Soc20: Research Methods- Regression Analyses- Chi-Square Test- Sociology Assignment Help

Internal Code: 1HHFI

Sociology Assignment Help:

Task: In this assignment, you will add regression analyses, a conclusions section, a literature review, update your tables, and properly number your tables and figures. Instructions are outlined below beginning at part IV. Use the same Word document that you used for Paper 1. Write the sections in the order listed in this document. I. Research Question a. State your research question. Then discuss why you feel it is important to know the answer to your research question. b. What is the hypothesis? c. What is the dependent variable? How is it operationalized in the data? d. What is the independent variable? How is it operationalized in the data? e. What (if any) changes will you need to make to how your dependent and independent variables are operationalized, and why? II. Descriptive Information on Your Variables a. In a table, present the mean, standard deviation, max and min for your independent and dependent variables. Provide at least three sentences interpreting the table. b. In a table, present a crosstab of your two variables. Your table should report means, standard deviations, and frequencies. Provide at least three sentences on any patterns you notice. c. If both your dependent and independent variables are categorical, construct a bar graph. Make sure the graph has all of the relevant titles. Interpret your graph. d. If both your dependent and independent variables are continuous, construct a histogram for each continuous variable. Make sure the histograms have all of the relevant titles. Interpret each histogram. e. If you have one continuous and one categorical variable, you will do two things: 1) create a bar graph using both variables and, 2) create a histogram of the continuous variable. Make sure the bar graph and histogram have the relevant titles. Interpret the bar graph and the histogram. For the histogram, discuss what changes you made to the way the histogram looks (e.g., the number of bins you decided to use compared to the number of bins Stata used) and why you made them. III. T-test For Mean Differences & Chi-Square Test of Independence If you have a continuous dependent variable and a categorical independent variable that is dummy coded, then conduct a t-test for mean differences across the two groups. Present the output in a table labeled “Table 2” and give it a title. In the table, show the independent variable groups, the mean for each group, the standard deviations, the t-value, and the p-value for Ha: ! = 0. Provide an interpretation of the table that explains what the table is telling you substantively about your variables. Do not just say what the numbers are, I can tell that from your table. Also, state whether your result is statistically significant and what number(s) you are using to make that determination. If you have two categorical variables, then conduct a chi-square test of independence. Present the output in a table labeled “Table 2” and give it a title. Provide an interpretation of the chi-square results that explains what the results are telling you substantively about your variables. Do not just say what the numbers are, I can tell that from your table. Also, state whether your result is statistically significant and what number(s) you are using to make that determination. IV. Analyzing the Relationship between Your Independent and Dependent Variables Run a correlation of your dependent and independent variables. Place this correlation matrix into a table titled Correlation of Dvname and Ivname. Do not put the Stata names of the variables in the tables; rather, use the formal names of the variables. For example, put “Completed Years of Education” rather than “educ”. Place an asterisk next to correlations that are statistically significant. Explain which correlation method you used and why. Then provide at least three sentences on the observed correlation including, whether the relationship is statistically significant and what the correlation results are saying substantively about the relationship between your variables. Paper 2 Starts Here First set of instructions: Go back to the sections you wrote in Paper 1 and correct any mistakes with operationalization, or failing to state what your dependent variable is, or anything that your TA told you to correct. Any mistakes you made in the first assignment that are not corrected will be marked off on this assignment. This includes text, tables, table titles, and so forth. V. Regression Instructions a. You will make 1 regression table for instructions in VI. You will title it Linear, Logistic, Multinomial (whichever you conducted) Regression of Dvname Ivname. Round decimals to the second or third place. Place an asterisk next to the coefficient if it is significant at the .05 level. VI. Add Control Variables to Your Regression a. Pick (10) control variables to analyze in your regression. b. First, regress your DV on your IV. This will be Model 1. c. Second, regress your DV on your IV and control variable #1. This will be Model 2. d. Remove control variable 1. Next, regress your DV on your IV and control variable #2. This will be Model 3. e. Continue this pattern examining one control variable at a time. You are looking to see how your main relationship changes as each control variable is added to the model. f. Finally, regress your DV on your IV and all of