Practice 1, 2016 EF5133 Practice Assignment 1 Due: 2016-03-04, 9am Introduction In this assignment you will analyze a cross-sectional data set of student test scores from California similar to the one used in Stock and Watson (2011). The data used in this assignment come from a more recent period (year 2007 or later) and are based on individual schools rather than districts. Each student will be assigned a data file for a particular academic year based on the student id number. To find out which data file you have been assigned, modify the sample code and replace the id number in the code with your student id number. ##***change this number to your student id number*** idnum = 0; A description of the variables in your assigned data file is given in the file caYYapi.html where YY is your assigned year number. The outcome variable of interest is the academic performance index (API) apiYYb. "The API is a single number, ranging from a low of 200 to a high of 1000, which reflects a school's performance level, based on the results of statewide assessments. [...] The API is calculated by converting a student's performance on statewide assessments across multiple content areas into points on the API scale. These points are then averaged across all students and all tests. The result is the API." The analysis sample is restricted to middle schools (excluding special education schools) with at least 100 students included in the calculation of API. Assignment (a) Report summary statistics for the following variables: apiYYb (academic performance index), acs_46 (average class size for grades 4-6), p_el (percent English learners), meals (percent students receiving free or reduced price lunch), pedu (average parent years of education). Display the scatter plot of the outcome variable of interest apiYYb against the other four variables. Which pair of variables appear to have a linear relation between them? (b) How many school districts are there in your sample? How many schools are there in the district with the largest number of schools? How many schools are there in the district with the smallest number of schools? (c) Report regression estimates for the following four regressions in one table. apiYYb = !0 + !1acs_46 + u apiYYb = !0 + !1acs_46 + !2p_el + u apiYYb = !0 + !1acs_46 + !2p_el + !3meals + u apiYYb = !0 + !1acs_46 + !2p_el + !3meals + !4pedu + u Explain the likely relation between class size and test performance based on these four regressions. (d) Consider two hypothetical schools A and B that differ only in p_el and meals (any other characteristics are assumed to be the same). For school A, p_el is equal to the sample mean and meals is zero. For school B, p_el is zero and meals is equal to the sample mean. Compute a 95% confidence interval for the expected difference in API between these two schools using estimates from the last specification with k = 5 regressors. Is this difference 'significant'? (e) In 1992 California passed a charter school law, the second state to do so after Minnesota passed one in 1991. Since then the number of charter schools in California has grown from 238 (2.8% of all public schools) in 1999/2000 to 1,065 (10.7% of all public schools) in 2012/2013. The assigned data sample contains two dummy variables dcharti and dchartd for charter schools. Report regression estimates to examine whether student performance as measured by API and its relation with the explanatory variables considered in part (c) differs between charter and non-charter schools and between the two types of charter schools. Carefully interpret the results you obtained. Submission Instructions If you want your submisstion to be considered for marking, you must electronically submit your report by the due time. Following the link in the course page from Loop to submit your report. No submissions after the due time are accepted. (The due time is based on the server's system clock, not your computer's clock. You should submit your report well in advance of the due time.) For this practice assignment, you can (and I encourage you to) work as a group. For a group report, work with one of the data files assigned to a member of the group. Report Format Your report must be typeset as an A4-sized document and submitted as a single pdf (portable document format) file. The report should begin with a title section that clearly lists your name, student ID number, and DCU email address. Use a reasonable (i.e. legible) font size, line spacing, and margins. There is no limit or requirement in terms of word counts or page numbers. However, your report should be kept short and to the point. Read the questions carefully. Marks will be deducted for long answers that are not directly relevant to the question. Do not include program or code in the main body of the text. Put all such information in an Appendix with a smaller fixed width (monospaced) font at the end of your report. I may ask you to submit the code you used for the assignment. You must have the code ready to submit if I ask you to do so. Use tables and figures effectively. Each table and figure must be accompanied by a caption or note explaining the contents of the display (e.g. what are the numbers in parentheses?). Do not place tables and figures at the end of the report; place them in the main body of the text where they belong. When you report estimation results, do not copy-and-paste output from the software. Properly format them into tables or graphs as is done in academic journals. In particular, do H.Kawakatsu 1 of 2Practice 1, 2016 EF5133 not report too many digits just because the software prints them. When you quote results from other sources, clearly indicate the source. For example, there is no need to rederive standard statistical results in your report; just quote from the relevant sources. Improper attribution can result in plagiarism. In the text of your report, it is standard to refer to these citations simply as author (date), e.g. Smith (1992) or Smith (1992, p.27). Consult Citing & Referencing for further guidance such as styles for quoting a web site. Do not use footnotes or endnotes. Do not 'decorate' your report. No marks for using fancy fonts, glittering colors, or blinking images. This is not a business report; it should be written as an academic paper. Re-read the sample report carefully. References Stock, James H. and Mark W. Watson (2011) Introduction to Econometrics: Pearson, 3rd edition. H.Kawakatsu 2 of 2