6.1 Structured interviews and individual reading sessions
6.2 The subjects
6.3 Selection of issues for interviews
6.4 Findings from interviews
Structured interviews were conducted with a view to investigating whether there were any consistent differences in background and behaviour between high and low scorers. Individual reading sessions were intended to explore the difficulties in reading reported by high and low scorers, and the extent to which they differed.
In order to carry out the structured interviews and discussions on individual reading, three research assistants were selected and, 3 months before the main data collection, received training totalling some 10 hours. Unfortunately two of the trained assistants in Malawi, and one of the trained assistants in Zambia had to withdraw very shortly before the data collection exercise. They were replaced, but the replacements received extremely brief preparation which inevitably had a negative effect on data gathering.
The structured interviews and individual readings took place in a single session. Pupils were not interviewed alone, but each was asked to choose two friends to accompany them, in order to help put the "target pupil" at ease. Questions were also put to the friends, but only answers from the "target pupil" were analysed. The structured questionnaire was administered first, and the reading and discussion were audio-recorded immediately afterwards. Discussion took place in either English or the local language.
Structured interviews and individual reading sessions were conducted with pupils from each class, selected according to their score, high, mid, or low. In this analysis only the high scoring and low scoring groups are dealt with. As the operation proceeded on a school-by-school basis, high and low are relative to the performance within each class. The range of scores for each level over all 6 schools was as follows:
Malawi:
High-scoring pupils (N=12; 6 boys, 6 girls)
Range of scores: 13-30
Low scoring pupils (N=12; 6 boys; 6 girls)
Range of scores: 3-9
Zambia:
High-scoring pupils (N=12; 5 girls; 7 boys)
Range of scores 19-30
Low scoring pupils (N=12; 7 girls; 5 boys).
Structured Interview Group. Test scores: range 0-8
Reading Session Group. Test scores1: range 1-17
1 The Zambian low-scorers in the reading session group were not identical to those in structured interview group. Eight of the latter turned out to be non-readers who could not contribute to the reading session and had to be discarded from the analysis. They were replaced by eight from a mid-scoring pool. The numbers of males and females did not change
The selection of school-related factors focused on those that are likely to contribute to pupil achievement, and where there was also a reasonable chance of collecting reliable data. Home related factors were selected on the same basis. For reasons of space, this report presents findings on only some of the factors investigated, namely:
School related factors:
- age and years of schooling
- absences from school
- language preference in reading
Home related factors:
- composition of the home group
- home socio-economic status
- home literacy in English and Nyanja (or other local language)
- books per household
- reading preferences
6.4.1 Age of starting
6.4.2 Pupil absences
6.4.3 Preferred language for reading
6.4.4 Family socio-economic status
6.4.5 Family composition
6.4.6 Literacy in the home
6.4.7 Books per household
6.4.8 Preferences in text types
In Zambia the differences between high and low scorers are small; high scorers start at a mean age of 7 years 8 months, and low scorers a little earlier at 7 years 6 months. In Malawi high scores start at a mean age of 6 years 1 month, and low scores exactly 1 year later. This does not appear to be a decisive factor.
These were established by asking the pupil, the teacher, and checking in the class register if it was available. In Malawi the term had been running for a mean of 30 school days at the time of the interviews, and in Zambia for a mean of 18. The results were:
Table 32: Absences over 30 days of high and low scorers in Malawi
|
|
Total pupil-days absent (max: 360) |
Mean days absent per pupil |
|
High Scorers (N = 12) |
30 |
2.5 (8.3%) |
|
Low Scorers (N = 12) |
42 |
3.5 (11.6%) |
Table 33: Absences over 18 school days of high and low scorers in Zambia
|
|
Total pupil-days absent (max: 216) |
Mean days absent per pupil |
|
High Scorers (N = 12) |
5 |
0.42 (2.3%) |
|
Low Scorers (N = 12) |
30 |
2.50 (13.9%) |
Low scorers tend to be absent more than high scores, with low scoring Zambian pupils having 6 times more absences than high scoring pupils. (It should be pointed out that one low-scoring pupil accounted for 14 days of the absences in that group; nonetheless even if that pupil is removed, the rate of absence of low scoring pupils is still more than 3 times that of high scoring pupils). In Malawi low scoring pupils have roughly 40% more absences than high scoring pupils. Pupil absenteeism could be a significant factor contributing to intra-school pupil variation.
