• Tidak ada hasil yang ditemukan

Is the Verbal Versus Nonverbal SS Difference Unusually Large?

KAUFMAN BRIEF INTELLIGENCE TEST–SECOND EDITION (KBIT- 2)

Step 3: Is the Verbal Versus Nonverbal SS Difference Unusually Large?

Verbal vs.

Nonverbal SS Difference

Size of Difference Needed of Abnormality

Does size meet unusually

large criteria?

15 24 No

Step 3 Decision Box If an unusually large

difference is found

then this unusual discrepancy should be interpreted.

Explain the unusually large Verbal and Nonverbal difference.

If no unusually large difference is found

then you must determine if noted difference is interpretable.

Explain the noted difference.

Rapid Reference 2.13

Examiners are cautioned against interpreting implications of a meaning ful Verbal and Nonverbal KBIT- 2 difference, and should recommend com- prehensive testing.

Determining Scaled Scores for the Verbal Knowledge and Riddles Subtests

Kaufman and Kaufman (2004b) did not design the two KBIT- 2 verbal subtests, Verbal Knowledge and Riddles, to measure distinct verbal abili- ties. However, there may be circumstances in which it is useful to compare performance on the two measures, or situations may arise when one or the other subtests cannot be administered. For these purposes, Table B.8 in the KBIT- 2 Manual (Kaufman and Kaufman, 2004b) provides norms for these two subtests using scaled scores (mean of 10, standard deviation of 3). While a scaled score on one subtest cannot be used for obtaining a Verbal score or IQ Composite standard score, the purpose of scaled scores is to provide useful normative information about an individual’s abilities.

When comparing Verbal Knowledge and Riddles scaled scores, a differ- ence of three points or greater is signifi cant at the .05 level.

STRENGTHS AND WEAKNESSES OF THE TEST

In the next section, we highlight major strengths and weaknesses of the KBIT- 2 in the areas of test development, administration and scoring, re- liability and validity, standardization, and interpretation. Information is provided using the Rapid Reference format to provide information in a user- friendly format. (See Rapid Reference 2.14.)

CLINICAL APPLICATIONS OF THE TEST

In order to provide evidence of the clinical sensitivity of the KBIT- 2, data were collected with various special populations including individuals with learning disabilities, speech / language disability, Attention- Defi cit /

Strengths and Weaknesses of KBIT- 2

Strengths Weaknesses

KBIT- 2 can be used across a wide range of ages 4–90.

Normative IQs range from 40–

160 for most ages and provide a high ceiling for children and low fl oor for adolescents and adults.

Test authors provide theoretical explanation for using lower start points for individuals ages 76–90 on the Matrices subtest.

KBIT- 2 is normed such that if the examinee fails the second trial of a teaching item on Riddles and Matrices subtests, examiner may use own words to restate directions or de- scribe the task more generally using additional examples. It is also permissible to use a dif- ferent language for teaching.

Care was taken to drop poten- tially offensive or biased items from Verbal Knowledge sub- test.

Subtests have many common rules that facilitate easy ad- ministration. For example, basal and ceiling rules are con- sistent across subtests. Simi- larly, all items are worth 0 or 1 point.

Rapid Reference 2.14

Because a one- word response is required for Riddles sub- tests, limited information is obtained regarding an exam- inee’s expressive vocabulary abilities.

Split- half reliability is not ex- cellent for the Nonverbal score when assessing young children ages 4 and 5 (mean = .78).

KBIT- 2 IQ Composite corre- lates .76 and .77 with the WISC- III and WISC- IV Full Scale IQs.

KBIT- 2 Nonverbal score cor- relations with WISC- III PIQ (.53), WISC- IV PRI (.56), and WAIS- III (.79) are low.

While verbal subtests assess general information, receptive vocabulary, and verbal rea- soning, the ability to measure expressive vocabulary is lim- ited.

The KBIT- 2 samples only lim- ited areas of cognitive func- tioning and does not measure processing speed or working memory constructs.

(continued)

Matrices subtest may be admin- istered in languages other than English as long as the language is shared by examiner and ex- aminee.

While Riddles subtest is not to be administered in languages other than English, responses provided by examinee in other languages may be scored as correct if the examiner knows the language and the answer is equivalent to a cor- rect English response. To fa- cilitate score, correct Spanish responses are provided in record form for Riddles Items 9–48.

Artwork is colorful and appeal- ing.

