Research Question and Statement of Purpose | 034 011, Papers of Introduction to Sociology

Download Research Question and Statement of Purpose | 034 011 and more Papers Introduction to Sociology in PDF only on Docsity! Jamie L. Lee 034:011 Baller Spring 2006 Research Question and Statement of Purpose Research Questions: Due to my interest in future employment in the United States school system, I selected a research question pertaining to the differences in public and private school students’ math, reading, and science test scores to conclude whether or not competition is necessary for schools. I also wanted to study the relationship of teacher performance (student performance) regarding rational choice theory. My research questions ask: Is competition necessary for school systems? And also, do male students benefit more from private schooling than females? These research questions require some explanation of what exactly is being studied. Competition between school systems questions if private schools or public schools are better in terms of student performance. If the private schools produce higher test scores than the public schools, competition would be necessary for schools, but if private schools do not produce higher test scores than the public schools, competition would not be necessary for the school systems. The independent variable is the type of school: private or public. The dependent variable in the study is students’ math, reading and science test scores. There will be a comparison of test scores between private and public schools. These scores can be regarded as the determining factor of the success of a private or public school. The interaction variable I decided to focus on was gender. I question if the relationship between private schooling and test scores is stronger for one gender over the other. Males may benefit more from private schools than females if they are monitored more than at public schools. (Monitoring refers to the ability to guide and watch over the student at a more intimate 1 level). There may also be the similar possibility that males in private schools are expected to join in more activities. Better grades and high performance may result from the higher expectations that may be placed on males in private schools. I anticipate that among male students, the beneficial impacts of private schooling should be greater than they are for females. If neither gender shows a positive relationship to benefits from private schooling, then competition is not necessary for the school system. Statement of Purpose: My study of the relationship between school sector, student test scores and gender is important because of both policy and theoretical reasons. The association of these three variables is important for United States’ policy because if competition is necessary in schools, current policies regarding schools will have to change. The government may have to consider making public schools more competitive in teacher salaries and expectations so that students are receiving the same type of learning in all different school sectors. This idea of competition for teacher expectations brings us to rational choice theory. Rational choice theory is significant in this study because it will help decide if private school teachers are expected to perform better because students can leave the private school at almost anytime. If my outlook is correct in relating school competitiveness with teachers’ expectations, then rational choice theory explains this relationship. Relevant Sociological Theory: Rational choice theory is relevant to the research questions of: Is competition necessary for school systems? And also, do male students benefit more from private schooling than 2 The interaction hypothesis involves the benefits of private schooling and gender. Considering my alternative hypothesis one of competition being necessary for schools, I also suspect that males will perform better at private schools due to the greater attention males might receive in private schools. My null hypothesis two is: Male students will not benefit more from private schools than females do. My alternative hypothesis two is: Males will benefit more from private schools than females do. Research Design In this section, I will describe the research design that will be used to test my research questions. My first question is: do students in private schools outperform those in public schools, controlling for family income? In other words, is competition necessary for school systems? And my second question is: Do males benefit more from private schooling than females? I will be addressing the operationalization of key concepts, the sampling strategy that will be used, and my technique of data collection. I will also cover my design of proof and the limitations I may encounter. Operationalization The key concepts of my research questions are reading, science and math test scores, gender, school sector and family wealth. The test-score variables were constructed following principles of Item Response Theory, which is considered a strong approach for measuring student performance. Interested readers are encouraged to consult appendix H of the NELS: 88 Second Follow-Up Student Component Data File User’s Manual. I will study test scores for 5 reading (“by2xrstd”), math (“by2xmstd”), and science (“by2xsstd”). Higher scores on these measures indicate better performance. School sector