However, it should be noted that absence from school may not be readily amenable to change, as it is likely to be affected by the family attitude to education, and family circumstances, especially poverty, as suggested in this extract from the interview with a 13 year old Zambian girl, who had been absent 14 times that term. Italics indicate that the original Nyanja has been translated into English. The pupil has just said that she cannot read either the English or Nyanja texts (I = interviewer; P = pupil):
I: There's nothing you know?
P: No.
I: Oh. Can't you try even a bit?
P: No.
I: How often do you come to school?
P: Ah... when I don't have a pen, I don't come to school.
I: If you don't have a pen, you don't come?
P: Yes. So like - since we opened, I have only been twice, because I didn't have a pen.
I: You usually miss school. Doesn't your mother scold you when you miss school?
P: But because I did not have the books.....
I: I see, even if you miss class, they don't scold you if you don't have books.
P: That's right.
This girl scored 3 in the English test, and 2 in the Nyanja test (both below the chance score of 4); she would appear to be for all practical purposes a non-reader. Moreover, there is probably little academic compensation in her background, which is economically poor. A further example is the following, from Zambian pupil, again translated from Nyanja:
ZCB04:
I: (.....) you are sometimes absent?
P: Yes.
I: When do you miss school?
P: Like when I don't have an exercise book.
I: When you don't have books?
P: Yes, or when I wash my uniform.
I: Oh, when you miss school, does your mother scold you?
P: Yes, she does.
I: And when she scolds you, what do you do?
P: Nothing.
While some children do not dare to come to school without books, shoes, or uniform (no longer compulsory according to the Ministry officials in both countries, but often insisted on by teachers), some may face problems at home if their parents disapprove of absence:
ZDB12:
I: (...) How often do you come to school?
P: Very often.
I: Do you ever miss school?
P: Once in a while.
I: Like when?
P: Like when I have a pain in my leg.
I: When you miss school doesn't your mother scold you?
P: She does.
I: What about at home?
P: My father beats me.
Children in both countries were asked whether they preferred to read in English or in the local language. Their responses were:
Table 34: Preferred Language for Reading in Malawi
|
|
High Scorers |
Low Scorers |
|
Prefer reading in English |
9 |
6 |
|
Prefer reading in Chichewa |
2 |
6 |
|
No preference |
1 |
0 |
There is a pronounced tendency for high scorers to prefer to read in English, while low scorers are equally divided between preferring English and Chichewa. Pupils were also asked to give reasons for their choice, with the following results:
Table 35: Reasons for Preferring English or Chichewa for Reading
|
|
In order to learn |
Because it's easy |
Other |
|
Reasons for preferring to read in English (N=15) |
11 |
0 |
4 |
|
Reasons for preferring to read in Chichewa (N=8) |
0 |
8 |
0 |
High and low scorers display instrumental motivation in reading English for self- improvement or to make progress in school, whereas Chichewa is preferred because it is easy to read. No one suggests that they read English because it is easy, or that that they read Chichewa in order to learn. The response to this question may reflect the fact that there is relatively little "academic" information available in print in Chichewa. The response also suggests that the goal of reading in English for pleasure may be difficult to attain, given that reading in English is not perceived as easy.
