This chapter provides an introduction to the main pilot study by describing the procedure and methodology to be followed. A more detailed description of the individual tasks involved and the rationale behind the selection of tasks included will be given in the subsequent chapters. The aim of this study was to assess the content and procedure of the Kilifi Assessment battery as a whole; and to determine the reliability of the individual tasks. Results were to be used to establish the final format of the battery.
The Assessment Team
This consisted of:
1. A Field Worker fluent in several of the local vemaculars to: a) recruit the subjects and b) interview parents.
A teacher was recruited to fulfil this role. Her training included familiarisation with the objectives of the project, to enable her to provide informed explanations to parents; and training in the administration of the questionnaire.
2. A team of four Assessors trained in the use of the Assessment Battery, and also fluent in the appropriate languages, to assess the children. This team consisted of three special needs teachers, from a local educational assessment unit. The fourth member was the associate who had been involved in the development of the material.
3. The Principal Investigator, acting as Observer, to complete a Behavioural Observation Schedule on the children, and to evaluate the procedure of the Assessment sessions.
Content of the Assessment Battery
The Pilot study consisted of:
1. An Interview with parents completed by the field worker. This was to elicit brief information about the health, school experience and family background of the child. (This included questions on the source of family income, and the size of
family). The aim was to learn more about the community, and to gain experience in how family details can be best elicited.
2. A Parent questionnaire to elicit information about the child's behaviour in the home (regarding the parents' perception of their child's level of independence, general affect, and level of concern about their child's general development). 3. A Behavioural observation of the child working in a one-one assessment situation.
4.14 Assessment Tasks, divided between two sessions of approximately 1 to 1.5 hers. long. The first session began with a test of visuo-motor speed, and then concentrated upon assessing those cognitive functions which are sampled by the Kaufman -Assessment Battery for children, to measure Sequential and Simultaneous Processing. The second half included assessments tasks concerned with Achievement, Planning, and Language.
Procedure
The pilot study was carried out between January and May 1994. Following initial recruitment the child was seen three times. On the first two visits the whole Assessment Battery was applied, and the child was seen at both visits by one of the team of four Assessors. The third visit, to collect data for a re-test reliability study, was a repetition of either Visit One or Two, and was carried out between 6 -10 weeks later by an assessor new to the child.
The first two visits were carried out at the children's homes. This meant working out of doors (with the sole exception of one home), as far from the main activity of the household as possible, and using the shelter of trees. When these were completed Visit Three, the re-test visits were begun. At this point the lack of shelter became a major problem as the rains had begun. However this period also coincided with the school holidays, and it proved possible to obtain permission from the local Education Department to use empty classrooms. In addition to problems of shelter other disadvantages of working in the home included the distraction of visitors arriving who had not been briefed about the
nature of the assessments, and the family livestock wandering through during the assessment session. However, as most schools do not have spare classrooms during term time, using schools was not an option available as an altemative to working in the home. However schools were very willing to accommodate the assessors where possible, and the children seemed relatively relaxed coming to the local school compound. It was planned to use school compounds for the main study as far as possible, and a canvas awning was made to provide shelter where necessary.
Subjects
The target group for assessment were children in the year up to their seventh birthday. To minimise language problems children who came from the majority Mgiriama sub-group of the Mijikenda were sought. Initially children were identified from the census record which is held at the KEMRI- WELLCOME Research Unit in Kilifi. The census area is divided into zones. Those zones where the population was known to be largely non-Giriama were excluded, as were those zones deemed relatively inaccessible, for logistical reasons. 4-5 children were selected at random from each of the remaining zones, and recruitment visits made to each of the identified households. Travelling between households was reduced by locating a child of the same age group living near to each of these randomly selected children. In order to ensure no duplication between the pilot and main study subjects, children known to have had cerebral malaria were excluded, which may have limited the range of abilities sampled.
Results
Description of Subiect Group
A total of 56 children were seen. (34 girls 22 bovs). Their ages ranged between 5 years 7 months and 6 years 11 months, with a mean age of 6 years 4 months.
All the children came from the Mijikenda tribal group, with the majority (95%) speaking Kigiriama as their first language. Two subjects preferred to answer in
Kiswahili, although all children appeared to understand the Kigiriama instructions. A third child spoke Kauma, which is a very closely related Mijikenda language. For the Main Study it was deemed necessary to establish the child's language of choice prior to the main assessment session. Although all the Mijikenda languages are very similar there are differences in vocabulary and pronunciation. These differences made it necessary to be prepared for changes in the instructions, and be prepared with a list of acceptable responses depending on the preferred language of each subject. These modification were relatively easily made, as the Assessors were all Mjikenda, and therefore either fluent in, or highly conversant with, the Mijikenda languages spoken in the study area. The fact that some children used Kiswahili in preference to their tribal vernacular, meant that instructions had also to be available in this language. Again the familiarity of the Assessors with this language meant that this too was not a major hurdle.
Familv Background Interview:
a) The size of the families from which the children came ranged from 2 to 11, with a mean of 5.6.
b) School attendance:
no school
stopped going to school attends nursery school attends primary school
No. of children .22 11 17 6 % 39.3 19.6 30.4 10.7
c) The health of the child was reported as Sickly Average Good No. of children 11 34 10 % 19.6 60.7 17.9 160
d) Coma: 7 children were reported as having been in coma at some time in the past. Consultation with doctors working at the unit suggested that this rate was unusually high, it was therefore concluded that the format of this question was inappropriate, and it was discarded from subsequent questionnaires.
e) The main source o f fam ilv incom e: responses fell into the following categories:
% Employed /Professional 3.6
Employed/ Manual 85.0
Farming/ Cash Crops 1.8
Farming/ Surplus Crops 1.8
Do not know 4.0
This data was compared to that collected in the area from previous studies (Marsh et al in prep.), in particular to try to sub-divide the largest of the categories reported above. The conclusion was that this data reflects the largely homogenous nature of the employment in the region, and other measures may prove more useful as socio-economic indicators.
CHAPTER 12 INFORMATION PROCESSING - THE K-ABC AND THE KILIFI