True or False: Simple linear regression is suitable for a binary dependent variable.

True or False: Simple linear regression is suitable for a binary dependent variable.

• A. True
• B. False
• C. Only sometimes
• D. It depends on the data

Answer: (B)

Simple linear regression is indeed not suitable for a binary dependent variable. In linear regression, the dependent variable is assumed to be continuous and normally distributed. When the dependent variable is binary—taking on only two possible values (e.g., yes/no, success/failure)—this assumption is violated. Using linear regression in this scenario can lead to several issues. The predicted values may fall outside the range of 0 and 1, which is not appropriate for probabilities associated with binary outcomes. This can result in nonsensical interpretations of the data and inaccuracies in the model. To analyze binary outcomes, logistic regression is typically employed instead. Logistic regression provides a method to model the probability of the dependent variable being in one category versus the other, catering specifically to the nature of binary data. This approach correctly constrains the predictions to fall within the 0-1 interval, making it more appropriate for handling binary outcomes.

Simple linear regression is indeed not suitable for a binary dependent variable. In linear regression, the dependent variable is assumed to be continuous and normally distributed. When the dependent variable is binary—taking on only two possible values (e.g., yes/no, success/failure)—this assumption is violated.

Using linear regression in this scenario can lead to several issues. The predicted values may fall outside the range of 0 and 1, which is not appropriate for probabilities associated with binary outcomes. This can result in nonsensical interpretations of the data and inaccuracies in the model.

To analyze binary outcomes, logistic regression is typically employed instead. Logistic regression provides a method to model the probability of the dependent variable being in one category versus the other, catering specifically to the nature of binary data. This approach correctly constrains the predictions to fall within the 0-1 interval, making it more appropriate for handling binary outcomes.