Quantitative Data Analysis - Essay Example

?Quantitative Data Analysis Introduction: Cognitive ability of a grown-up person is a very vital factor which plays an important role in the determination of wages as well as outcomes form the labor market. Hence, many researchers in the area of social science have asked for implementing alternative policies for improving the cognitive attainment of underprivileged people since their childhood. One probable policy is to support financially poor families so that they can offer better learning environments for their children. The present study will make an effort to examine the factors that determine cognitive ability of individuals in their childhood. There are a number of factors that can be expected to have important role in determining the cognitive ability of individuals in their childhood. These factors include various characteristics of a child including gender, birth weight, breast feeding, etc., parental factors like parent qualification, parents’ social class etc. socioeconomic issues of family including ethnicity, religion, income, housing etc., other family related issues like childcare, bedtime, ability to read, etc. and geographical factors. This study will make an attempt to find which of these factors have significant effect on cognitive ability of children in UK. The thesis statement for this research can be stated as follows: Various children specific, parental, family specific and geographic factors affect cognitive ability of individuals at early age. Against this thesis statement the following research questions can be made: How do children specific factors like gender, birth weight, breast feeding etc affect the cognitive ability of a child? How do various parental factors like parents’ qualification, social class, etc. affect the cognitive ability of a child? How do socioeconomic factors related to a family, like ethnicity, religion, income, housing, etc. affect cognitive ability of a child? How do family related factors like bedtime, ability to read, childcare, etc. affect cognitive ability of a child? How do geographical factors affect cognitive ability of a child? Given these research questions, the following hypotheses will be tested in this quantitative study. Hypotheses related to first research question: Null hypothesis: Children specific factors like gender, birth weight, breast feeding etc do not significantly affect the cognitive ability of a child Alternative hypothesis: children specific factors like gender, birth weight, breast feeding etc significantly affect the cognitive ability of a child. Hypotheses related to second research question: Null hypothesis: Various parental factors like parents’ qualification, social class, etc. do not significantly affect the cognitive ability of a child. Alternative hypothesis: Various parental factors like parents’ qualification, social class, etc significantly affect the cognitive ability of a child. Hypotheses related to third research question: Null hypothesis: Socioeconomic factors related to a family, like ethnicity, religion, income, housing, etc. do not significantly affect cognitive ability of a child. Alternative hypothesis: Socioeconomic factors related to a family, like ethnicity, religion, income, housing, etc. significantly affect cognitive ability of a child. Hypotheses related to fourth research question: Null hypothesis: Family related factors like bedtime, ability to read, childcare, etc. do not significantly affect cognitive ability of a child. Alternative hypothesis: Family related factors like bedtime, ability to read, childcare, etc. significantly affect cognitive ability of a child. Hypotheses related to fifth research question: Null hypothesis: Geographical factors do not significantly affect cognitive ability of a child. Alternative hypothesis: Geographical factors significantly affect cognitive ability of a child. In order to test these hypotheses a statistical test of regression will be conducted. The data and method for the test will be discussed in the next section. Data and Method: Sample data: For the purpose of this research, the data selected are a subset of data from the Millenium Cohort Study (MCS) and is a random sample of 12888 children born in the UK in 2000/1. The sample population for the study was collected from all live births in the UK over 12 months from 1 September 2000 in England & Wales and 1 December 2000 in Scotland & Northern Ireland. The sample was selected from a random sample of electoral wards, disproportionately stratified to ensure adequate representation of all four UK countries, deprived areas and areas with high concentrations of Black and Asian families. The study has been tracking the Millennium children through their early childhood years. The four surveys of MCS cohort members carried out so far – at age nine months, three, five and seven years – have built up a uniquely detailed portrait of the children of the new century. For the present analysis, this study will be looking at cognitive assessment of children at age of 5. British Ability Scales (BAS) were used to assess children cognitive ability. A summative test score adjusted for age and standardized. The summative score includes three components: Naming Vocabulary, Construction Pattern and Picture Similarities. Method: For the purpose of the present study quantitative research method will be applied. Quantitative research is conducted on the basis of