EC2017 Individual Report How to submit? Upload your work via submission point on Moodle. What to submit? Submit ONLY 1 MS Word file with your answers. Do not forget to

copy all your commands from your do-file at the end of your work. Your results should be replicable using those commands. Word limit? Recommended word limit is 1,500 words. 10% below or above this word limit is allowed. Figure/table captions, Stata commands, text in tables, footnotes, and regression specifications do not count towards the word limit. Data: You are asked to work with European Social Survey (ESS) data, a cross-national survey that has been conducted across Europe since 2001. In this assessment, we will use the final round only (data published in 2023). The description of all variables and data collection processes is available together with this assessment setup on Moodle. In this task, you are asked to proceed with regression analysis, and write a report. Read the questions carefully, full marks can be achieved only if you answer the questions fully. Use all the countries in the dataset. Question 1) You are interested in learning about the likelihood of voting in an election. (Hint: define a person as having voted (=1) if they have voted in the previous election and 0 if the person has not voted in the previous election). a) You are interested in whether there are gender differences in voting behaviour. You estimate a regression model with being a woman as the single explanatory variable. Clearly write down the regression model. Then estimate it and interpret the estimated coefficients. Can you see any gender differences in voting behaviour? b) Your friend argues that when having a paid job in the last 7 days is excluded from the estimated model in part 1.a, the coefficient on being a woman will be biased. Explain to your friend the conditions under which excluding the variable of having a paid job in the last 7 days makes the coefficient on gender to be biased. Intuitively discuss the direction of bias. c) Then use the formula for omitted variable bias to calculate the size of bias. Confirm your answer by comparing the coefficient on gender from estimating the model in part 1.a with the new model. (Hint: you might need to restrict your sample to those observations without missing values to be able to obtain the correct size of bias). d) You now want to test whether the impact of having a paid job on voting behaviour depends on the individual's age. Write down a model which allows you to directly test that hypothesis. Describe step-by-step how you build your null and alternative hypotheses, how you select the test, how you calculate the test statistic, and how you make your decision over the rejection/not rejection of your null hypothesis and conclude. (You can exclude gender from the explanatory variables). e) Then proceed and estimate the model in part 1.d. Are there any age differences in the impact of having a paid job on voting behaviour? Comment on your findings and carefully interpret all the estimated coefficients in that model. f) Continue to use the regression model in part 1.a. You think that education is an important predictor of voting behaviour. You want to add years of education as an explanatory variable to this model, but your friend argues that the impact of education on voting behaviour is concave. Write down a model which allows you to test this hypothesis. g) Estimate the model in part 1.f and carefully interpret the relationship between years of education and voting behaviour. Question 2) You are interested in investigating whether the impact of the determinants of trust in the police in Scandinavian countries is different compared to the impact of the determinants of trust in the police in other countries. (Hint 1: Scandinavian countries include Denmark, Finland, Iceland, Norway, and Sweden. Hint 2: You need to write down a single regression model to discuss this question). a) Write down a regression model which allows you to test this. Define your dependent and independent variables and justify your choice. Describe whether you expect positive/negative estimated coefficients for each independent variable. b) Using the variables you selected, estimate your regression model. Export the results into a table and carefully comment on your findings. c) Write a short report summarizing your overall findings (suggestion: 200 words). Question 3) At the end of your work, include a copy of your do-file containing all your commands. Ensure that your do-file has comments explaining the purpose of each line, to clarify what each command is doing. Marking Criteria Assessment criteria Knowledge and understanding of theory and techniques Analysis and evaluation of evidence from data and statistical investigation Fail (0-39) Limited, confused and not accurate spelling mistakes; accurate citation referencing no clear ability of applying such techniques Discussion 1S understanding of theory and statistical techniques learned in the methodologies module; mainly descriptive; no evidence of deep understanding; no ability to properly analyse evidence presented graphs, tables and statistics; and mainly Pass (40 - 49) Evidence of some but rather basic Quality of Poor writing Broadly a style style, quality of difficult to report: writing graphs and understand; tables; no poorly and positive report even if the writing style should improve; some spelling mistakes; not always accurate referencing and citation inaccurate citation referencing understanding and ability of application of the and learned throughout the term Satisfactory (50-59) There 1S evidence of learning even if not always as accurate and detailed as required by the investigation There 1S a Good than tendency for descriptive rather analytical interpretation of the outcome of the statistical investigation interpretation the investigation results with some ability of analysing the evidence even if this is not always as informative and accurate as it should be Generally, a well written report but with some inaccuracies in citation and referencing. No attempt to create a narrative of the evidence of Good (60-69) Clear evidence of learning and engagement with the subject matter taught during the term. Shows a firm understanding and ability of application Good analytical skills and, overall, shows a 'control' of the subject matter and a clear ability to interpret and, at least partly, critically evaluate the evidence Excellent (70 100) Strong and firm understanding of the required statistical methodologies; clear evidence of an ability to apply them in context Excellent analytical skills with clear evidence of an ability to critically evaluate findings emerging from the investigation Very Overall, a well written report written without spelling mistakes and good referencing citations. There is an attempt to create narrative that, however, is not always successful a the well report with a nice attempt to create a narrative of the evidence from the investigation. Good overall referencing and no spelling mistakes