The responses of the Zambian children as to whether they preferred to read in English or in Nyanja, were:
Table 36: Preferred Language for Reading in Zambia
|
|
High Scorers |
Low Scorers |
|
Prefer reading in English |
12 |
9 |
|
Prefer reading in Nyanja |
0 |
3 |
There is a tendency for high and low scorers to prefer to read in English. Since some of these low scoring pupils could not read in either language, the issue of preference is problematic; inquiry on this point indicated that they were referring to the English reading lesson (as opposed to the Nyanja reading lesson), rather than the process of independent reading. Pupils were also asked to give reasons for their language choice, with the following results:
Table 37: Reasons for Preferring English or Nyanja for Reading
|
|
Because it's easy |
Other |
|
Reasons for preferring to read in English (N=21) |
18 |
3 |
|
Reasons for preferring to read in Nyanja (N=3) |
2 |
1 |
It is rather surprising that none of the pupils reports an instrumental motivation in reading English, as the Malawian pupils did. These findings are also surprising in that many pupils who prefer to read English and say that it is "easy", report difficulties in reading it. The most likely explanation is that they mean "English is relatively easier to read than Nyanja" (which is born out by their test results and negative reaction to Nyanja texts - see below), rather than simply "English is easy to read".
Categorisation according to the British system is not appropriate for Zambia or Malawi, since they have a very different socio-economic structure. Even establishing if the carers of the pupils are working or not can be difficult, as there is a tendency to count only waged occupation as "work"; thus one Zambian boy said of his uncle and grandparents, with whom he lived "They don't work, they are only farmers." For the purposes of this research, a three part socio-economic categorisation was drawn up, as follows:
0 = predominantly manual, informal training (e.g. subsistence farmer; cleaner)
1 = manual and non-manual, formal training (e.g. policeman, driver)
2 = non-manual, with secondary/tertiary education (e.g. teacher, civil servant)
The families were categorised according to the "highest status" adult in the household. The categorisations reveal the following:
Table 38: Socio-economic Status of the Families of High and Low Scorers in Malawi
|
|
Total in Category 0 |
Total in Category 1 |
Total in Category 2 |
|
High Scorers |
3 |
3 |
6 |
|
Low Scorers |
7 |
2 |
3 |
Table 39: Socio-economic Status of the Families of High and Low Scorers in Zambia
|
|
Total in Category 0 |
Total in Category 1 |
Total in Category 2 |
|
High Scorers |
3 |
3 |
6 |
|
Low Scorers |
3 |
8 |
1 |
This finding indicates a moderate positive relationship between family background and score, with high scores associated with relatively privileged family background. It suggests that in Malawi and Zambia, as elsewhere, social background influences school performance: 75% or more of low scorers are in category 0 or 1; 50% of high scorers are in the "highest" category 2. It is therefore not likely that pupil performance in general will be altered only by what happens in school. Absenteeism of children from poor backgrounds, for example, will probably persist. However, although background influences performance, it does not necessarily determine performance, and there are exceptions to the general pattern: for example 25% of high scorers in both countries are in the "lowest" socio-economic category.
Assigning Zambian and Malawian children to families of a given composition is difficult. Some children move in temporarily with different relations (e.g. in order to be closer to school), while those that live in nuclear families may have considerable contact with other relatives.
Family composition here refers to the pupils' situation at the time of the interview, and is reported here in terms of four main categories:
(i) nuclear family (mother and father, plus siblings if applicable)
(ii) mother only (plus siblings if applicable)
(iii) extended family (at least one parent, plus at least one adult relative)
(iv) other (living with relatives but with neither mother nor father)
There were no cases of children living with their father as sole carer.
Table 40: Composition of the Families of High and Low Scorers in Malawi
|
|
Nuclear |
Mother sole parent |
Extended |
Other |
|
High Scorers |
4 |
1 |
6 |
1 |
|
Low Scorers |
7 |
1 |
3 |
1 |
Table 41: Composition of the Families of High and Low Scorers in Zambia
|
|
Nuclear |
Mother sole parent |
Extended |
Other |
|
High Scorers |
9 |
1 |
1 |
1 |
|
Low Scorers |
6 |
1 |
0 |
5 |
There is no overwhelming association of any family composition with high or low scoring, although we see that nearly half of the low scoring children in Zambia live with a group that does not include either mother or father. It is possible that such children are carrying out a heavier than normal load of domestic work. However, this is not the case in Malawi. Children of families headed by a lone mother are probably under-represented in this research, since they are less likely to be attending school regularly, and hence less likely to feature in the research.
