4. Barbados study: Quantitative survey and case studies

4.1 The quantitative survey background:
4.2 Barbados case study: Harbour, 6th form co-educational secondary school
4.3 Barbados case study: South, single sex, female-only (newer, non-sixth form) secondary school

4.1 The quantitative survey background:

This report began with the quantitative assertion that females were performing better in schools than boys in Trinidad and questioned whether this finding generalised to other Caribbean countries. Further, we were concerned to ascertain whether there was a 'drop-out' among boys as they progressed through the school years.

The generalisation of gender differentiated achievement and staying-on in the Caribbean required that both quantitative and qualitative studies be undertaken in another country and Barbados was chosen to provide this comparative perspective. A similar argument will be presented in the next chapter for the inclusion of St Vincent. These two additional islands present structurally similar studies, but are reported separately because they represent dissimilar populations. The school system in Barbados allows for universal primary and secondary education, similar to the system in Trinidad. The school system in St Vincent offers universal primary education. Secondary education is only available for 60% of the primary school leavers.

The Barbados section of this report establishes quantitative rates and background factors correlated to academic success and failure as well as using small, focused case studies to obtain qualitative insights into female and male learning strategies. The Barbados study substantiates speculative information concerning girls' successful educational performance at both primary and secondary school levels. To substantiate the speculative information quantitative information was collected concerning primary and secondary pupil performance throughout the country, using a representative sample similar to the sample used by Jules & Kutnick (1990). Information collected included: paternal and maternal occupation, sex of pupil, age of pupil, birth placement in family, end-of-term test scores and Common Entrance Examination scores (where appropriate). Additionally, children were asked to provide further information about their lives at home and this included: whether they undertook jobs (chores) in and around the home, who the child lived with, who helped the child with homework and whether the child attended a pre-school. Upon completion of quantitative data collection and preliminary analysis, secondary schools were selected for case study (using ethnographic techniques).

THE SAMPLE:

To report on within-class and national achievement scores for children attending the state funded schools of Barbados a sample representative of the school population as a whole was required. The sample included a proportional and focused selection of all types of the stratified primary and secondary schools within the island. Stratification within primary school types are found in:

a) all-through co-educational primary schools,

b) male-only and female-only primary schools,

c) composite schools (primary schools with additional year groups for those children who did not pass the CEE but had not reached the age at which they could leave compulsory schooling), and

d) senior schools (similar to composite schools, but without the lower years of the primary school).

There are very few private primary schools in Barbados, and they were not included in the sample. Also, there are no state-funded, religious maintained primary schools in Barbados. Of the eighty-four primary schools on the island, twenty-one schools were identified for inclusion in the study. Selection of actual schools was proportional to type of primary school within the whole population of primary schools. Actual number and types of primary schools selected included: 13 co-educational primary schools, 2 male-only primary schools, 2 female only primary schools, 3 composite schools, and 1 senior school.

All secondary schools in Barbados are comprehensive, as they provide a full range of curriculum subjects for study. All but four of the secondary schools are co-educational. Secondary schools are differentiated or stratified on two criteria: whether the schools have a sixth form, and criteria based on CEE results. Stratified types of secondary schools within Barbados include:

a) long established, co-educational schools that include a sixth form,
b) established, co-educational schools without a sixth form,
c) newer co-educational schools without a sixth form, and
d) single-sex schools without a sixth form.

There are twenty-two secondary schools in Barbados. The sample selected twelve schools for inclusion in this study. The twelve schools included: 4 sixth form, co-educational schools; 3 established co-educational schools without sixth forms; 3 newer co-educational schools without sixth forms; 1 boys-only school without a sixth form; and 1 girls-only school without a sixth form.

Within the sample, information was not collected from all children in each of the schools. In parallel with the strategy used in Trinidad, five year levels were identified for study. In each primary school one class of the Standard 1 level (children aged approximately 7 to 8 years) was randomly selected and all classes at the Standard 4 level (children aged 10 to 11 years, the top year level for primary schools) were selected. Random selection of Standard 1 classes was made because these were mixed ability classes. All Standard 4 classes were selected to include any streaming (tracking) strategy used in the school. In each secondary school a Form 2 and a Form 4 class were randomly selected for inclusion. In the sixth form schools, random selection also included a Form 6 class. Overall, data was recorded for 2255 children. There were 1551 primary school children and 704 secondary school children surveyed. The sampling strategy drawn upon was proportional (in relation to the number and types of schools in Barbados), focused (in the choice of survey school within the appropriate proportions), stratified (by type of primary and secondary school, and by year level in school), random (by choice of class per year level except for Standard 4), and clustered (in that information was obtained for all of the children in each of the classes chosen for inclusion).

INFORMATION PROVIDED BY THE SURVEY:

1. Who is succeeding in school? This first, and main, question was not simple to answer even if we focused on differences between males and females solely. Initially, consideration must be given to success within-class and by curriculum subject. These questions were asked to confirm the generalisability of female success found in the Trinidad results and to ascertain whether stereotypical course preference (male preference for mathematics and sciences and female preferences for humanities and social sciences) was shown in class attainment. Within-class scores are the essential starting point for understanding academic success; for it is within the intensity of classroom interaction that children are provided the encouragement and motivation to succeed through the feedback and responses given to their work. Further, the question of success should move beyond the within-class level to the national level to ascertain whether the attainments of males and females within-class can be generalised to a different assessment situation.

