Note
1. Level of significance *p<.05 ** p<.01
2. AV is the average effect obtained from Meta-Analysis
below the mean approximately 9 per cent in Orissa (-.094) and 6 per cent in Tamil Nadu (-.066), whereas boys (.056) tend to score approximately 5.6 per cent above the mean in both the states.
For 10 years of teacher's schooling, the girls and boys respectively score below the mean 21 per cent and 6 per cent in Orissa and 26 per cent and 14 per cent in Tamil Nadu. For 14 years of teacher's schooling, the girls tend to achieve approximately 7 per cent in Orissa and 19 per cent in Tamil Nadu above the mean, whereas, boys tend to score approximately 22 per cent in Orissa and 31 per cent in Tamil Nadu above the mean. Though both the boys and girls are benefited by the higher qualification of teachers. However, the LOWESS regression curve for Orissa (Fig. 5) shows that boys gain more than girls.
In mathematics, teacher's qualification is negatively associated (-.158, p<.01) with the gender gap in Kerala. The girl-TCHQUAL slope (-.433) is very steep for the teachers having 14 years of schooling as compared to the slope (.004) of those teachers who have 10 years of schooling. Almost similar trend is found for boys. It indicates that students tend to score lower than average score, if teachers with higher qualification are appointed in the primary schools.
The meta-analysis (Table 6) shows that, in language, the average slope of the variable stability of teachers is very weak (-.002). This indicates that continuous stay of teachers in a school is not significantly associated with the language achievement of students.
38 IER: SPECIAL NUMBER 1995
FIG. 5 Effect of teachers qualification on gender gap in language achievement (Orissa)
In Orissa, the variable is negatively associated (-.042, p<01) with the gender gap. The girls score 19 per cent of standard deviation of achievement below the mean provided teacher stays in the school for a period of 7 years. Whereas, the girls score near to the average if the service period of teacher is only 3 years. Similar trend has been observed in case of boys. This indicates that stay of a teacher beyond the state average (5 years in Orissa) in a school may affect adversely on the learning achievement of children.
The teacher's experience and the in-service training provided to the teachers are not associated with gender gap and therefore do not make significant effect in the learning achievement of children in language
In Karnataka, teachers experience has negative association with the gender gap in both the mathematics (-.012, p<.05) and language (-.016, p<.01 achievement. Whereas, in service training has a marginal positive impact on the learning achievement of language in Madhya Pradesh and Kerala. It may be noted that the average experience of teachers is more than 12 years in each state. Except in Assam and Orissa, more than 50 per cent of teachers have received in- service training.
STUDIES ON DPEP 39
FIG. 6 Effect of teachers on gender gap in language achievement (Karnataka)
EFFECT OF SCHOOL RESOURCES For mathematics, the variable access to teaching material has positive association (.095,p<.05) in Assam (Table 7) and negative association (-.038) in Karnataka. Instructional material used for teaching of mathematics in the classroom has negative association with the gender gap in Assam (-.079,p<.05) and Haryana (-.026) and no significant association is found in other states for both the subjects. Out of 12 items, on an average. 6 items are available in the states. It seems either the educational materials are not in good condition or not properly used by the teacher.
Physical facility available in the school has a positive association with the gender gap in Karnataka, whereas, in Assam it has negative association on both the subjects. In Orissa, the facility is negatively associated (-.057, p<.05) with language achievement. The meta-analysis shows that the variable do not have significant impact on the learning achievement of students. Out of 14 items, on an average, 4 to 8 items are available in the school.
In language achievement, meta-analysis (Table 7) shows that, on an average, number of teachers in a class has negative association (-.021, p<.05)
40 IER : SPECIAL NUMBER 1995
Note
1. Level of significance * p<.05; **p<.01
2. A V is the average effect obtained from Meta-Analysis
with the gender gap in Haryana, Karnataka, Madhya Pradesh and Tamil Nadu. In Karnataka, the predicted NUMTCH-girl slope is 0.076 for the two teachers, -.105 for average number of teachers and -.159 for seven teachers, indicating that large number of teachers in a school has a negative effect on the learning achievement of girls. Similar trend is found for boys also. On an average, from three teachers in Assam, Madhya Pradesh and Maharashtra to 10 teachers in Kerala are working in a school. One can see if the number of teachers exceed the state average, the students achieve lower than what they achieve with fewer teachers. Graph for Karnataka (Fig. 6) shows the similar trend for the effect of number of teachers in a school on language achievement. Pupil-teacher ratio has a negligible impact on the learning achievement of both the subjects,
STUDIES ON DPEP 41
For language achievement, schools having only primary sections has a negative association with the gender gap in language achievement in Haryana (-.262, p<.05), Maharashtra (-.183) and Orissa (-.155). The average effect (-.200) is significant at 5 per cent level of significance. The predicted girl-primary achievement slope shows that girls studying in primary schools of Haryana have a weak slope (-.085) than those girls studying in extended primary schools with steep slope (.180). Almost similar trend is found for boys. Similar observations are made in Orissa also. In Orissa. LOWESS regression curve has shown that in fact girls tend to perform better than boys in extended primary schools. In mathematics achievement, similar observations are made for Orissa also, However, in Assam, girl slope (.273) is not significantly associated with the gender gap.
