The Kaufman Assessment Battery for Children (K-ABC)(Kaufman & Kaufman 1983), was selected to provide the core of the current assessment battery because: a) its theoretical foundations are rooted in an information processing approach to the definition of ability (Kaufman & Kaufman 1983; Ashman & Conway 1989). The definition it uses of intellectual ability is consistent with the theoretical framework of the current study (Vygotsky-Luria-Das), including a dichotomous division of processing skills.
b) The K-ABC procedure allows for a period of familiarisation for the child with the testing technique, incorporating a more dynamic, teach-test approach into a more structured format.
c) As a consequence of these two attributes it has also been recommended for use in cross-cultural contexts, including recent application in Sub-Saharan Africa (Giordani et al 1996).
d) Finally the K-ABC attempts to provide a description of the processes underlying competence and guidelines for remediation of the defective processing (Das 1984, Campione 1989).
More detailed justification for the selection of this battery, including: an assessment of the theoretical foundations; its applicability to the current context; the potential value of the information provided in making recommendations for future interventions; and the reason for the rejection of other batteries will follow. In order to familiarise the reader with the content of the battery, and to clarify some of these arguments, the original sub-tests, and their current adaptations, will be described first.
The K-ABC consists of two scales. The Mental Processing Scale and the Achievement Scale. The Mental Processing Scale is further divided into two forms of processing, simultaneous and sequential. Three factors, corresponding to the simultaneous-sequential-achievement divide, have been identified as underlying the structure of the K-ABC (Kaufman & Kaufman 1983).
Description Of The Battery Of Sub-Tests:
It is appropriate at this point to provide a description of the sub-tests which constitute the K-ABC, and the following descriptions will make reference to both the processes and abilities involved in each sub-test.
The three scales, Simultaneous Processing Scale; Sequential Processing Scale; and the Achievement Scale are represented in the K-ABC by 16 sub-tests, with the number administered varying according to age. The following descriptions will focus on the sub-tests from the original battery which were also selected for use in the current study (Kaufman & Kaufman 1983, Das 1984, Kaufman et al 1987, Spreen & Strauss 1992).
Simultaneous Processing Scale Magic Window (MW)
requires the child to identify and name an object whose picture is rotated behind a narrow slit, such that the picture is only seen in segments. The input is visual and it measures processing at the perceptual level. Abilities required for the successful completion of the sub-test include memory, visual perception and integration, and verbal expression. Although the stimuli are presented sequentially, simultaneous processing is required to integrate the parts into a whole. It was the best measure of simultaneous processing found for pre schoolchildren in the pre-standardisation try-out of the K-ABC (Kaufman et al
1982). Magic Window was included for children up to the age of 4 years 11 months, after which it ceased to load on the mental processing factors. However it was deemed appropriate for inclusion in the current test battery even though the children were older (6 years old), because in terms of school experience, at least half of the children would be pre-schoolers, and this can affect their processing style (Das et al 1990). The inclusion of material which might act as an ice breaker with children unfamiliar with formal instruction, would also be advantageous.
Face Recognition (FR)
Requires the child to identify and select one or two faces initially shown individually, and only briefly, from a selection of faces subsequently shown in group photographs. The input is visual, and it measures processing at the mnestic (memory) level (see the earlier section on the Information Integration Model for details). Face Recognition requires the application of the following abilities for the successful completion of the items; visual perception and integration and short term visual memory. As with Magic Window the task was designed for a younger age group than that included in the present study. The reasons for its inclusion here are identical to those given the inclusion of Magic Window, in that where the children's educational experience is limited, the informality of this assessment task makes it attractive, and a good motivator. Prior to inclusion of both of these out of range sub-tests checks needed to be made for possible ceiling effects amongst this older age range, and on the contribution that they made to the underlying construct of the test battery as a whole.
Gestalt Closure (GO
Requires the child to name partially completed silhouette drawings. It is a visual task, which measures processing at the perceptual level. It also requires the application of visuo-perceptual organisation, and spatial abilities, and involves a memory component and verbal expression.