Within-class attainment: Drawing upon the 'raw' scores (actually assigned by teachers on the end-of-term examinations), girls were found to perform at significantly higher levels than boys in the core subjects (of English, mathematics and social studies) and integrated sciences (see Table 4.1.1). The scores show a superior performance of girls at a general level, but do not account for variations between teachers (in their ability to write 'fair' tests and the difference in subject matter that may be covered in each individual class). Standard deviated scores for each subject by class within each school were calculated so that valid comparisons could be made between schools. (Standard deviated scores assume that the scores from each class form a normal distribution for that class and a standard deviation for the class mean can be calculated for each child. Thus, while class scores may vary, deviations from the mean remain a consistent and comparable feature between classes. This calculation is useful for comparisons with regard to the interpersonal variables such as sex, parental occupation, etc., but cannot be used to compare between whole classes (and schools) - as these variables will contain the same range of standard deviated scores.)

Table 4.1.1: Raw and standard deviated scores for core subjects and significance of differences between boys and girls.

Core Course	Average Male Score	Average Female Score	Difference based on Raw Score	Difference based on S.D. Score
English	58.23	64.33	F1,2129= 52.186, p<0.0001	F1,2128= 64.292, p<0.0001
Maths	57.36	63.02	F1,2147= 37.679, p<0.0001	F1,2146= 23.530. p<0.0001
Integrated Science	59.79	63.19	F1,1508= 9.578, p<0.002	PI,1502= 8.896, p<0.003
Social Studies	57.84	64.25	F1,1790= 46.786, p<0.0001	F1,1787= 22.555, p<0.0001

Table 4.1.1 shows that girls scored significantly higher than boys when drawing upon the raw or standard deviated scores. On average the girls scored six percentage points higher than boys. These scores (as shown in the Trinidadian case studies) allow a higher placement in classroom ranking and the possibility of more (and more positive) feedback from their teachers.

Within the secondary schools, where differentiated and more extensive courses were run, the differences between males and females were not so clear cut (see Table 4.1.2).

Table 4.1.2: Sex differences among secondary school students by subject and reporting raw and standard deviated scores (number of students taking these courses in brackets).

Courses	Average Male Score	Average Female Score	Difference based on Raw Score	Difference based on S.D. Score (compari-sons for coed. schools only)
English	56.97 (303)	58.77 (337)	N.S.	F1,521-22.548 p<0.0001
Maths	52.80 (311)	56.13 (343)	F1,652= 5.165, p<0.023	N.S.
Social Studies	58.04 (276)	61.53 (305)	F1,578= 5.905, p<0.015	F1,482= 9.440, p<0.002
Biology	48.29 (38)	55.10 (82)	N.S.	F1,109= 10.38, p<0.002
Chemistry	48.30 (43)	52.75 (77)	N.S.	N.S.
Physics	52.56 (36)	54.02 (50)	N.S.	N.S.
Foreign Language	57.27 (220)	65.53 (258)	F1,476= 27.654, p<0.0001	PI,407= 37.436, p<0.0001
Business Studies	55.84 (88)	59.04 (99)	N.S.	F1,136= 3.863, p<0.051
Industrial Arts	49.89 (91)	51.26 (38)	N.S.	F1,110= 11.292, p<0.001
Fine Arts	51.46 (54)	51.10 (87)	N.S.	N.S.

Within secondary schools, girls were found to perform consistently better than boys although then-performance (based on raw scores) may not have achieved significance in all subjects. Use of the standard deviated scores provides a clearer analysis, but comparative results can only be reported for students in co-educational schools. In no subject did boys perform better than girls. Girls maintained

their comparative and significant superiority in English, social studies, biology (but not the other differentiated sciences), foreign languages, business studies and industrial arts. Table 4.1.2 also shows that a greater number of girls than boys were found taking courses in each subject except Industrial Arts. The superior number of girls was found (especially) in 'traditional' male subjects of mathematics, chemistry and physics.

When raw score performance was compared by sex in relation to type of school attended the consistent female dominance was not maintained (Table 4.1.3). This table used an average of the raw scores for the core subjects (including an average score for the individual sciences at the secondary level). As noted, girls scored consistently higher than boys but did not achieve a significant difference in the Senior Primary schools (where very few pupils attended) or in the newer co-educational secondary schools.