your control variables. This will be your final regression model for this Table. g. Include the constant in each model, but only interpret the constant in the first model (Model 1). h. In the last line of your table (the very bottom row within the table border), make an 2 R row. List the 2 R for each regression you conduct. i. In your table, place an asterisk next to significant coefficients. *j. Now discuss what you found in your analyses. k. Use your full regression model to create a new table that has 1) all of your predictors from the full model 2) their effects, and 3) p-values for each coefficient. Give it the title “Trimmed Regression Results” and number it in sequence. No discussion of this table is necessary, but it needs to be in your paper. *Discussing Your Regression Output Say which type of regression you conducted (linear, logistic, multinomial) and why. Then, in at least two paragraphs, say what the main relationship is that you are examining, what the control variables are, why you chose them, and why you recoded them in the ways you did. Now state the number of observations your results are based on. Next, discuss the relationship of interest (DV-IV) beginning with Model 1. Interpret the effect of your independent variable. Here you should say what the direction of your effect is (positive/negative), whether your relationship of interest (DV-IV) is statistically significant and say what the p-value is. Then say what the statistical significance means for the relationship between your two variables. Here, give a substantive explanation rather than a mechanical one. That is, do not say “a one-unit change in this leads to a one-unit change in that” and stop there. Say what those units are and talk briefly about what this is saying about the real world. Next, talk about how this relationship changes with the addition of each control variable. Here you should frame your discussion as a summary of what’s happening to the main relationship as you move across models, as opposed to repeating yourself from Model 1 to Model 2 to Model 3... etc. Most of your discussion should center on what changes you see in your main relationship as you controlled for various factors. Point out control variables that appear to be important in their effects either uniquely on your dependent variable, or on your main relationship. Be sure to convey the story of the model like we’ve been doing with the class examples. For instance, what are you left to conclude about your DV/IV relationship after you have done this statistical assessment? Lastly, interpret the 2 R from your first model (DV-IV). Then provide a few sentences on how the 2 R changed from your first model to your full model and what that is telling you about the main independent variable and controls you selected. A good discussion will likely be close to 1-page. Update your previous tables (and do your tables correctly) Place all Tables and Figures at the end of your paper (after the References section). a. Run a correlation of your IV, DV, and controls. Put them all in the correlation together. Place an asterisk next to any significant correlation coefficients. Summarize the table in 1-2 paragraphs. This new correlation table will replace the correlation you did for the first assignment. In other words, you should turn this assignment in having only 1 correlation table that includes all of your variables. Be sure to focus on the largest and smallest correlations (if they are significant) and discuss their importance to your model and the larger U.S. population – think about the big picture here. b. Go back to your descriptives table and update the means, standard deviations, min & max for all of your variables (IV, DV, controls). This should be Table 1 in your document. After you make this update, you should have only one Table 1 that includes all of your variables. Provide a discussion of the variables in the Table focusing on several of the means that you feel are interesting/important. c. Keep the Figures in your paper that you made in the first paper assignment. Update them if you did them incorrectly in the first paper. Your final paper will be marked off if there are no Figures that you were required to do in the first assignment, or if they are still done incorrectly. d. Pick two control variables and create either a bar graph using both variables or two separate histograms. Make sure the bar graph and histogram have the relevant titles. Interpret the bar graph/histograms (whichever you chose). For the histograms, discuss what changes you made to the way the histograms look (e.g., the number of bins you decided to use compared to the number of bins Stata used) and why you made them. Number the additional Figure(s) sequentially. VII. Conclusions Section Write a conclusions section (close to 1-page) recapping what you did in your research paper. This section begins by telling the audience what question you set out to answer, then provides a summary of what you found. Only discuss what you feel were the key analyses that gave you a good picture of what’s happening between your two main variables. This section is meant to provide the audience with a discussion about the importance of your research question, and what the findings mean for the real world. Your TA will help you think about this. VIII. Literature Review Write a 1-page literature review on your main variables (dependent and independent). Your TA will guide you on this part.