Matrix Analogies (MA)
Requires the child to select the picture/design which best completes a 2 by 2 visual analogy. This, too, is a visual task, and all but the initial items are non verbal. It measures processing at the conceptual level. The abilities involved here include perceptual organisation, and reasoning.
Triangles/Construction (Con)
Triangles uses coloured rubber pieces to construct a design depicted in two dimensions. Early piloting of this task in the current cultural context indicated that the material was so unfamiliar that it was extremely difficult to engage the
children in the task. It was therefore replaced by a very similar task which used more familiar materials (suggested by a construction task reported in Rutter et a! 1970a). This task required the child to construct a shape with wooden sticks to match a template left exposed. Like Triangles, the new task, labelled Construction, is a non-verbal task, and measures processing at the conceptual level. It also requires spatial ability and visual-motor co-ordination.
Also part of the K-ABC Simultaneous Scale are Spatial Memory, which requires the child to recall the location of randomly arranged pictures; and Photo Series, which requires the organisation of a randomly placed sequence of photographs. These were excluded following initial piloting, which determined that these two sub-tests were difficult to introduce to, and engage the children in. The two pre school tasks described first were retained instead.
Sequential Processing Scale
All three of the K-ABC sub-tests comprising the Sequential Processing Scale were included.
Hand Movements (HM1
Requires the child to repeat a series of hand movements in a sequence which matches that produced by the examiner. This non-verbal assessment involves visual short term memory and visual motor-co-ordination. It therefore measures processing at the mnestic level.
Number Recall (NR)
Requires the child to repeat a series of numbers spoken by the examiner. This too measures processing at the mnestic level, as it involves auditory short term
memory.
Word Order (WO)
Requires the child to point at a sequence of silhouettes in the same order as previously named by the assessor, both with and without an intervening interference task. In common with the two other sub-tests included in the
Sequential Scale this too measures processing at the mnestic level. It involves auditory short term memory and verbal comprehension.
Attainment Scale
One sub-test was taken from the Attainment Scale. This was Arithmetic (Arith)
It requires the child to count, compute, identify numbers and demonstrate understanding of mathematical concepts. The process of modification and adaptation of this sub-test will be discussed in more detail in a separate section.
Pilot Study of Kilifi Simultaneous - Sequential Sub-tests The aims of the pilot study were to:
a) to calculate the test-re-test reliability for each sub-test; b) to evaluate the clarity of the procedures and the instructions c) to make a final assessment of the task content and difficulty level.
An evaluation questionnaire was completed by each assessor on their experience of the materials and procedures. Their conclusions contributed to the modifications made prior to the initiation of the Main Study. Some sub-tests required substantial modifications. When this occurred re-piloting took place, using children from two nursery classes. Both the results of the first and modified task will be reported. Tables which contain the results of the final version of the tasks will be labelled MB (main battery). The results of the sub-tests will be reported in the order that they were administered in Visit 1 of the Main Pilot Study.
Procedure
The outline of the procedure has been described in Chapter 11. To familiarise the child with each sub-test it was introduced by demonstration, and began with two, taught, sample items. For the first two scored items initial failure was followed by further teaching and a re-test. Initial success was followed by a detailed affirmation of the answer given. For these two items a pass was
recorded for success at either test or re-test. All subsequent items were administered as test only, and were divided into sections of equal difficulty and length. Failure on all items in one section led to a suspension of further administration of that sub-test. No time limits were administered on any sub-test.
Results
Reported in the following tables are the test-re-test reliabilities of each sub-test. Also reported, below the appropriate table, is the significance level of the difference in means between each trial, as an investigation of the practice effect.
Magic Window
Possible maximum score: 14
Table 12.1 Magic Window (MB)
Mean s.d. min max corr. 2-tailed siq.