Table 4.1.3: Raw score performance of boys and girls by type of school attended (for co-educational schools only)

Type of school	Average Male Score	Average Female Score	Difference based on Raw Score	Difference based on S.D. Score
Primary -general	59.42	67.04	F1,1082= 40.668, p<0.0001	F1,1082= 34.633, p<0.0001
Primary -Composite	55.02	65.86	F1,127-8.424, p<0.004	F1,127= 9.481, p<0.003
Primary-Senior	54.22	59.62	N.S.	N.S.
Secondary (older)	54.91	58.82	F1,156= 6.427, p<0.012	F1,156= 8.501, p<0.004
Secondary (newer)	51.34	54.70	F1,144= 3.499, p<0.063	F1,144= 3.563, p<0.061
Sixth Form	58.22	61.82	F1,260= 5.277, p<0.022	F1,259= 11.540, p<0.001

A closer inspection of the results of students in the sixth form schools shows that there were no significant differences between boys and girls in any of the separate core curriculum subjects. Girls' slightly higher averages in each curriculum subject obviously was the contributing factor to a significantly higher averaged end-of-term raw score. Nonsignificant differences in the separate curriculum subjects was characteristic of each of the three form levels (2nd, 4th, and 6th). A similar pattern was found when the same analysis was undertaken using the standard deviated scores. Nonsignificant differences also characterised the separate core curriculum subjects in the newer secondary schools. Within the older secondary schools, girls maintained their superiority over boys (on both raw score and standardised scores) in English and social studies, although differences became non-significant in mathematics and science. These findings both confirm the expected female superiority in overall classroom performance and indicate that both male and female scores can vary to include high scoring males and low scoring females.

One final way of looking at the raw scores for these children was to compare average raw scores for each year level and in each subject (Table 4.1.4). This comparison showed a gradual improvement of males during the years of secondary schooling although a greater number of females were found to be studying in each of the subjects. The gradual increase in male scores is partially accounted for by the academic selectivity of males who remained in secondary schooling through to sixth form (see further subsection on drop-out).

Table 4.1.4: Comparison of raw scores of males and females by year level and subject (number of students per individual science course in brackets)

LEVEL/ Subject	STANDARD ONE	STANDARD FOUR	FORM 2	FORM 4	FORM 6
English:
male	66.39	55.34	59.92	52.18	59.33
female	74.69	63.29	61.59	56.95	52.77
Social Studies:
male	65.93	54.45	60.90	53.71	54.35
female	76.03	61.35	65.89	58.26	54.08
Maths:
male	65.85	56.31	55.18	49.74	51.59
female	73.45	62.86	55.13	56.74	57.86
Science:
male	68.32	57.34	59.32
female	76.56	61.46	58.91
Biology:
male				50.74(34)	55.00(2)
female				58.88(65)	57.58(12)
Chem.:
male				49.24(34)	50.38(8)
female				56.97(62)	40.77(13)
Physics:
male				54.36(25)	59.22(9)
female				56.44(36)	51.46(13)

CEE, national comparison: Common Entrance Examination scores are the one example of a national examination, with consistent criteria for scoring across all schools. Girls showed significantly higher scores than boys on the CEE (F1,1730= 67.139, p<0.0001; boys average = 51.36, girls average = 60.68). Table 4.1.5 displays that girls scored higher than boys in the co-educational school types but the difference was not significant in the schools with Sixth Forms. Scores for single-sex schools are added for information only.

Table 4.1.5: Average CEE scores by sex within types of secondary schools (number of students contributing to each average in brackets)

Type of school	Average Male Score	Average Female Score	Difference
Secondary (older)	71.07 (71)	74.20 (78)	F1,147= 5.583, p<0.019
Secondary (newer)	40.03 (81)	47.14 (59)	F1,138= 4.265, p<0.041
Secondary Boys only	42.58
Secondary Girls only		45.48
Sixth Form	82.51 (93)	83.16 (156)	N.S

The letter grade assigned to written work on the CEE again confirmed that girls scored higher than boys (see Table 4.1.6). The differences shown in the table are confirmed statistically (X\^) = 91.862, p<0.0001).

Table 4.1.6: Number of students obtaining each of CEE letter grades by sex (percentage of sex by grade in brackets)

SEX	LETTER GRADE					Row Total
	E	D	C	B	A
male	85	175	264	180	99	803
	(10.6)	(21.8)	(32.9)	(22.4)	(12.3)
female	30	116	258	275	202	881
	(3.4)	(13.2)	(29.3)	(31.2)	(22.9)
Column Total	115	291	522	455	301	1684

Streaming: CEE were not taken until children were ready to leave primary school, thus performance of boys and girls in primary schools could not be compared nationally. Some insight can be gained through the inspection of the placement of males and females into streamed classes; selection into top stream was based on the school's perception of the child's academic ability. If the school was large enough to sustain more than one class per age group, pupils were placed in academic ability streams for the duration of their final year in primary school (Standard 4). Among the streamed classes, girls and boys were found at all levels. Table 4.1.7 shows that a higher proportion of boys were found in the lower streams and more girls were in the top stream in schools; this difference was significant (X² (3) = 16.591, p<0.0009).

Table 4.1.7: Placement of boys and girls in various stream levels in the top year of their primary school (percentages by sex per stream in brackets)

SEX	STREAM				ROW Total
	bottom stream	high bot torn stre	next to top stre	top stream
male	40	97	129	121	387
	(54.1)	(59.5)	(53.1)	(41.2)	(50.0)
female	34	66	114	173	387
	(45.9)	(40.5)	(46.9)	(58.8)	(50.0)
Column	74	163	243	294	774
Total	(9-6)	(21.1)	(31.4)	(38.0)	(100.0)

2. What was the distribution of male and female pupils in each of the year levels, and was there evidence of a male 'drop-out' rate at the secondary school level?

