[SOLVED] CS PUBL0055 Moodle page, is entitled resource_curse.csv and contains informati

Section 1: Natural Resources and Democracy (50 marks)
In 2001, famously asked Does Oil Hinder Democracy?, a question which his analyses lead him to answer with Yes. At least since then, there has been a long-standing interest in political science in the so-called resource curse, which describes the tendency of countries that are rich in natural resources to have worse economic, political and social outcomes than countries without lots of natural resources. In particular, some countries whose economic model relies heavily on oil production are among the most autocratic regimes.
In this section, you will re-examine the relationship between oil production and democracy in more recent years. The data file you will use, which can be downloaded on the PUBL0055 Moodle page, is entitled resource_curse.csv and contains information about 170 countries in the year 2015. The data includes the following variables:
Variable name

Copyright By Assignmentchef assignmentchef

country_code
fh_political_rights
log_gdp_cap
log_population
Description
Country name
Three-letter country code
World region where the country is located (Africa, Americas, Asia, Europe, Oceania)
Oil revenue as percentage of total GDP (Data from the World Bank) Polity V score which measures how a countrys government is selected, competed for, and constrained, on a scale from -10 (strongly autocratic) to +10 (strongly democratic)
Freedom House political rights score which measures the degree of political freedom in a country, on a scale from 1 (not free) to 7 (very free)
Logged real GDP per capita in $ (Data from the World Bank)
Logged population size (Data from the World Bank)
You can load the data by using the following command:
curse <- read.csv(“resource_curse.csv”) Answer the following questions:Question 1 (10 marks)Lets begin your analysis by looking at the two variables that measure the outcome of interest (the degree of democracy): polity5 and fh_political_rightsa. Calculate and interpret the correlation coefficient in substantive terms. Note that, as there are missing values, you have to specify the option use = “complete.obs”. (5 marks)b. Create a scatterplot with, on the one axis, the Polity V score, and, on the other axis, the FH political rights index. Comment on what you see. (5 marks)Question 2 (10 marks)Now, have a look at the independent variable, the share of oil production in each countrys GDP.a. Which countries have missing values in oil_share? (3 marks)b. Visualise the distribution of the variable oil_share. What do you observe? (4 marks)c. How many observations have no oil revenue at all (0% of total GDP)? (3 marks) 3 Question 3 (15 marks)In the next step, lets investigate the bivariate relationship between the dependent and independent variables.a. Run two simple linear regressions, both with oil_share as the independent variable, one with polity5 as the dependent variable, one with fh_political_rights as the dependent variable. Present the regression results in a table. (5 marks)b. Interpret each of the regression coefficients for oil_share in substantive terms. (3 marks)c. Compare the regression coefficients for oil_share between the models. Can we conclude that the association is stronger for one of the measures of democracy than the other? If not, why not? (3d. Interpret and compare the R2 (focus on the multiple R2, not the adjusted R2) for each model insubstantive terms. (4 marks)Question 4 (15 marks)When investigating the relationship between oil share and the measures of democracy, we may want to hold some other variables constant.a. Using polity5 as a dependent variable, run a multivariate regression with oil_share, log_gdp_cap, log_population and region as independent variables. Present the results in a table. (5 marks)b. Interpret coefficient for oil_share in substantive terms. How does it compare to the one in the bivariate regression with polity5? (5 marks)c. How does the interpretation of the intercept compare between the bivariate and the multivariate regression with polity5? (5 marks)Section 2: Combatting COVID Vaccine Hesitancy (40 marks)During the first roll-out of COVID-19 vaccines, many governments were faced with high levels of vaccine hesitancy. Focusing on Germany, Heike Klver and co-authors conducted a survey experiment in March 2021 to try to find out whether individuals willingness to get vaccinated could be increased with specific policies designed to reward vaccination. More specifically, the researchers designed an online survey, in which they randomly allocated survey respondents to different groups, that were presented with different COVID-policy proposals.In this question, you will focus on the effect of a policy proposal that would grant vaccinated people more freedoms than unvaccinated people. Specifically, the treatment group was shown the following:Imagine the following scenario: Special regulations apply to vaccinated people. For example, even when the Corona incidence is high, they can travel again, visit cinemas, restaurants or concerts and are not subject to any contact restrictions.By contrast, the control group was shown:Imagine the following scenario: There are no special regulations for vaccinated people even when the Corona incidence is high. For example, they cannot travel again, visit cinemas, restaurants or concerts and are still subject to contact restrictions.The outcome was measured on a scale from 0 to 10 with the survey question:Please use this scale to indicate how likely it is that you would be vaccinated against corona under these conditions.The data file you will use, which can be downloaded on the PUBL0055 Moodle page, is entitled vaccine.Rdata and contains information on 3062 respondents from the experiment. Note that the data does not contain anyone who was already vaccinated. The data includes the following variables: Variable namewillingnessuniv.degreeDescriptionRespondents pre-treatment stated willingness to get vaccinated. Factor variable with 3 levels: Willing, Unwilling and Undecided.Dummy variable indicating whether the respondent is male (1) or female (0). Age group of the respondent. Factor variable with 4 levels: 18-34, 35-49, 50-65, 66+.Dummy variable indicating whether the respondent has a higher education degree (1) or not (0).Treatment variable: dummy variable indicating whether the respondent was in the treatment group (1) or the control group (0).Outcome variable: respondents post-treatment-stated likelihood of getting vaccinated. Numeric variable ranging from 0 (I will definitely not get vaccinated) to 10 (I am sure to get vaccinated).You can load the data by using the following command:load(“vaccine.Rdata”) Answer the following questions:Question 1 (10 marks)Lets begin by looking at the attitudes towards vaccination before treatment.a. What are the proportions of respondents in each category of the variable willingness? (2 marks)b. What are the share of those willing to get vaccinated among male respondents? How does that compareto the share of those willing to get vaccinated among female respondents? (4 marks)c. What was the proportion of people who were willing to get vaccinated and had a university degree? Among those who had a university degree, what is the proportion who were willing to get vaccinated?Explain the difference between these two proportions. (4 marks)Question 2 (15 marks)We will now turn our attention to the experimental part of the data.a. What feature of experiments allows researchers to estimate convincing causal effects and why? (5 marks)b. Have a look at the proportions of the pre-treatment variables male, univ.degree and age in the treatment and control group, respectively. Are the distributions broadly similar? (5 marks)c. What do your results in b. tell us about whether the conditions for estimating a convincing causal effect are met in the data? (5 marks)Question 3 (15 marks)Lets look at the outcome to estimate the treatment effect.a. Calculate, report and interpret the difference between the mean outcome in the treatment group and the mean outcome in the control group in substantive terms. (3 marks)b. Calculate and report the difference in the mean outcomes between the treatment and control groups, but only for those respondents who were undecided at the start of the survey. (2 marks)c. Calculate and report the difference in the mean outcomes between the treatment and control groups, but only for those respondents who were unwilling to get vaccinated at the start of the survey. (2 marks)d. Calculate and report the difference in the mean outcomes between the treatment and control groups, but only for those respondents who were willing to get vaccinated at the start of the survey. (2 marks)e. Compare your results from b., c. and d. to your response to a. What do you conclude from this about the effectiveness of the treatment? In particular, have a think about what those willing to getvaccinated would have answered to the outcome question, in the absence treatment. (6 marks) CS: assignmentchef QQ: 1823890830 Email: [email protected]