Time 1 3.6 2.2 1 11
0.76 <0.01
Time 2 5.8 3.4 2 13
Paired Diff. Mean -2.12, s.d 2.25, t-value -4.73, df 24, 2-tail Siq <.001 ,95% Cl (-3.05, -
1.10)
The materials appeared to attract the children's attention although the medium was relatively unfamiliar. Highlighted was the need for sight screening prior to testing. Consistently poor results from two items suggested that they needed to be replaced with more familiar images. Also recommended was an alteration in the introductory instructions.
Face Recognition
Possible maximum score: 7
Table 12.2 Face Recognition
Mean s.d. min max corr. 2-tailed siq.
Time 1 2.58 1.22 0 6
0.68 <0.01
Time 2 3.50 1.68 1 7
Although this correlation was significant the modifications suggested were followed in order to a) extend the possible range of scores by increasing the number of items; b) to provide clearer instructions when the probes were
increased from one to two photographs. The longer version, with an improved reliability, was re-piloted on 12 children.
Possible maximum score: 15
Table 12.3 Face Recognition (MB)
Mean s.d. min max corr. 2-tailed siq.
Time 1 10.00 2.41 5 13
0.91 <0.01
Time 2 10.00 1.95 7 13
Paired Diff. Mean .00, s.d. 1.04, t-value .00, df 11 2-tail Siq > 1 .0 .95% Cl (-.66, .66)
Gestalt Closure
Possible maximum score: 16
Table 12.4 Gestalt Closure (MB)
Mean s.d. min max corr. 2-tailed siq.
Time 1 5.96 2.86 2 11
0.81 <0.01
Time 2 6.84 2.44 0 12
Paired Diff. Mean -.89, s.d. 1.70, t-value -2.65, df 25 2-tail Sia <.05 95% Cl (-.1.57, -.20)
Modifications suggested included giving clearer guidelines to the Assessors to make the administration of the training items more uniform; and to allow a wider interpretation of one item for which a second response was relatively common. As these required only minor alterations no re-piloting was deemed necessary.
Matrix Analogies
Possible maximum score: 15
Table 12.5 Matrix Analogies
Mean s.d. min max corr. 2-tailed siq.
Time 1 5.92 2.86 2 13
0.23 0.23
Time 2 7.15 2.88 2 11
The teaching items did not proceed smoothly, with some stumbling over the instructions, and random guessing by the children. The drawings in the verbal items proved to be a little ambiguous. It was therefore decided to improve the procedure; remove the picture items, and replace them with some simpler design questions. The modified version was re-piloted on 19 children.
Table 12.6 Matrix Analog ies (MB)
Mean s.d. min max corr. 2-tailed siq.
Time 1 7.68 2.71 2 11
0.35 0.14
Time 2 8.89 2.18 4 12
Paired Diff. Mean -1.21, s.d. 2.82, t-value -1.87, df 18 2-tail Siq <.10 95% Cl (-2.57,1.49 )
These modifications resulted in improved overall performance, but the reliability was not significant. Further changes were made to the procedure, but without further test-re-test data being collected. It was decided to retain this item, although because of the poor reliability data, there may be limitations in the sensitivity of this sub-test to detect group differences.
Construction
Possible maximum score: 10
Table 12.7 Construction
Mean s.d. min max corr. 2-tailed siq.
Time 1 3.57 1.9 1 10
0.27 0.17
Time 2 4.50 2.0 1 9
The selection of items needed to include a wider variety of difficulty levels, to allow a greater lead in time in particular. Also recommended was a change in the ordering of the items, as the progression of difficulty found did not correspond with the original ordering. The new longer version was re-piloted on 19 children.
Possible maximum score: 28
Table 12.8 Construction (MB)
Mean s.d. min max corr. 2-tailed siq.
Time 1 18.42 5.72 6 27
0.73 <0.01
Time 2 20.89 4.59 12 27
Paired Diff. Mean -2.47, s.d 3.96, t-value -2.72, df 18, 2-tail Siq <.05 ,95% Cl (-4.38, .56)
Implemented for the Main Study was also the procedure that the final score for each item was arrived at after consultation between the Assessor and Observer, both of whom kept sketches of the responses.