Overall, there was a roughly equal distribution of male and female students in the survey; 48.5% of the sample children were male and 51.5% were female. The breakdown of males and females by age group (or year in school) shows an increasing 'drop-off in the number of males attending school as students progressed through the secondary years (see Table 4.1.8); the nearly equal distribution of males and females in the standard one year changed to nearly two-thirds female in the sixth form. Evidence of this drop-off is first shown at the Form 4 level, although this is seen even more dramatically at the sixth form.

Table 4.1.8: Distribution and percentage of participation of males and females by year in school (percentage by year group in brackets)

SEX	SCHOOL YEAR					Row Total
	standard one	standard four	form 2	form 4	form 6
Male	239	517	169	133	25	1083
	(49.8)	(48.8)	(52.2)	(44.3)	(35.7)	(48.5)
Female	241	542	155	167	45	1150
	(50.2)	(51.2)	(47.8)	(55.7)	(64.3)	(51.5)
Column	480	1059	324	300	70	2233
Total	(21.5)	(47.4)	(14.5)	(13.4)	(3.1)	(100.0)

Explanations for the drop-off in male attendance were explored (post hoc) statistically. Initial speculation questioned whether the boys who dropped out may have come from a particular social class background. Social class of fathers was identified in the categories used in the Barbados census, and the individual categories were placed into three broader groupings (of professional-managerial, sales-skilled, craft-unskilled). Distribution of the surveyed boys into these groupings (Table 4.1.9) shows a predominance of boys coming from a craft-unskilled background, and approximately one-third of the boys did not know the occupation of their fathers (or did not know their fathers at all).

Table 4.1.9: (Grouped) Paternal occupations of boys surveyed

Value Label	Frequency	Percent	Valid Percent	Cum Percent
prof/ man	239	22.0	31.0	31.0
sales/ skilled	146	13.5	18.9	49.9
craft/ unskilled	386	35.6	50.1	100.0
no occupation identified	313	28.9	Missing
Total	1084	100.0	100.0

Table 4.1.10 displays the distribution of boys by social class at each year level. The table shows a skew of higher social class in the primary school year level and the opposite at sixth form; the difference in parental occupation by year level was statistically significant (X² (8)=51.859, p<0.001). Of the twenty-one boys who identified their father's occupation at Form 6, three-quarters were from a craft-unskilled background.

Table 4.1.10: (Grouped) paternal occupation of surveyed boys by year level

YEAR	FATHER'S OCCUPATION			Row Total	No occupation provided
	Manager	Skilled	Unskilled
standard one	110	31	37	178	61
	(61.8)	(17.4)	(20.8)	(23.1)
standard four	194	69	97	360	157
	(53.9)	(19.2)	(26.9)	(46.8)
form 2	41	26	55	122	47
	(33.6)	(21.3)	(45.1)	(15.8)
form 4	37	18	34	89	44
	(41.6)	(20.2)	(38.2)	(11.6)
form 6	3	2	16	21	4
	(14.3)	(9.5)	(76.2)	(2.7)
Column	385	146	239	770
Total	(50.0)	(19.0)	(31.0)	(100.0)

As paternal occupation was negatively correlated with staying-on of boys to sixth form, it was worthwhile to explore whether the same relationship was true of maternal occupation and boys who stayed in school. Table 4.1.11 presents a very different picture for the support of the academic staying-on of these boys and maternal occupation. A majority of the boys in sixth form who provided information about maternal occupation showed that their mothers were well educated and held responsible positions.

Table 4.1.11: (Grouped) maternal occupation of surveyed boys by year level

YEAR	MOTHER'S OCCUPATION			Row Total
	professional /man	sales/ skilled	craft/ unskilled
standard one	26	51	62	139
	(18.7)	(36.7)	(44.6)	(20.2)
standard four	68	148	128	344
	(19.8)	(43.0)	(37.2)	(50.0)
form 2	36	51	28	115
	(31.3)	(44.3)	(24.3)	(16.7)
form 4	22	34	19	75
	(29.3)	(45.3)	(25.3)	(10.9)
form 6	8	5	2	15
	(53.3)	(33.3)	(13.3)	(2.2)
Column	160	289	239	688
Total	(23.30	(42.0)	(34.7)	(100.0)

A further post hoc comparison for males, showed that there were few differences in the home backgrounds of those who stayed to sixth form from the younger boys. The comparison included variables with regard to chores and responsibilities that they undertook at home, whether they had attended pre-school, or with whom they lived. The males who stayed-on to sixth form had a significantly higher CEE score than other boys in secondary schools (F2,299=13.453, p<0.0001); their average score was 85 while boys in fourth form was 58 and second form was 60. Those in sixth form also had a better attendance record than other boys (F2,201=4.227, p<0.016); those in sixth form only missed an average of 2.4 days in the current academic year, while those in fourth form missed an average of 6.7 days and second form missed 4.8 days.