fundamental hypotheses where various statistical processes are applied to test the hypotheses (Glitz, 1997). The quantitative research method is used to describe variables and examine relationships among those variables (Creswell, 2001). In this case quantitative research will be conducted to find out the relationship between cognitive ability of children and various types of children specific, parental, family socio-economic, and geographical factors. (Creswell, 1994) To get in depth view regarding the pattern of relationships of the dependent variable, i.e. cognitive ability indicated here in terms of summative scores, and a range of independent variables; i.e. characteristics of a child including gender, birth weight, breast feeding, etc., parental factors like parent qualification, parents’ social class etc. socioeconomic issues of family including ethnicity, religion, income, housing etc., other family related issues like childcare, bedtime, ability to read, etc. and geographical factors, a multiple regression can be run. To run a multiple regression with the help of the data at hand, the following regression equation has to be estimated- Y = ? + ?1 Xch + ?2 Xpar + ?3 Xreg + u ……………………………………………………………………………….(1) In the above equation, the variable Y is cognitive ability measurement, Xch is the matrix of the independent factors relating to child’s characteristics, such as gender, birth weight etc., Xpar is the matrix containing parental and family socioeconomic and other family relted variables and Xreg are geographical variables (rural vs urban, country etc.). ?1 is the coefficient matrix of the variable matrix of children characteristics. The sign of this coefficient represents the direction of the effect of independent variable on dependent variable, while its magnitude reflects the extent by which dependent variable is affected by independent variable. ?2 is the coefficient of the variable matrix of parental and family related factors. Similarly, ?3 is the coefficient of the variable matrix of geographicl factors. Equation 1 is a multiple regression model. Before running regression, it is necessary to make certain assumptions. The assumptions that are required to be made here are as follows: The error factors in the equation (1), i.e. ut have zero mean. There is a common variance of the error factors. So it be written that V(U) = ?2 for all t. The error terms are independent of one another. Every independent variable is self-determining. It simply means that E(Xch U) = 0, E(Xreg U) = 0 and E(Xpar U) = 0. This assumption will follow instantly if the explanatory variables are considered to be nonrandom. This follow the statement that the allocation of error terms is not dependent of the distribution of any independent factors in the regression equation. The error term is additionally considered to be normally allocated. In combination with the first set of assumptions relating to the error terms, it can be stated that errors are autonomously and normally allocated with zero mean and a fixed variance. (Berk, 103-105) Result: In this section the results of the multivariate regression analysis will be presented. Table 1 presents the results for the first set of explanatory variables, i.e. children’s characteristics: Table 1: Explanatory Variables Unstandardized Coefficients Standardized Coefficients t Sig. gender 2.621 0.595 0.055 4.402 0 Birth weight low -6.825 2.065 -0.041 -3.305 0.001 Birth weight medium -2.612 0.753 -0.044 -3.468 0.001 Birth weight high 1.244 0.961 0.016 1.294 0.196 breastfeeding_3mths 1.452 0.815 0.029 1.782 0.075 breastfeeding_3_6mths 4.335 0.971 0.074 4.464 0 breastfeeding_6mths 4.051 1.025 0.066 3.952 0 In the table above it is found that t statistics for the coefficients of gender, low and medium birth weight, breast feeding at 3-6 months and 6 months are statically significant as the P value of the t statistics for the coefficients of these variables are less than 0.05. Table 2 presents the results for the second set of explanatory variables, i.e. parental and family related factors Table 2: Explanatory variables Unstandardized Coefficients Standardized Coefficients t Sig. parents qualification none -10.967 1.997 -0.098 -5.493 0 parents qualification overseas -10.678 2.943 -0.047 -3.629 0 parents qualification nvq1 -9.72 1.963 -0.083 -4.95 0 parents qualification nvq2 -6.537 1.344 -0.112 -4.865 0 parents qualification nvq3 -3.263 0.897 -0.052 -3.637 0 parents qualification nvq5 -2.166 1.108 -0.044 -1.955 0.051 parent level social class never worked -3.278 2.297 -0.019 -1.427 0.154 parent level social class prof/manage 4.049 1.2 0.083 3.376 0.001 parent level social class intermediate 1.193 1.28 0.017 0.932 0.351 parent level social class emp -0.42 1.576 -0.004 -0.267 0.79 parent level social class tech -0.995 1.425 -0.011 -0.698 0.485 ethnicity not known 1.905 3.965 0.008 0.48 0.631 ethnicity mixed -1.889 4.427 -0.007 -0.427 0.67 ethnicity indian 5.446 3.649 0.029 1.492 0.136 ethnicity pakistani -7.061 2.957 -0.047 -2.388 0.017 ethnicity blk -9.385 2.609 -0.047 -3.597 0 ethnicity others 4.422 3.057 0.039 1.446 0.148 religion protestant 0.944 0.742 0.018 1.273 0.203 religion catholic 0.78 1.016 0.011 0.767 0.443 religion other christian -5.54E-02 1.043 -0.001 -0.053 0.958 religion other religion -8.347 2.304 -0.084 -3.623 0 income poor above at both 1.968 2.281 0.04 0.863 0.388 income poor below at both 0.325 2.468 0.004 0.132 0.895 income poor missing 0.311 