| Ever since the controversial findings of the Coleman Report,a vast amount of literature have been published,albeit with contrasting results,there has been a never-ending debate on whether environmental factors have a large effect on student achievement,or rather whether it was student and family level factors that had a more significant effect on student achievement.As the years have passed,researchers drifted away from the American data.This led to the introduction of educational production function research in developing countries,and the introduction of cross-country comparisons between and within developed and developing countries.Additionally,as educational achievement research became popular overtime,more sophisticated methodology and research design has been introduced.The OLS estimation technique has come under intense scrutiny.Better methodologies started to proliferate in the 80s,and hierarchical linear modeling started dominating in many papers.The argument for the adoption of such models was due to the nature of education data,which is nested or clustered in nature.Educational data was seen to violate a critical assumption of OLS,that of observational independence.That is,individuals clustered within one group are likely to be more similar to other members of that group than to individuals clustered within another group.This resulted in the estimation of less efficient parameter estimates.Surprisingly,the change in data,country contexts,methodologies and time frames has not brought an end to the debate on what better explains student achievement.Rather,all these innovations have added more to the contrasting evidence.This review shows that there is still a need to ascertain what better explains student achievement,especially in developing countries where there has been limited research in this area.In a nutshell,the reviewed literature shows no conclusive evidence on which variables better explain student achievement the world over.In addition,the strength and direction of the relationships between factors and performance has appeared to be inconsistent across studies.It has also shown that it is not possible to assume that results obtained in one country apply in another country without empirically establishing such assumption.According to the Organization for Economic Co-operation and Development(OECD),Shanghai-China and Singapore were amongst the first and second top performing countries and economies.The position of Indonesia is lower than Malaysia,even though both were at the bottom line.Thus,this study not only analyzed the inadequacy of mathematics content in Indonesia and Malaysia compared to Shanghai-China and Singapore,but also described some lessons that can be drawn in relation to the content tested in the Program for international student assessment(PISA)mathematics literacy test from the selected Asia countries and area.PISA uses a two-stage stratified sampling procedure.In the first stage of sampling,schools having age-eligible students were sampled systematically with probabilities proportional to the school size,which is a function of the number of eligible students enrolled.A minimum of 150 schools was selected in each country.This was followed by randomly selecting a number of students around 15 years of age in the selected schools.The student sample from Indonesia,Malaysia,Shanghai and Singapore were 5622,5197,5177 and 5546.The data used for the analysis was obtained from the official PISA website.The data were analyzed using descriptive statistics,and a hierarchical linear modeling(HLM)approached with the HLM version 7.0 computer programme.The HLM analysis started with the baseline model where none of the level-1 or level-2 explanatory variable was included.Results from this model suggest that considerable differences in total average school mathematics performance existed across countries.Similarly,the intra-class correlation(ICC)was found different from one country to another.In Indonesia and Shanghai,all of the student background variables were statistically significant predictors of math performance except for preschool education attendance for one year or less;whereas in Singapore only students,family economic,social and cultural status,and preschool education attendance for more than one year showed statistically significant relationships with math scores.Similarly,the relationship between math performance and gender,and preschool education attendance for one year or less seemed to vary significantly across schools in Singapore.Meanwhile,in Malaysia only gender seemed to vary significantly across schools.The relationship between math performance and quality of school educational resources were statistically significant across schools in Indonesia,Malaysia and Shanghai,except for Singapore that appeared to vary significantly across schools.Findings from the present study show that there is significant relationship between teacher-student ratio and mathematics achievement,in all the selected Asia countries and area.The results suggested that this model worked differently across the selected countries and area.Overall,the empirical analysis found that student-level predictors had higher significant coefficients to some extent,which led to the inference that at least for all the selected Asia countries and area under study,student-level factors better explain their mathematics achievement scores.However,this does not mean that other factors should be disaffirmed.It should be recalled that Shanghai and Singapore came 1st and 2nd respectively,while Malaysia came 52nd and Indonesia 64th among the 65 countries and area that participated in PISA 2012.Although system-level variables were not examined in this study,future research studies may apply a 3-level HLM analysis including various countries in order to explore whether country differences in terms of mathematics performance can be interpreted with respect to differences in educational systems,policies and practices.More so,future study can be conducted using different existing large scale international achievement data such as TIMSS,SACMEQ,and PIRLS.Finally,because the current study did not explain why certain relations between math performance,contextual and background factors were present or to no avail,further study can be conducted within each country to gain deeper understanding of the reason underlying these relationships. |
PDF Full Text Request