Download as pdf or txt
Download as pdf or txt
You are on page 1of 3

Practical Exercises

Nonlinear Regression

N.B. For each exercise, the working SPSS data set is indicated. The download link for
the data sets can be found in the section called “Download Your Resources Here”.

1
Data set: health_funding.sav

This data set contains information about the health care funding in 50 cities. Your task is
to study the relationship between the health funding (funding) and the number of
reported diseases per 10,000 inhabitants (disease). You suspect that this relationship is
nonlinear, and its form is quadratic or cubic.
1. Build a quadratic and a cubic regression model with these variables. (The health
care funding is supposed to be the dependent variable.)
2. Compare the prediction accuracy of these models. Which of them describes the
relationship most appropriately?
You can perform the tasks above by either linearizing the quadratic and cubic
equations, or by using the nonlinear regression procedure in SPSS.

2
Data set: album_sales.sav

This data set contains information about various music albums. You task is to study the
relationship between the album sales (Sales) and the number of plays on radio (airplay)
using both a quadratic and a cubic regression model.
Which of the two models gives the best prediction accuracy? (You either linearize the
quadratic and cubic equations, or use the nonlinear regression procedure in SPSS.).
3
Data set: car_sales.sav

This data set contains information about 157 car models. Your task is to identify the
best nonlinear model for the relationship between the car price (price) and engine size
(engine_s). To that effect, you have to do the following:
1. Build a quadratic and a cubic regression model with these variables.
2. Compare the prediction accuracy of these models. Which of them describes the
relationship most appropriately?
You can perform the tasks above by either linearizing the quadratic and cubic
equations, or by using the nonlinear regression procedure in SPSS.

4
Data set: car_sales.sav

Study the relationship between the car price (price) and fuel efficiency (mpg) using the
inverse function. You can either linearize this function, or use the nonlinear regression
procedure in SPSS.

5
Data set: exam_anxiety.sav

This data set contains information about the exam anxiety levels for 103 students. You
have to predict the anxiety level (Anxiety) based on the time spent revising (Revise)
using the inverse function. You can either linearize this function, or use the nonlinear
regression procedure in SPSS.
6
Data set: car_sales.sav

Study the relationship between the car price (price) and power (horsepower) using the
exponential function.

7
Data set: marketvalues.sav

This data set contains information about the purchase prices of 94 houses. Study the
relationship between price (value) and surface in square feet (sqft) using:
- the exponential function
- the logistic function

Which of these functions ensures the best prediction accuracy?

8
Data set: exam_anxiety.sav

Try predict the exam performance (Exam) based on the time spent revising (Revise)
using the logistic function (sigmoid).

You might also like