List of tasks


You have three tasks to complete. For tasks 2 and 3 you should also include the SPSS outputs of
your analysis (correlation in task 2 and regression in task 3). To provide these outputs either
directly export the tables or figures from SPSS or use the printscreen command and paste the
results into your file. In case of any technical troubles do not hesitate to contact me. Everything
needed to solve the task can be found in the chapters of the textbook by Field that I included as
readings to our session. Deadline is January 20, 2021. Ideally deliver one text file (docx, pdf or
anything similar) with all your answers and solutions.


Task 1

For each of the 10 enlisted variables decide whether they are nominal, ordinal or scale and shortly
explain your decision.

Variables:

-          Names of countries

-          Percentage growth of GDP of countries

-          Military ranks (captain, general etc.)

-          Number of members of households

-          Blood type of hospital patients

-          Age of people measured in years

-          Size of countries in square miles

-          Educational levels of people (coded as 0 for people without any education, 1 for people
with elementary education, 2 for high school graduates and 3 for college graduates)

-          Color of hair

-          Unemployment rate of countries


Task 2

The file PresidentCZ.sav contains the results of presidential election in Czech Republic for each
district in the country. The elections use a two round system and if none of the candidates obtains
majority of votes in the first round, two strongest candidates continue to the second round where
the one with more votes wins. Your task is to calculate two correlations. First, calculate the
correlation between the support of Zeman (in the first round) with support of Zeman (in the second
round). Second, calculate the correlation between Schwarzenberg (in the first round) with Sobotka.
For both calculations choose the proper correlation coefficient and explain why you have chosen
this coefficient. Also answer the following questions:


What is the correlation between the support of Zeman (first round) and Zeman (second round)?

What is the correlation between the support of Schwarzenberg (first round) and Sobotka?

Are these correlations significant?

Briefly describe your findings.


Task 3


The file GreenParty.sav contains a dataset of electoral results of Green party in all districts of
a country. The dataset includes one dependent variable (GreenParty) and three independent variables
(Graduates, Pensioners and Campaign). Your task is to provide a regression analysis to find out the
effects of independent variables on your dependent variable. The characteristics of the variables
are as follows:

Region – only the ID of regions (not used for the regression)

GreenParty – percentage share of votes obtained by Green Party in region (dependent variable)

Graduates – percentage share of people in the region with university degree (independent variable)

Pensioners – percentage share of pensioners in the population of the region (independent variable)

Campaign – a binary variable which has value 1 if the Green Party has campaigned in the region
before the election and value 0 if the party has not campaigned in the region before the election
(independent variable)


In this task you have to calculate a linear regression model and answer these questions:


Based on R^2 what is the fit of your regression model?

What is the effect of Graduates on the dependent variable GreenParty?

What is the effect of Pensioners on the dependent variable GreenParty?

What is the effect of Campaign on the dependent variable GreenParty?

Which of these effects are statistically significant?

Which independent variable is the best predictor of the dependent variable and based on which
indicator can we know that?

In one paragraph explain briefly what are the main findings based on your regression model.