42 IER: SPECIAL NUMBER 1995
Percentage of female teachers in a school has a weak association with the gender gap on both the subjects. In Orissa, with average 20 per cent of female teachers, girls score 10 per cent below the mean and boys score 6 per cent above the mean. However, if the percentage of teachers are increased to 50 per cent the girls and boys tend to score 1.6 per cent and 18 per cent respectively above the mean. The LOWESS regression curve (Fig. 7) has shown that girls tend to score higher than boys.
FIG. 7 Effect of female teachers on gender gap in language achievement (Orissa)
EFFECT OF ACADEMIC CLIMATE OF SCHOOL The academic activity in teaching of mathematics is negatively associated (Table 8) with the gender gap in Haryana, Madhya Pradesh, Maharashtra and Orissa. indicating, on an average, girls score lower than boys. The meta-analysis also shows that the average effect size (-.215) is significant at 5 per cent level of significance. The effect of the variable is smallest (-.084) in Maharashtra and largest (-.433) in Orissa. We consider two extreme cases (i) when teacher does not assign any problem to solve and (ii) when teacher assigns problems regularly to predict the achievement slope for boys and girls. In Orissa, for first case, girls and boys score approximately 48 per cent and 63 per cent
STUDIES ON DPEP 43
FIG. 8 Effect of teachers commitment on gender gap in mathematics achievement (Tamil Nadu)
respectively above the mean. Whereas, girls and boys score approximately 38 per cent and 23 per cent below the mean. In Maharashtra, for the first case, girls and boys score approximately 8.7 per cent and 14.4 per cent above the mean, whereas for the second case, girls score 88 per cent below the mean and boys score 2.4 per cent above the mean. In both the states, students, in particular girls, are losers if the teacher assigns mathematics problems regularly in the class. Shukla (1994) has also indicated in her study that teacher's knowledge of the subject-matter was highly inadequate.
As regards academic activity of teaching of language, only state of Assam has a negative association (-.537) with the gender gap.
Academic activity that teacher assigns homework in mathematics has positive association in Assam (.226) and Karnataka (.168) and has negative association (-.275) in Tamil Nadu. In language, the variable has a positive association (.287, p<.05) in Assam only. One can predict the score in two extreme situations, i.e., (i) if teacher does not assign the homework, the students score below the mean and (ii) if he/she assigns regularly, students score above the mean. Whereas, the situation is reverse in Tamil Nadu. Similar pattern has been observed in case of teacher's commitment in teaching of mathematics, i.e., if he/she pays attention towards the class, the
44 IER: SPECIAL NUMBER 1995
students tend to, score higher. A graph showing effect of teacher's commitment on achievement of mathematics is presented in Fig. 8.
For mathematics, parent's involvement, i.e., parents attend the parent-teacher meetings and meet teachers individually to discuss the performance of child' has positive association in Assam (.142) and negative association in Haryana (-.092) and Karnataka (-.091).
The leadership of the headmaster has a positive association (.100, p<.05) with teaching of language. It indicates that if teacher meets the headmaster to review the performance of his class and all primary classes of the school, improves the achievement of the children by 10 per cent of standard deviation above the mean.
Summary of the Findings The analysis examined a sample of 1,746 schools, comprising 4,879 teachers and 23,700 students collected under Baseline Achievement Study in 1993 to study the achievement gap between boys and girls. The results from test and HLM analysis revealed that achievement gap exists between boys and girls. Further, it shows that the achievement gap increases in both the subjects after controlling the pupil level variable in most of the states and it has negative association with the achievement. Kerala is the exception, where the gap has decreased marginally in language. It is also estimated that girls score is approximately 12.2 per cent and 11.3 per cent standard deviation lower than boys in mathematics and language achievement respectively. The achievement gap in mathematics is larger than language.
The results of second level of HLM analysis answer the second research question, i.e., to what extent the school level factors are associated with the gender gap. It shows that the higher qualification of teachers helps in improving the achievement of students to some extent. However, students, in particular girls, tend to perform higher provided teacher's with graduate and above qualification are not appointed at the primary stage. Further, continuous stay of teachers for not more than five years in the same school is likely to improve performance of girls. The teachers do not have the access to teaching material or it has not made any discernible impact on the learning achievement of children, particularly on the gender achievement gap. Large number of teachers in a school has a negative effect on the achievement of children.
The students perform better in the extended primary schools than primary schools. The extended primary schools, i.e., middle and secondary schools are generally available at block and district level. Hence, they attract good students and thus the peer group effect on the learning achievement of children. The presence of more than 50 per cent female teachers in a school is likely to
STUDIES ON DPEP 45
improve the performance of girls. it is possible that the girls feel more homely with the presence of female teachers in the school.
The frequent assignment of mathematics problems is negatively associated with the achievement gap between boys and girls. However, teacher's commitment and leadership of headmaster has made a positive impact on the learning achievement of children.
