I only need the format done first. That is due on the 29th The final project is due on dec 12th. SOCI 303 Statistical Analysis Final Paper I. Overview In this paper you will use statistical tests to explore a sociological topic. Specifically, you will perform the following: 1) 2 correlations 2) 2 scatter plots and 3) a bivariate linear regression model. This paper will be an essay with an introduction, one paragraph to describe the variables and data, three paragraphs to report the findings of each type of test, and a conclusion. Your paper should be readable to someone who is not familiar with statistics. II. Data Sets On our course website, I have provided to you several data sets from which to choose. You must choose a data set of interest to you from the following: Global data (global), Census level City data on crime (city), the Health Interview Health Information National Trends Survey, 2012 (HINTS), or one of the topical datasets made from the General Social Survey (GSS). All tests in the paper should use one dataset. III. Paper Format and Content The following format should be followed. Each heading should comprise one (1) paragraph except where noted. Use these headings and put them in bold. No cover pages please. Introduction v Give a general overview of your topic and its sociological significance. What is the research question your paper is examining? Data description v Describe the data set used in your study. Give the official title of the data and who collected it, etc. (provided in codebook). Variable description v Describe the variables you chose. Explain variables using normal language and give variable names or abbreviations that you may use during the paper. For example, “Population is the number of people living in the city. It is an interval-ratio variable with a mean of xx and a standard deviation of xx and a minimum and maximum of xx and xx.” v For ordinal variables you will need to explain the coding (e.g., health status was coded as follows: 1=poor, 2=fair, 3=good, and 4=excellent). v Give descriptive statistics (mean, n, standard deviation, minimum, and maximum) for each of the ordinal or interval-ratio variables and interpret those (e.g., the average health status was 3.2, which corresponds to good, or the average number of homicides across cities was 20.6). v None of your variables should be nominal (e.g. Marital Status is nominal because it has no rank or order). Correlation and Scatter Plot (2 paragraphs, one for each pair of variables tested) v Two correlations and scatterplots. For each correlation, you will choose two interval- ratio or ordinal (with 5 or more categories) that you theorize should be related. At least one of your correlations should have a Pearson’s correlation coefficient greater than 0.10 (r>0.10) so you can use it for your regression analysis. Include a scatter plot for each correlation. Scatter plots involving ordinal variables may not be easily interpreted, and if that is the case, you can still use the graph and state that it was not informative or could not be interpreted. Otherwise, make some interpretation of the scatter plot and relate it to the correlation. Report the results from your correlation, including the direction, magnitude, and sample size. See example in “Example Statistics Paper” on TITANium. Bivariate Linear Regression v Next you will perform a bivariate linear regression for your strongest correlation (correlation coefficient must be greater than 0.10). Report and interpret the R-square. Report and interpret the regression equations. You should also predict the dependent variable for a given value of the independent variable. You will choose the value of the independent variable because it is of interest to you (e.g., I am interested in the number of drinks consumed by a student with GPA of 2.0) or you may use the mean of X. See example in “Example Statistics Paper” on TITANium. Conclusion v Finally, in your last paragraph of the paper, summarize your findings without using numbers. Give your overall conclusion about each of the relationships, not using numbers, based on studying the data. You should also discuss whether the results were surprising or what you expected. State any weaknesses of your study, such as small sample size, wording of survey questions, sample composition, variables that were omitted or not available in the data, level of measurement of variables, etc. You may wish to suggest future research or what you would like to do next to understand your research questions. IV. Tables All tables should be copied and pasted from SPSS into Word and can be either within the text or at the end of the essay. If copy/paste doesn’t work, you can take screen shots or recreate the tables in word. Each table should be labled properly (Table 1, Table 2, Figure 1… etc.), have a title (e.g. Table 1: Correlation 1) and be formatted to be easily understood by your reader (tables should be same size and formatted the same/consistently). Refer to tables and figures in the text. Cite notable statistics from the tables using sentences in the paper. V. Formatting and Writing The format of the report is an essay where you use statistics to support your statements. Your paper should be readable to someone who is not familiar with statistics. Remember to round to one or two decimal places and be consistent when reporting statistics. You should also not start a sentence with a numerical symbol. State your results clearly and simply. This is not creative writing. The goal is to be clear and precise. The standards of formal and professional writing and grammar should be observed. Points 5 5 10 30 20 5 100 Paper Planning Form submitted by due date (see TITANium or syllabus) Introduction Description of data & variables Correlations (tables 5 each, interpretations 10 each) Scatter plots (plots 5 each & interpretations 5 each) Conclusion Writing / Grammatical mistakes will deduct points Total 25 SLR (equation & interpretation of equation 10, prediction of dependent for some independent value 5, R-square & interpretation 5, table 5) Steps to get started on the writing assignment ü Look at the codebooks for each of the data sets. ü Choose one data set of most interest to you. ü Look for types of variables in the data set (ordinal w/5+ categories or interval-ratio). ü Decide which variables you think may be related to each other. ü Decide which variables to use for which tests using table on the next page. ü Calculate descriptive statistics for each of the variables. ü Calculate 2 correlations. Make 2 scatter plots. ü Use two variables from correlation to calculate regression. ü Write the paper following the above format. To be completed to help you while writing your paper. You do not need to turn this in. This is not the planning form. Correlation 1 results (r, direction +/-, magnitude) Interpretation: Correlation 2 results (r, direction +/-, magnitude) Interpretation: Scatter plot 1 (linear? Positive or negative? Strong or weak relationship?) Interpretation: Scatter plot 2 (linear? Positive or negative? Strong or weak relationship?) Interpretation: Regression Equation: (?? = ??$ + ??&??) Interpretation: R2= Interpretation: Predict Y for a Given X (you choose X within range of X variable)

# I only need the format done first. That is due on the 29th The final project is

Leave a reply