You have successfully completed the online course modules. Congratulations! Now, you can test what you have learned in solving a practical task! This link provides you with data of Campbell’s trial-heat model1, which predicts the incumbent party’s candidate popular vote (i.e., the dependent variable) in U.S. presidential elections from two independent variables:

septpoll: Voter support for the candidate of the incumbent party in polls conducted in September of the election year
gdpqtr2half: GDP growth in the second quarter of the election year

Given this data, your task is the following:

Install R/RStudio
Import the data described above into R
Describe and interpret the data (also using visualizations where appropriate) with R, in

particular
a. The types of variables b. Numerical summaries c. Correlations with
2. The other independent variables
4. The dependent variable
Use R to estimate Campbell’s model by regressing the incumbent party’s candidate popular vote on the two independent variables described above and interpret the model coefficients (Course: Multiple and logistic regression).
Calculate the model forecast for the 2020 U.S. presidential election and compare the forecast to the outcome.

Hint

You might need to go back to the course modules to remember the different commands. Alternatively, you might need to look online for help. This is normal and very common when using software such as R. The big advantage of R is that it’s not only free, but it also has a large community that offers lots of help online. Therefore, it is part of the learning goal that you accomplish the skill to look for – and find – information that will help you to solve the problem.

Structure of your project thesis

1 It’s not necessary to read the research article that describes the model. That said, reading it would probably help you to better understand the model.

(Course: Working with the RStudio IDE)

(Course: Importing Data in R)

(Course: Introduction to data)

(Course: Exploratory data analysis)

(Course: Correlation and regression)

Introduction ⇒ Describe the model by Alan Abramowitz
Data ⇒ Describe the data as outlined above, provide the R code for each step of your

analysis, use useful visualizations such as tables or charts
Model estimation ⇒ Describe the model estimation as outlined above, provide the R

code for each step of your analysis, interpret the results. This is very important! Don’t

just publish the numbers. Tell the reader what the numbers mean!
2020 forecast

Requirements of the project thesis

● The project report should have a length of approx. 5 to 7 pages, try to be brief.
● Follow the official APA style guidelines when reporting the results. For results from

regression analyses, use tables
● Font size, typography, quotation styles (if appropriate), and so forth, are to be applied

according to the scientific guidelines as commonly taught and practiced at Macromedia

University of Applied Sciences.
● Make sure to place value on how you present the results in an appealing way. For

example, put effort in a nice visualization of the data.
● The project report has to be submitted online (max. 20 MB).