Asked Dec 1, 2019

Discuss the reasons and situations in which researchers would want to use linear regression. How would a researcher know whether linear regression would be the appropriate statistical technique to use? What are some of the benefits of fitting the relationship between two variables to an equation for a straight line?


Expert Answer

1 Rating
Step 1

Reasons and situations to use linear regression:

A linear regression is suitable in the following situations:

  • If a plot of the response variable and the explanatory variable show a linear relationship;
  • If the residuals are homoscedastic;
  • If the residuals are independent of one another;
  • If the residuals have a normal distribution (or can be manipulated to attain an approximate normal distribution).
Step 2

How to decide whether linear regression should be used:

The easiest and most suitable method to determine whether a linear regression is suitable in a given situation, is by using a graphical plot. When a dataset is available with observations on the potential response variable, with those of the corresponding explanatory variable, a researcher should draw a scatterplot of the response variable versus the explanatory variable. If the plot seems quite linear, then it is suitable to fit a linear regression model to the data. If the plot is non-linear but in such a pattern that a transformation (such as, logarithmic, square, cubic, square root, cube root, etc.) on either one or both the response and the explanatory v...

Want to see the full answer?

See Solution

Check out a sample Q&A here.

Want to see this answer and more?

Solutions are written by subject experts who are available 24/7. Questions are typically answered within 1 hour.*

See Solution
*Response times may vary by subject and question.
Tagged in




Related Statistics Q&A

Find answers to questions asked by student like you
Show more Q&A

Q: Do males and females differ in the amount of time they talk on the phone? A study reported that wome...

A: a. The null and the alternative hypotheses for the given scenario are given below:


Q: Assume that on a standardized test of 100 questions, a person has a probability of 80% of answering ...

A: It is given that the probability of answering any particular question is 0.80. The total number of s...


Q: An electronics retailer would like to investigate the relationship between the selling price of a di...

A: The null and the alternative hypotheses using for testing the correlation coefficient is:


Q: Number 16

A: Given data


Q: defcaluvulate the mean ang standard deviations of the distribution

A: The tree diagram is obtained below:


Q: Use a normal approximation to find the probability of the indicated number of voters. In this case, ...

A: From the given information, there are 170 eligible voters aged 18-24 and the percentage of eligible ...


Q: A lightbulb is advertised as lasting 1000 hours with a standard deviation of 150 hours. Assume the l...

A: It is given that the mean and standard deviation are 1,000 and 150, respectively.


Q: The null and alternate hypotheses are: H0 : μ1 = μ2H1 : μ1 ≠ μ2 A random sample of 12 observations f...

A: Hypotheses for the test is given below:


Q: To investigate the tradeoff between time to work and time to sleep, the following equation is estima...

A: 6 1. Testing for independent variable “educ”: State the null and alternative Hypotheses: Null Hypoth...