3. What explanations can be offered/or school-based success and failure. Within-class and CEE results and parental occupation: As would be expected, there were significant differences in core curriculum scores and CEE scores found between children whose parents worked in different occupations. As an example, Table 4.1.12 displays differences in average end-of-term scores and CEE scores by paternal and maternal occupation. This table shows that children whose fathers and mothers work in higher, more educated occupations scored consistently high on end-of-term tests and on the CEE examination. These results also characterised separate core curriculum end-of-term tests.

Occupation	Average end-of-term score by father occupation	Average end-of-term score by mother occupation	CEE score by father occupation	CEE score by mother occupation
top managers	64.14 (134)	64.81 (62)	70.55	72.85
professional	67.40 (213)	67.33 (264)	74.06	73.56
tech assoc prof	64.79 (158)	67.18 (61)	70.69	72.98
clerks	67.70 (76)	64.42 (321)	71.82	66.49
service/ sales	60.43 (213)	59.13 (271)	50.65	51.00
skilled agric/fish	61.07 (14)	68.46 (3)	59.96	79.17
crafts	59.58 (464)	59.56 (99)	51.12	51.99
machine operators	61.50 (138)	63.33 (37)	49.77	57.54
elementary	59.71 (201)	56.93 (370)	46.06	45.10
difference	F8,1579= 6.522, p<0.0001	F8,1456= 10.726, p<0.0001	F8,1234= 43.036, p<0.0001	F8,1161= 46.237, p<0.0001

Table 4.1.13: Distribution of pupils in types of primary school by grouped paternal occupation (percentages of occupation by school type presented in brackets)

SCHOOL TYPE	FATHER'S OCCUPATION			Row Total
	prof/ man	sales/ skilled	craft/ unskille
Primary coed.	421	162	245	828
	(67.8)	(73.6)	(86.9)	(73.7)
Primary - boys only	74	12	12	98
	(11.9)	(5.5)	(4.3)	(8.7)
Primary - girls only	58	24	14	96
	(9.3)	(10.9)	(5.0)	(8.5)
Primary - composite	54	20	11	85
	(8.7)	(9.1)	(3.9)	(7.6)
Senior school	14	2		16
	(2.3)	(0.9)		(1.4)
Column	621	220	282	1123
Total	(55.3)	(19.6)	(25.1)	(100.0)

Table 4.1.14: Distribution of pupils in types of primary school by grouped maternal occupation (percentages of occupation by school type presented in brackets)

SCHOOL TYPE	MOTHER'S OCCUPATION			Row Total
	prof/ man	sales/ skilled	craft/ unskille
Primary coed.	182	321	248	751
	(82.7)	(76.2)	(65.1)	(73.5)
Primary - boys only	13	26	43	82
	(5.9)	(6.2)	(11.3)	(8.0)
Primary - girls only	17	39	38	94
	(7.7)	(9.3)	(10.0)	(9.2)
Primary - composite	8	33	41	82
	(3.6)	(7.8)	(10.8)	(8.0)
Senior school		2	11	13
		(.5)	(2.9)	(1.3)
Column	220	421	381	1022
Total	(21.5)	(41.2)	(37.3)	(100.0)

Distribution of types of secondary school attended by parental occupation presents a slightly different picture. There was a predominance of top managerial and professional parents sending their children to schools with sixth forms and the older secondary (non-sixth form) schools. And, there was a higher proportion of children from crafts and unskilled parental occupations attending the single sex schools and newer secondary schools (see Tables 4.1.15 and 4.1.16).

Table 4.1.15: Distribution of pupils in types of secondary school by grouped paternal occupation (percentages of occupation by school type presented in brackets)

SCHOOL TYPE	FATHER'S OCCUPATION			Row Total
	prof/ man	sales/ skilled	craft/ unskille
Secondary - Coed (older)	52	20	37	109
	(23.3)	(24.1)	(20.3)	(22.3)
Secondary - Coed (newer)	28	16	46	90
	(12.6)	(19.3)	(25.3)	(18.4)
Secondary - Male Only	7	7	30	44
	(3.1)	(8.4)	(16.5)	(9.0)
Secondary - Female only	4	4	24	32
	(1.8)	(4.8)	(13.2)	(6.6)
Sixth form - Co-educational	132	36	45	213
	(59.2)	(43.4)	(24.7)	(43.6)
Column	223	83	182	488
Total	(45.7)	(17.0)	(37.3)	(100.0)

Table 4.1.16: Distribution of pupils in types of secondary school by grouped maternal occupation (percentages of occupation by school type presented in brackets)

SCHOOL TYPE	FATHER'S OCCUPATION			Row Total
	prof/ man	sales/ skilled	craft/ unskille
Secondary - Coed (older)	41	50	19	110
	(24.6)	(28.7)	(15.2)	(23.6)
Secondary - Coed (newer)	9	26	32	67
	(5.4)	(14.9)	(25.6)	(14.4)
Secondary - Male Only	3	22	18	43
	(1.8)	(12.6)	(14.4)	(9.2)
Secondary - Female only	5	9	25	39
	(3.0)	(5.2)	(20.0)	(8.4)
Sixth form - Co-educational	109	67	31	207
	(65.3)	(38.5)	(24.8)	(44.4)
Column	167	174	125	466
Total	(35.8)	(37.3)	(26.8)	(100.0)

Insights into the distribution of children of various parental occupations into the types of primary and secondary schools is important because the distinct types of school may provide differential status and feedback to the child - especially with regard to the scores that the child obtains in the school.