2.298 0.006 0.135 0.892 housing own 3.271 2.038 0.028 1.605 0.108 housing rent LA or HA -3.906 1.022 -0.062 -3.822 0 housing rent privately -1.614 1.824 -0.017 -0.885 0.376 housing other -2.067 1.528 -0.031 -1.353 0.176 childcare none -7.43E-03 0.856 0 -0.009 0.993 childcare other informal -1.354 1.802 -0.01 -0.751 0.452 childcare miss -0.848 0.768 -0.017 -1.105 0.269 bedtime always -2.335 0.805 -0.038 -2.9 0.004 dummy read missing -3.508 4.358 -0.01 -0.805 0.421 dummy read yes 3.168 0.695 0.061 4.556 0 In the table above it is found that among parental factors t statistics for the coefficients of parent qualification, and among family’s socio-economic factors ethnicity in respect of Pakistani and Black are statically significant as the P value of the t statistics for the coefficients of these variables are less than 0.05. On the other hand, among family related other factors t statistics for the coefficients of housing rent, bed time always, ability to read are statically significant as the P value of the t statistics for the coefficients of these variables are less than 0.05. Table 3 presents the results for the third set of explanatory variables, i.e. geographical factors Table 3: Explanatory variables Unstandardized Coefficients Standardized Coefficients t Sig. england dis 0.199 0.788 0.004 0.252 0.801 england ethnic -0.788 1.946 -0.006 -0.405 0.686 wales dis -1.627 2.764 -0.01 -0.588 0.556 scotland adv -4.61 2.03 -0.043 -2.271 0.023 scotland dis -8.339 2.287 -0.061 -3.646 0 NI adv 0.679 1.974 0.007 0.344 0.731 country england -4.218 1.66 -0.074 -2.541 0.011 country wales -3.579 2.463 -0.031 -1.453 0.146 town -0.48 1.215 -0.006 -0.395 0.693 village -0.689 0.94 -0.012 -0.733 0.464 In the above table, among the geographical factors, only t statistics for Scotland’s disadvantageous group and Country Wales are statistically significant as their values are less than 0.05. Discussion: On the basis of the above results, the following can be inferred: First, the cognitive ability of children is positively and significantly affected by gender. The cognitive ability is negatively and significantly affected by birth weights and positively and significantly affected by breast feeding in the sense that with lower birth weight cognitive ability falls and more breast feeding cognitive ability rises. Thus the first null hypothesis can not be accepted altogether as various children specific characteristics have significant effect on children’s cognitive ability. Second, in case of parental factors, parental qualification has negative and significant effect in the sense that as parental qualification rises children’s cognitive ability falls. This is against the normal expectation. On the other hand parental social class does not have any significant influence on children’s cognitive ability. In case of family’s socio economic factors, it is found that only in case of Pakistani and Black ethnicity, cognitive ability is negatively associated with the ethnicity in the sense that cognitive ability falls for Pakistani and Black people. Among other family related other housing rent, bed time always, ability to read significantly and positively affect cognitive ability of children in the sense that the family that goes to bed early and able to read are more probable to have children with high cognitive ability. Finally, third, the geographical factors do not really very important in deciding cognitive ability of children. Conclusion: The present study has attempted to find out he determinant of children’s cognitive ability. For this purpose it has conducted quantitative study using a large scale of data. On the basis of the discussion above, it can be concluded that the cognitive ability of individuals at early age is very well shaped by a number of factors. Among these factors some are children specific characteristics, while some are parental and family’s socioeconomic and some other family related factors. Among various factors the most influential factors are gender, breast feeding, birth weight, parents’ qualification, ethnicity of family, habit of going to bed, ability to read etc. A very important thing to be noted here is that at the time of the study a huge range of factors have been taken in to account under each set of factors considered in the multivariate regression analysis. For example, children specific features included gender, birth weight, and breast feeding, parental factors like parent qualification and parents’ social class, socioeconomic issues of family included ethnicity, religion, income, and housing other family related issues included childcare, bedtime, and ability to read, and there were various geographical factors as well. But at the end of the study it was found that only a few among this large range of explanatory variables affected children cognitive ability. But, since each group, some of the factors have been found to have produced significant effects on children’s cognitive ability, none of the null hypotheses pf this study can be rejected altogether. References: Glitz B. 1997. The Focus Group Technique in Library Research, An Introduction. Bulletin of the Medical Library Association, 85(4), 385-390. Creswell, J. W. 2001 Educational research: Planning, conducting and evaluating quantitative and qualitative research. OH: Merrill/Prentice-Hall. Creswell, J. W. 1994. Research design: Qualitative, quantitative approaches. Sage Publications, Inc. Thousand Oaks: CA Read More