Hand Movements
Possible maximum score: 22
Table 12.9 Hand Movements
Mean s.d. min max corr. 2-tailed siq.
Time 1 8.93 2.18 4 17
0.37 0.06
Time 2 8.81 2.73 3 16
Although it was easy to introduce the task to the child, its length made it harder to sustain interest. The task was therefore broken down, to reduce the number of items that needed to be passed at each stage. The format used was similar to that employed in the British Ability Scales, where the child is credited with the score for the whole section even after having successfully completed the first two items. The modified version was re-piloted on 21 children.
Possible maximum score: 22
Table 12.10 Hand Movements (MB)
Mean s.d. min max corr. 2-tailed siq.
Time 1 14.95 2.48 6 20
0.50 0.02
Time 2 14.81 3.12 11 20
Paired Diff. Mean .14 s.d. 2.87, t-value .23, df 20 2-tail Siq <1.0 95% Cl (-1 .1 6 ,1 .4 5 )
Some improvement was found in the reliability measure with these minor changes. The sub-test was retained, although, given the relatively poor reliability, potential limitations exist in the sensitivity of this sub-test to between group differences.
Number Recall
Possible maximum score: 18
Table 12.11 Number Recall (MB)
Mean s.d. min max corr. 2-tailed siq.
Time 1 11.54 2.69 4 18
0.70 <0.01
Time 2 12.88 2.69 8 17
Paired Diff. Mean -1.33 s.d. 2.08, t-value -3.14, df 23 2-tail Siq <.005 95% Cl (-2.21, -.46)
This sub-test proved easy to transfer for use in the current context, although the importance of the order of the numbers needed to be stressed in the demonstration and instructions.
Word Order
This was carried out in two parts, a memory task, a) simple b) with interference.
Possible maximum score: 10
Table 12.12 Word Order-Sim ple (MB)
Mean s.d. min max corr. 2-tailed siq.
Time 1 5.00 2.30 2 10
0.75 <0.01
Time 2 5.48 1.97 2 10
Paired Diff. Mean -.48, s.d 1.55, t-value -1.61, df 26, 2-tail Siq <.15 ,95% Cl (-1.10, .13)
Table 12.13 Word Order -Interference
Mean s.d. min max corr. 2-tailed siq.
Time 1 3.34 2.12 1 8
-0.39 0.85
Time 2 3.73 1.76 0 9
Interest was hard to maintain over the whole task, particularly following the introduction of the interference task. This was poorly administered, and led to some confusion. In the original K-ABC this age group was at the borderline for the administration of the interference section. Given all these issues it was decided to discard the interference section from this sub-test, and to extend the number of items in the simple task.
Practice Effects
A significant practice effect was found, even after the 6 to 10 week interval, but only on four of the eight sub-tests (Construction, Gestalt Closure, Magic Window, and Number Recall.)
Goals of the K-ABC
The following section aims to evaluate the use of the K-ABC format as an assessment tool for the current context, in the light of the results of the current adaptation, and of previous applications.
In the K-ABC Interpretive Manual the primary goals of the battery were identified as:
2. to separate acquired factual knowledge from the ability to solve unfamiliar problems
3. to yield scores that translate to educational intervention 4. to include novel tasks
5. to be easy to administer and objective to score
6. to be sensitive to the diverse needs of pre-school, minority group, and exceptional children. " (Kaufman & Kaufman 1983 p. 5).
Each of these will be discussed primarily with reference to the original K-ABC. In order to take account of any possible alterations following the modifications here adopted, the results of studies reported on the K-ABC sub-tests will be compared to the results of factor analysis carried out on the final version of the sub-tests included in the Kilifi battery. The Kilifi sample thus included subjects from the main study, (N=174, Age = 6 years).
Goals 1 and 2. The Theoretical and Research Basis, and the Definition of