Pupils were asked to what extent the adults with whom they lived could read in English and in any local language. Answers were categorised as "Yes", "A little" or "No" based on the claim made for the most proficient adult in the household.
Table 42: Family Literacy in English of High and Low Scorers in Malawi
|
|
Literate: |
Literate: |
Literate: |
|
High Scorers |
10 |
1 |
1 |
|
Low Scorers |
7 |
3 |
2 |
Table 43: Family Literacy in English of High and Low Scorers in Zambia
|
|
Literate: |
Literate: |
Literate: |
|
High Scorers |
9 |
2 |
1 |
|
Low Scorers |
10 |
1 |
1 |
There is a weak relationship in Malawi between family literacy in English in the expected direction with higher scoring pupils tending to come from more English literate backgrounds. However, there seems to be no relationship between family literacy in English, and scoring level in English in Zambia.
Table 44: Family Literacy in Local Language of High and Low Scorers in Malawi
|
|
Literate: |
Literate: |
Literate: |
|
High Scorers |
10 |
1 |
1 |
|
Low Scorers |
11 |
0 |
1 |
Table 45: Family Literacy in Local Language of High and Low Scorers in Zambia
|
|
Literate: |
Literate: |
Literate: |
|
High Scorers |
10 |
0 |
2 |
|
Low Scorers |
10 |
1 |
1 |
There seems to be no relationship between family literacy in local language and scoring level in English in either country. However, the investigation may not have captured a crucial element, namely the family attitudes to, and practices of, literacy.
Children were asked to estimate how many books were in their home. Their answers had to be cautiously dealt with by the interviewers, since pupils may count magazines, and school exercise books taken home, as books. Responses were divided into two categories: 0 to 20, and 21 or more. Results were:
Table 46: Books per Household of High and Low Scorers in Malawi
|
|
0 - 20 books |
21 + books |
|
High Scorers |
7 |
5 |
|
Low Scorers |
12 |
0 |
Table 47: Books per Household of High and Low Scorers in Zambia
|
|
0 - 20 books |
21 + books |
|
High Scorers |
6 |
6 |
|
Low Scorers |
6 |
6 |
The expected picture emerges in Malawi with nearly half of the high scoring pupils claiming more than 20 books in the house, while none of the low scoring children do so. The Zambia finding is unexpected, and there is no obvious explanation, especially in view of the result in Malawi. However, having books is not of course the same as reading them: one child reporting a few books in the home, added that they were a gift from her father's employer, and that nobody ever looked at them! The types of books most frequently possessed were: religious (15 Zambia; 12 Malawi), educational (2 Zambia; 9 Malawi), stories (2 Zambia; 7 Malawi), related to family's work (2 Zambia; 2 Malawi).
The answers to the question "What sort of things do you like to read?" was left open, so as not to suggest responses to the children. Answers were grouped together into the following classifications, with some children expressing multiple "likes". There may be some overlap between the classification "school books" and "stories", since the school books also contain stories.
Table 48: Reading Preferences of High and Low Scorers in Zambia
|
|
High Scorers |
Low Scorers |
|
Stories |
10 |
6 |
|
School Books |
2 |
3 |
|
Religious books |
1 |
1 |
|
Other |
1 |
1 |
|
Nothing |
1 |
3 |
Table 49: Reading Preferences of High and Low Scorers in Malawi
|
|
High Scorers |
Low Scorers |
|
School books |
7 |
4 |
|
Stories |
4 |
2 |
|
Religious books |
3 |
3 |
|
Other |
3 |
4 |
School books and stories are the most popular in both countries, with religious books third, although they were the books most frequently possessed.