Raw scores and CEE scores by type of school:

Secondary schools: End-of-term scores and CEE scores corresponded with prestige of secondary school such that the higher the status of the school the higher the average score. Table 4.1.17 displays the average end-of-term scores, CEE scores, and scores in the core curriculum subjects by type of secondary school.

Table 4.1.17: Average end-of-term, CEE, and core curriculum subject scores for types of secondary school

Type of school	Average score	CEE score	English	Maths	Science	Social Studies
Older	56.89	72.71	58.63	52.17	50.07	59.76
Newer	52.82	43.03	54.89	49.73	53.66	53.02
Boys	53.64	42.58	56.97	48.78	57.77	56.17
Girls	49.69	45.48	47.74	51.88	39.93	48.79
6th Form	60.46	82.83	62.45	61.36	66.93	66.58
Diff.	F4,687=17.739 p<0.0001	F4,657=328.692 p<0.0001	P4,645=14.492 p<0.0001	F4,659=13.778 p<0.0001	F4,361=23.204 p<0.0001	F4,585=21.333 p<0.0001

The table clearly shows that students in the sixth form schools were assigned higher average and separate curriculum scores than the other schools, and had a much higher CEE score profile (these were confirmed on post hoc Scheffe tests to the 0.05 level of probability). One anomaly arises in this table, girls in the girls-only school scored higher than the median on CEE but were assigned consistently low average scores overall, in English, science and social studies. The same pattern was found in a breakdown by form levels at both second form and fourth form. An explanation for this anomaly may reside in the significant social class differences of the children attending different types of secondary school. Students attending single-sex schools (especially the girls school) were more likely to come from a craft/unskilled (paternal) background than any of the other school types.

One further, but substantial, point should be made from this data. When the analysis focused solely on the schools with a sixth form, there were no significant differences found between children from the various paternal and maternal occupations with regard to the raw and standardised scores in the core and other subjects. There was a significant difference for the CEE scores that were taken before entry to secondary school (paternal occupation F2,199= 11.773, p<0.0001, maternal occupation F2,193= 4.444, p<0.013). This finding was similar to the study undertaken in Trinidad and demonstrated a 'democracy among the elite'; once the child gained entry to the top schools, social class differences became non-significant.

Primary schools: Between types of primary school a number of differences were found (Table 4.1.18), although these differences were not as dramatic as those found between the secondary schools.

Table 4.1.18: Average end-of-term, CEE, and core curriculum subject scores for types of primary school

Type of school	Average score	CEE score*	English	Maths	Science	Social Studies
Co-ed	63.34	52.89	64.71	64.98	63.98	61.97
Boys	56.92	41.89	57.28	55.22	60.32	55.70
Girls	60.38	51.91	59.04	56.79	61.66	64.98
Compos	60.86	46.76	58.29	59.61	64.44	63.42
Senior	55.04	13.54	51.90	52.32	-----	59.23
Diff	F3,1080= 5.992 p<0.0005	F4,1082= 20.563 p<0.0001	F4,1498= 9.092 p<0.0001	F4,1502= 11.626 p<0.0001	N.S.	F4,1217= 3.181 p<0.013

*CEE scores obtained only from standard 4 year.

Table 4.1.18 shows that there were significant differences for the end-of-term and CEE scores between types of primary schools. There is no simple explanation for the differences, with the exception that the Senior schools produced consistently worse scores than the other types. Co-educational primary schools assigned higher average scores and produced better CEE scores than the other school types. Girl only primary schools assigned higher curriculum scores and produced higher CEE scores than equivalent boy only primary schools.

Pre-school attendance: Attending a pre-school was also found to be of significance in both end-of-term scores and CEE scores (average end-of-term F2.2134 = 23.829, p<0.0001; CEE F2.1705 = 54.805, p<0.0001). In both cases, those who attended pre-school had significantly higher scores than non-attenders and those who were unsure of attendance. It is worthwhile noting that pre-school attendance was significantly correlated to social class (rho= 0.2254, p<0.001); children were more likely to attend a pre-school if their parents worked in skilled and higher managerial positions (Table 4.1.19).

Table 4.1.19: Number and percentage of children that attended pre-school by grouped paternal occupation.

Pre-school Attend.	FATHER'S OCCUPATION			Row Total
	prof/ man	sales/ skilled	craft/ elem
no	95	80	280	455
	(19.3)	(27.7)	(36.0)	(29.2)
unsure	36	34	136	206
	(7.3)	(11.8)	(17.5)	(13.2)
yes	361	175	361	897
	(73.4)	(60.6)	(46.5)	(57.6)
Column	492	289	777	1558
Total	(31.6)	(18.5)	(49.9)	100.0)

Table 4.1.19 shows that a greater percentage of children whose fathers were from a professional or managerial occupation attended pre-school than other paternal occupations; the difference between the grouped paternal occupations was significant (X²(4) = 94.201, p<0.0001).