Split- half reliability coeffi cients are excellent for the IQ Com- posite and Verbal standard score with means of .93 and .90, respectively.

Test- retest reliabilities are fa- vorable for the IQ Composite, Verbal standard score, and Nonverbal standard score with means of .90, .91, and .83, re- spectively.

The KBIT- 2 was well stratifi ed and closely matches U.S.

population based on March 2001 Current Population Sur- vey.

KBIT- 2 Matrices subtest is not appropriate for individuals with signifi cant visual defi cits.

Hyperactivity Disorder, Mental Retardation, gifted and talented, trau- matic brain injury, and dementia. Study results are reported in the KBIT- 2 Manual (Kaufman & Kaufman, 2004b, pp. 164–65). Across studies, sam- ple sizes were small, ranging from 12 to 95 participants. Individuals with Mental Retardation exhibited very low means (Verbal SS = 63.1, Nonver- bal SS = 65.3, IQ Composite SS = 61.1) on the KBIT- 2, indicating that this measure demonstrates the ability to effectively assess individuals with lower cognitive abilities. The mean scores for children and adolescents in gifted programs ranged from 110 to 115 while individuals with ADHD or a learning disability had means near 90 on all three KBIT- 2 scores. Ex- aminees with speech / language disability exhibited slightly lower Verbal (M = 85) than Nonverbal scores (M = 88). The traumatic brain injury and dementia groups had mean standard scores of approximately 75 across all three KBIT- 2 indices.

Sample collection procedures and quality control measures were taken to ensure data in- tegrity.

Tables are provided in the KBIT- 2 Manual to indicate Ver- bal SS and Nonverbal SS differ- ences required for signifi cance.

Frequencies of Verbal SS and Nonverbal SS differences based on ability are also pro- vided.

Authors provide useful expla- nations on appropriateness of providing descriptive ranges and age- equivalent scores.

Although less reliable than interpreting the Verbal SS, ex- aminers may compare Verbal Knowledge and Riddles subtest scaled scores.

TEST YOURSELF

1. Test developers hoped to accomplish which of the following when revis- ing the original K- BIT?

(a) update norms

(b) develop alternative subtests to measure verbal ability (c) develop alternative subtests to measure nonverbal ability (d) both a and b

2. According to growth curve studies conducted by Kaufman and Kaufman (2004a),

(a) crystallized abilities decline at a later age than fl uid abilities (b) fl uid abilities decline at a later age than crystallized abilities (c) crystallized and fl uid abilities decline at roughly the same age (d) there is no consistent pattern of decline with age

3. The Verbal Knowledge and Riddles subtests may be administered in Spanish? True or false?

4. Basal and ceiling rules are consistent across KBIT- 2 subtests. True or false?

5. If an examinee fails a teaching item on the Riddles or Matrices subtests, the examiner should

(a) discontinue testing

(b) provide a second trial and give credit if the response is correct (c) provide a second trial but do not give credit if the response is cor-

rect

(d) immediately provide direction in your own words

6. Mike, age 5, is administered the KBIT- 2 and earns a raw score of zero on the Riddles subtest. You should

(a) calculate the IQ Composite only

(b) calculate neither the Verbal SS or the IQ Composite

(c) calculate the Verbal SS, Nonverbal SS, and the IQ Composite (d) consider the entire test invalid

7. For individuals age 7 years 4 months to 90 years, the Verbal, Nonverbal, and IQ Composite scores range from 60 to 130. True or false?

S S

8. According to Kaufman and Kaufman (2004b), providing descriptive categories

(a) is not recommended in most cases

(b) may give the false impression of a meaningful Verbal SS and Nonver- bal SS difference

(c) is useful when examinee’s Verbal and Nonverbal SS differ signifi - cantly

(d) all of the above

9. When determining whether Verbal and Nonverbal Standard Score dif- ferences are unusually large, Kaufman and Kaufman (2004b) advocate using which of the following cut- off percentages?

(a) 16%

(b) 10%

(c) 5%

(d) 1%

10. When might it be useful to determine the scaled score for the Verbal Knowledge and / or Riddles subtests?

(a) when either of the subtests cannot be administered

(b) to compare distinct receptive and expressive vocabulary abilities (c) to assess for a language- based learning disability

(d) none of the above

Answers: 1. d; 2. a; 3. False; 4. True; 5. c; 6. b; 7. False; 8. d; 9. b; 10. a

53

THE REYNOLDS INTELLECTUAL