EDUCATIONAL IMPLICATIONS The above finding bring out certain implications for school based intervention on teacher quality, use of educational facilities in school, academic press, school climate and school policies.
The variables studies under teacher quality revealed that teacher with average qualification (prescribed for primary school teacher) tend to reduce the achievement gap between boys and girls. Perhaps teacher with higher qualification may seek the job of primary teacher but not committed to teaching at this level. So this fact may be considered at the time of appointing teacher at this level. On the other hand longer duration of stay in the particular school enlarge the gap between boys and girls on achievement. It may be due to monotony and boredom. Training programme should be arranged-to equip the teacher with innovating teaching strategies, enhance their motivation and nurture commitment.
It is assumed that facilities are prerequisite of better learning. The study highlighted that, though- the facilities are available in the school but not affecting the learning performance of girls. Therefore, strict supervision should be arranged to see the usability condition of the materials and its constant use in teaching learning process. Also the teacher should be trained in the effective use of those materials.
Due to pressure of academic activity in mathematics, probably girls are not able to cope with the task and become remotivated. This may lead to lower achievement. Therefore, remedial classes can be arranged to overcome their deficiencies in mathematics through suitable teaching strategies.
With reference to policy implication, the findings. suggest that girls are performing better in the primary school which is attached to middle and secondary sections. It may be because of wide variety of experiences, opportunities and interaction with peer groups. The extension of primary school with higher grade may not be only beneficial for girls but also for other students. Second aspect emphasize the benefit of female teacher in school. It shows that girls are performing better in school with higher percentage of female teacher. This finding support the principle of Operation Blackboard scheme where female teachers were provided in the school.
46 IER : SPECIAL NUMBER 1995
ACKNOWLEDGEMENTS Authors gratefully record the contribution of Prof. J.D. Willms, UBC, Canada for intruding the Hierarchical Linear Model approach and providing the necessary guidance in preparation of this paper.
We are thankful to the NCERT and NIEPA faculty members who shared BAS data and Shri M. Arif Sadiq, Programmer for providing technical support in analysis of data.
Authors thank Dr. K.B. Rath and other faculty members of the group for giving their valuable suggestions to bring this paper in this form.
REFERENCES Bryk, A.S. & Raudenbush, S.W. (1992). Hierarchical Linear Models for Social and Behavioral Research; Application and Data Analysis Methods. Newbury Park, CA : Sage.
Buch, M.B. (1991). Fourth Survey of Research in Education, 1983-88. National Council of Educational Research and Training, New Delhi.
Dave, P.N. & others (1988). Achievement Under Project PECR (Mimeographed), National Council of Educational Research and Training, New Delhi.
Ministry of Human Resource Development, Department of Education, Government of India. (1993) : Education for All - The Indian Scene.
National Council of Educational Research and Training (1994) : Research Based Interventions in Primary Education - The DPEP Strategy.
Kulkarni. S.S. & others (1970). All-Indian Survey of Achievement in Mathematics, National Council of Educational Research and Training, New Delhi.
Shukla, S. & others (1994). Attainments of Primary School Children in Various States, National Council of Educational Research and Training, New Delhi.
STUDIES ON DPEP 47
ANNEXURE I
Indicators used in the Study
S.No. Label Indicator Treatment/Remarks
I. Pupil's Background Variable
1. GIRL Gender Girl = 1, Boy = 0 Centered at Pupil
(SS 3) Level
2. SCST Caste SC or ST = 1, Non-SC/ST = 0 Centred
(SS 6) at Pupil Level
3. DADED Father Education Applied Logit
(SS 9)
4. MUMED (SS9) Mother Education Applied Logit
5. DACOCC Father Unskilled worker = 1; Poultry
(SS 10 Occupation farming = 2; Picking forest produce
= 3; Agri, labourer = 4; Farmer 5;
Skilled worker = 6; Street vendor =
7; Others 8; Self employed = 9;
Domestic servant 10; Household =11;
Clerk = 12; Employer = 13 Senior
Officer = 14 Applied Logit
6. REPEAT Ever Repeated Repeat = 1 if yes in any class Else
(SS H14) Class = 0 Centre at Pupil Level
7. SES Socio-economic SES = (DADED + MUMED + DADOCC)
Status /Valid responses Composite &
standardised
II. Contestual Variables
1. MEANSES School Mean SES Aggregated from pupil to school
level Centered at School Level
2.PCTSCST Per cent SCST Per cent of SC and ST in the school
Centered at School Level
III. Teacher's Quality Variables
1. TECHQUAL Teacher's Class VIII = 8, Class X = 10,
(TS 4) Qualification class XI/XII = 12. graduate = 14,
post graduate =16 Aggregated &
centered at school level
2. TCHEXP Experience Subtracted year of first
appointment from 1993
(TS 10B) Aggregated & centered at school
level
3. INSERVICE In-service INSERVICE = 1, if received during
(TS 13) training last 3 years Aggregated & centered
at school level
4. STABLE Period in present Subtracted year of appointment
(TS 10C) School (from 1993) Aggregated & centered
at school level