Not only were children from an educated background more likely to attend pre-school, girls were more likely than boys to attend a pre-school (Table 4.1.20). The difference between number of boys and girls attending pre-school was significant (X²(2)=6.317, p<0.04).

Table 4.1.20: Number and percentage of boys and girls that attended pre-school.

Pre-school Attend.	SEX		Row Total
	male	female
no	351	328	679
	(51.7)	(48.3)	(31.1)
	(33.4)	(29.0)
unsure	150	153	303
	(49.5)	(50.5)	(13.9)
	(14.3)	(13.5)
yes	549	650	1199
	(45.8)	(54.2)	(55.0)
	(52.3)	(57.5)
Column	1050	1131	2181
Total	(48.1)	(51.9)	(100.0)

With whom does the child live? Especially in the Caribbean, personal development, educational attainment and occupation have been linked to the immediate family of the child. Research in Jamaica (Miller, 1986) refers to the 'marginalisation' of the black male as an explanation for low attainment in school and low motivation to productively enter the work force. To provide insight into family support and structure all children in the survey were asked to identify with whom they lived. Children were most likely to live with both parents, and a majority of those not living with both parents lived with their mother (Table 4.1.21). Another way of expressing this information is that children lived overwhelmingly in households with their mother present (nearly 87% of children lived with both parents or mother only). Those living with father only was a small percentage, but brings the total number of children living with a natural parent (in the survey) to above 90%.

Table 4.1.21: With whom does the child live?

	Frequency	Percent
mother only	873	40.1
father only	76	3.5
both parents	1013	46.5
grandparents	101	4.6
other relatives	67	3.1
guardian	46	2.1
other	1	.0
	79	Missing
Total	2256	100.0

Simply identifying the parent with whom the child lives does not provide information as to how the living situation affects the school-based performance of the child. Table 4.1.22 shows distinct advantages for children who live with both parents as opposed to mother only or father only (confirmed with Scheffe post hoc analysis at the 0.05 level); and these results remained constant for both boys and girls as measured by their CEE, the standardised within-class scores and the raw within-class scores.

Table 4.1.22: Averages and differences on within-class and national scores for children living with mother, father and both parents

CHILDREN/ Scores	ALL CHILDREN	MALES ONLY	FEMALES ONLY
Common Entrance Live with:
mother	51.89	47.69	55.95
father	50.90	46.06	57.68
both	62.04	56.95	66.54
Significance	F2,1520= 35.883 p<0.0001	F2,735= 14.718 p<0.0001	F2,767= 21.054 p<0.0001
Standardised Live with:
mother	-0.1070	-0.2874	0.0672
father	-0.0238	-0.1294	0.1318
both	0.1149	-0.0477	0.2785
Significance	F2,1932= 11.893 p<0.0001	F2,937= 6.485 p<0.0016	F2,972= 5.963 p<0.0027
Raw Average Live with:
mother	57.64	54.27	60.84
father	59.27	55.38	65.23
both	63.51	60.56	66.44
Significance	F2,1932= 25.924 p<0.0001	F2,937= 14.256 p<0.0001	F2,972= 13.051 p<0.0001

As this project was undertaken in the Caribbean and researchers (see Drayton, 1995) have identified that children living with their mothers achieve well in school, a further analysis was undertaken to ascertain school achievement to assess the effects of living with father present. An initial analysis was undertaken for both girls and boys and explored achievement by the averaged raw classroom scores, standardised scores and CEE scores. The analysis also explored for differences within each of the core subjects. In each of these analyses, children who lived in a household with father present (either father only or father and mother) scored significantly higher than households without father present. Further analyses were undertaken to ascertain whether these findings characterised the attainment of boys and girls separately; findings replicated the significant results (above) for both boys and girls. It should be noted that father presence in the household was correlated with social class, and this confounds arguments asserting the positive effects of mother-dominated households in the Caribbean.

Who helps with homework? Aside from identifying with whom the child lives and the relationship of living with to school attainment, it is also important to note who works with the child to promote their school attainment. Over 73% of the children that responded to the question Who helps you with your homework?', stated that someone at home helped them (Table 4.1.23).

Table 4.1.23: Who helps the child with homework

Who helps	Percentage of children receiving help
Father	29.8%
Mother	50.3%
Brother	14.3%
Sister	18.2%
Other Relative	21.4%
Tutor	6.6%
Friend	20.2%
Guardian	7.4%

In line with whom the child was most likely to live with, results showed that the mother was the person who helped with homework most often. Parental help with homework was more likely to be found among parents from the higher levels of occupation.

Receiving help with homework was a general phenomenon. Only in some categories of help was there a relationship to the attainment of pupils. High attainers received proportionally more homework help from their fathers than mid and low attainers (X² (1)=10.187, p<0.006). Lower attaining children were more likely to receive homework help from brothers, sisters and friends. Homework help also varied with regard to the sex of the child. Fathers helped sons and daughters equally. Mothers gave slightly more help to sons than to daughters (X² (1)= 7.214; p<0.007) as shown in Table 4.1.24. Brothers, sisters and other relatives provided approximately equal amounts of help to boys and girls. Girls were more likely to receive help from their friends (X² (1)=9.447, p<0.002), and shown in Table 4.1.25.

