth Scale Value in the appropriate column on page 1 of either Record Form 1 or 2. Growth Scale Values may also be recorded on the reproducible form at the end of this chapter. This form enables comparison of Growth Scale Values across multiple administrations of CELF–5. Growth Scale Values are not available for Understanding Spoken Paragraphs, Reading Comprehension, or Structured Writing
Recalling Sentences:
Q: Recalling Sentences is considered an expressive subtest, yet it requires receptive ability to process and recall prior to repeating. How is this only considered an expressive measure?
A: Yes! We, Secord and Wiig, totally agree. However, many school systems require a differentiation between Receptive and Expressive Skills, the only reason we feature it. You should have noticed that I stated ‘primarily expressive” or ‘primarily receptive’ as I talked about the receptive‐expressive distinction.
the student’s grade level and develop comprehension for all terms. For spatial terms, a procedure in which you use three or more tokens or pictures of entities and ask the student to place or manipulate the items has worked for me in the past. For temporal terms, I have used a clock face, class/bus/train schedules as references for interpretation.
Specific Limitations of the CELF‐5:
Q: What do you consider to be specific limitations of the CELF ‐ 5?
A: One limitation covers any and all formal tests. It is that no test can do everything you might want it to do. That is the most obvious reason for requiring us of more than one test to arrive at a diagnosis and also for using cross‐battery testing.
Within these limitations, we have done everything to provide a broad battery of tests with flexibility in using the tests and with clearly excellent statistical characteristics.
With every issue of a CELF test version we have tried to respond to the most recent trends in our field. To further extend the range of the CELF family of tests, we have added the CELF‐5 Metalinguistic battery. I have visions of other extensions in future stand‐alone tests or as a part of future CELF batteries.
Elisabeth Wiig. November 2017