is measured by a binary variable, “private,” that is scored 1 for students in Catholic or other private schools and 0 for students in public schools. Family wealth will be a control variable, which is referenced as “byses.” This variable is a composite measure that indicates parents’ educational attainment, parents’ occupation, and family income. As scores on this measure increase, family wealth increases. The effect of school sector will be assessed for males and females. The variable “female” will be used to measure gender. It is scored 1 for females and 0 for males. A multiplicative term that refers to the product of ‘private’ and ‘female’ will be used to test the interaction. Sampling My sample will consist of a total of 9,414 12th grade male and female students in who were studied in 1992. This was a third study completed after an initial study of 8th graders in 1988 and 10th graders in 1990. I will be attempting to generalize to the entire U.S. population of 12th graders. The sampling strategy used was a clustered, stratified national probability sample. One advantage to this type of sample is being able to study small subgroups that might usually be underrepresented. Also, the error is reduced when clustered and stratified samplings are used together. Weaknesses of this strategy will be discussed below in the limitations section. Data Collection Technique I will be using data that was collected and composed by the US Department of Education, National Center for Education Statistics. The distributor of this data is the Inter-university Consortium for Political and Social Research in Ann Arbor, Michigan. My data collection 6 technique is secondary data compiled from the National Education Longitudinal Study: Base Year Through Fourth Follow-up, 1988-2000. The original data was not collected by me. The advantage of using this data is that it was collected by social scientists with expertise in survey design and issues that pertain to education in America. Design of Proof My design of proof will be based on a series of multiple regression models. Each model will have an individual test score as the dependent variable. The effect of school sector will be estimated in each model while controlling for family wealth and gender. If I find that the variable “private” is positively related to a test score, I will reject the first null hypothesis and allow the first alternative hypothesis to stand, stating: Private schools produce higher test scores than public schools do. This hypothesis will be tested for three types of test scores: reading, math, and science. To test my interaction hypothesis, I will create a multiplicative term that is the product of the variables “private” and “female.” This multiplicative term will be added to the three models that will be run to test hypothesis one, so six models will be estimated in all. If the effect of the multiplicative term is negative and significant, that would mean that females benefit less than males do from being in a private school. Such a result would allow me to reject my second null hypothesis and allow the second alternative to stand, stating: Male students will benefit more from private schools than females do. 7 Results from Model 1 demonstrate that all factors are significant when it comes to reading test scores in relation to gender, family wealth and school sector. Family wealth plays the largest part in whether or not a student receives higher test scores. All variables produce positive effects on reading test scores; therefore, students in private schools outperform those in public schools, wealthy students outperform those from more deprived backgrounds, and females outperform males. Model 2 shows the math scores with the effect of family wealth, school, and gender. Model: MODEL2 Dependent Variable: BY2XMSTD MATHEMATICS STANDARDIZED SCORE Number of Observations Read 9414 Number of Observations Used 9414 Analysis of Variance Sum of Mean Source DF Squares Square F Value Pr > F Model 3 161574 53858 638.91 <.0001 Error 9410 793237 84.29720 Corrected Total 9413 954811 Root MSE 9.18135 R-Square 0.1692 Dependent Mean 52.30072 Adj R-Sq 0.1690 Coeff Var 17.55492 Parameter Estimates Parameter Standard Variable Label DF Estimate Error t Value Pr > |t| Intercept Intercept 1 52.05396 0.13966 372.73 <.0001 private 1 0.55270 0.31917 1.73 0.0834 BYSES SOCIO-ECONOMIC STATUS 1 5.59157 0.13237 42.24 <.0001 COMPOSITE female 1 0.08862 0.18951 0.47 0.6400 Parameter Estimates Standardized Variance Variable Label DF Estimate Inflation Intercept Intercept 1 0 0 private 1 0.01669 1.05224 BYSES SOCIO-ECONOMIC STATUS 1 0.40750 1.05407 COMPOSITE female 1 0.00440 1.00198 The results of Model 2 show that family wealth has the largest impact on math test scores. Gender is not significant when it comes to math test scores. Importantly, students in private schools outperform those in public schools. 