Table 4.1.24: Help with homework by mother for son or daughter

HELP BY MOTHER	CHILD		Row Total
	son	daughter
no	503	599	1102
	(46.40	(52.1)
yes	581	551	1132
	(53.6)	(47.9)
Column	1084	1150	2234
Total	(48.5)	(51.5)	(100.0)

Table 4.1.25: Help with homework by friend for boys or girls

HELP BY FRIEND	SEX		Row Total
	male	female
no	892	886	1778
	(82.3)	(77.0)
yes	192	264	456
	(17.7)	(23.0)
Column	1084	1150	2234
Total	(48.5)	(51.5)	(100.0)

4. With the range of significant results provided, what is the relative contribution of each result to the overall performance of pupils and students in schools? This chapter has described significant differences in school attainment explained by sex, social class (occupation of father and of mother), preschool attendance, school type, with whom the child lives and who helps with homework. Stepwise multiple regressions were undertaken to ascertain the amount and significance of each of these variables in explaining standardised average end-of-term results, raw end-of-term results and CEE scores. These regressions have to be conducted separately for primary and secondary school students as patterns of school difference were not consistent among these age groups. A number of nominal variables had to be receded to allow for inclusion into the analysis.

Primary schools: Based on an imputed parametric distribution of data within the variables (sex of child, occupation of father, occupation of mother, type of primary school attended, pre-school attendance, with whom the child lives and who helps the child with homework), regressions were undertaken. Using CEE scores as the dependent variable over 25% of the variance (Adjusted R square) was explained; the predominant contributions to the variance were occupation of mother (18.5%), occupation of father (5.1%), and sex of pupil (1.5%). Using the standardised end-of-term score as the dependent variable (which excluded the type of primary school) only 9% of the variance was accounted for in the regression; the main contributions were occupation of mother (7.3%), occupation of father (1.6%). Finally, using the raw end-of-term score as the dependent variable (which allowed inclusion of type of primary school) 11% of the variance was accounted for in the regression; occupation of mother (7.3%), occupation of father (1.9%), with whom the child lives (0.8%) and sex of pupil (0.8%). All of these regressions show that success on national examinations and within-class assessment was not simply explained by girls achieving higher scores than the boys. The most significant amount of variance in these regressions was explained by parental occupations (especially mother) and this was followed by sex of the child and with whom the child lived. Figure 4.1.1 displays how the correlations between these significant variables contribute to classroom attainment.

Figure 4.1.1: Correlations between key regression variables affecting primary school attainment

A profile of the most consistently successful child in primary school would be a girl whose mother and father (both) worked in professional/managerial positions and who lived with both parents.

Secondary schools: Regressions undertaken at the secondary level showed a different picture than the primary schools - especially because of the stratification of secondary schools and the different social class profile of the distinct types of secondary school. Variables used in these stepwise regressions included sex of child, occupation of father, occupation of mother, type of secondary school attended, pre-school attendance, with whom the child lived and who helped the child with homework. Using CEE scores as the dependent variable a massive 56.5% of the variance was accounted for; the main contributions to this variance were type of school - sixth form or not (44%), occupation of mother (12.2%), occupation of father (2%), sex of student (0.7%), and with whom the child lives (0.6%). Using the standardised end-of-term scores as the dependent variable, the explanation of variance was much more limited: only 6.8% of the variance was accounted for - and the sole contribution was made by the sex of the student. Finally, the raw end-of-term scores were used as the dependent variable and 14.5% of the variance was accounted for; the main contributions to the variance were type of secondary school (7.5%), attendance of pre-school (3.1%), sex of student (2.9%) and occupation of father (1.1%). Figure 4.1.2 displays how the correlations between these significant variables contribute to classroom attainment.

Figure 4.1.2: Correlations between key regression variables affecting secondary school attainment

A profile of the successful student at the secondary level shows a female who is attending a sixth form school, whose parents work in high managerial/professional positions, who lives with both parents and who attended a pre-school.

Regressions for both primary pupils and secondary students showed that academic attainment was not simply explained by the sex of the child. Occupation of mother and father, whether the child lived with one or both parents, type of school attended and pre-school experience were all likely to have a strong effect on the success of the child.

Choice of schools for ethnographic case studies

Results of the quantitative survey in Barbados led to an agreement that much more insight into the workings of achievement needed to be undertaken within classrooms. Case studies would allow us to focus on particular schools and students to ascertain how achievement is affected by sex of student, social class background and type of school. Within the funding limits of this project, it was agreed that the case studies would be limited to secondary schools. From the preliminary quantitative analyses in Barbados, the schools within which the case studies would take place include:

a) the female-only school which had average CEE scores but low end-of-term scores, noting that this school had a comparatively low social class intake; and

b) a sixth form school where there were no significant differences between boys and girls in core curriculum or CEE scores, yet there were more girls in the higher forms and a generally higher social class intake.

Within the limits of this project, these school-based case studies could not be as extensive as those reported from Trinidad. Case study researchers were only allowed part-time observations for periods between two and four months within the case schools. Reports of their findings are taken directly from their notes and, thus, could not be organised and interrogated as had been undertaken in the Trinidad studies.