10 Model 3 shows the results of science test scores with the effect of family wealth, school and gender. Model: MODEL3 Dependent Variable: BY2XSSTD SCIENCE STANDARDIZED SCORE Number of Observations Read 9414 Number of Observations Used 9414 Analysis of Variance Sum of Mean Source DF Squares Square F Value Pr > F Model 3 113054 37685 441.84 <.0001 Error 9410 802591 85.29126 Corrected Total 9413 915645 Root MSE 9.23533 R-Square 0.1235 Dependent Mean 52.03346 Adj R-Sq 0.1232 Coeff Var 17.74882 Parameter Estimates Parameter Standard Variable Label DF Estimate Error t Value Pr > |t| Intercept Intercept 1 52.64621 0.14048 374.77 <.0001 private 1 -0.22681 0.32105 -0.71 0.4799 BYSES SOCIO-ECONOMIC STATUS 1 4.60578 0.13315 34.59 <.0001 COMPOSITE female 1 -1.38119 0.19062 -7.25 <.0001 Parameter Estimates Standardized Variance Variable Label DF Estimate Inflation Intercept Intercept 1 0 0 private 1 -0.00699 1.05224 BYSES SOCIO-ECONOMIC STATUS 1 0.34276 1.05407 COMPOSITE female 1 -0.07000 1.00198 Model 3 results show that science is the only test score that is not positively related to private schooling. The results show that the relationship between the two is not significant but family wealth and gender are significant. In this case, males outperform females. Again, family wealth has the largest impact on student test scores out of the three variables. Model 4 explains reading test scores with the interaction of gender and private schooling. Model: MODEL4 Dependent Variable: BY2XRSTD READING STANDARDIZED SCORE Number of Observations Read 9414 Number of Observations Used 9414 Analysis of Variance Sum of Mean Source DF Squares Square F Value Pr > F 11 Model 4 139851 34963 414.93 <.0001 Error 9409 792820 84.26193 Corrected Total 9413 932672 Root MSE 9.17943 R-Square 0.1499 Dependent Mean 52.09153 Adj R-Sq 0.1496 Coeff Var 17.62173 Parameter Estimates Parameter Standard Variable Label DF Estimate Error t Value Pr > |t| Intercept Intercept 1 50.58890 0.14350 352.54 <.0001 private 1 1.32375 0.44496 2.98 0.0029 BYSES SOCIO-ECONOMIC STATUS 1 4.95023 0.13245 37.37 <.0001 COMPOSITE female 1 2.48881 0.19996 12.45 <.0001 prifem 1 -0.74791 0.62280 -1.20 0.2298 Parameter Estimates Standardized Variance Variable Label DF Estimate Inflation Intercept Intercept 1 0 0 private 1 0.04045 2.04593 BYSES SOCIO-ECONOMIC STATUS 1 0.36501 1.05580 COMPOSITE female 1 0.12498 1.11610 prifem 1 -0.01647 2.08220 Model 4 results display that gender does not condition the effect of private schooling. Model 5 shows math test scores with the interaction involving sector and gender. Model: MODEL5 Dependent Variable: BY2XMSTD MATHEMATICS STANDARDIZED SCORE Number of Observations Read 9414 Number of Observations Used 9414 Analysis of Variance Sum of Mean Source DF Squares Square F Value Pr > F Model 4 161702 40426 479.59 <.0001 Error 9409 793109 84.29257 Corrected Total 9413 954811 Root MSE 9.18110 R-Square 0.1694 Dependent Mean 52.30072 Adj R-Sq 0.1690 Coeff Var 17.55444 Parameter Estimates Parameter Standard Variable Label DF Estimate Error t Value Pr > |t| Intercept Intercept 1 52.01316 0.14352 362.40 <.0001 12 teachers performing better in private schools because of the benefits, private school student test scores are higher than public school student test scores. Because the results prove that reading and math test scores are positively related to private schooling, RCT works for my research question. RCT can be used to predict student test scores in different school sectors because their results are better or worse depending on teacher performance. These results have policy implications. School systems should start using more competitiveness when it comes to teacher salary, teacher performance and teacher involvement. With competition, teachers will have to perform well to stay at a job. If they are performing well, then the students will be performing better. If they are performing badly, then the teacher will have negative consequences. There should also be some consideration when it comes to family situations. Since family wealth is an important factor in test scores, students in families with lower income should be given special classes for no cost to help them with this gap. They should be given the same opportunities that the wealthier students are getting. Although the results presented are consistent with my alternative hypothesis and RCT, there are some weaknesses. There are possible prior variables that may have affected the results in this research question. In mentioning that students should be given the same opportunities, this implies that a prior variable of social upbringing might be important. This social upbringing could be clubs, camps, or other extracurricular activities that wealthier students can afford. Another prior variable that could have affected the results is what type of pre-school the student attended. In sum, the actual results may be from different social upbringings or the pre-school that was attended instead of the current school sector that is being attended. In review, competition is necessary for schools. RCT is a useful theory since it did a good job of predicting my results. Student reading and math test scores are positively related to 15 school sector. My analysis also showed that family wealth is the most important predictor of test scores and that gender gaps remain in reading, where females excel, and in science, where males excel. The results shown provide Rational Choice Theory with more credible evidence. 16 References: Scott, John. Understanding Contemporary Society: Theories of The Present, edited by G. Browning, A. Halcli, and F. Webster. (Sage Publications, 2000). Liska, Allen E. and Steven F. Messner. Perspectives on Crime and Deviance, 3rd Ed. (Upper Saddle River, NJ: Prentice Hall, 1999). National Education Longitudinal Study: Base Year Through Fourth Year Follow-Up, 1988-2000: United States Department of Education National Center for Education Statistics, First ICPSR Version, May 2004 17