Which of the following is a drawback of using LPM models?
A) LPM models cannot handle dichotomous dependent variables.
B) LPM fitted values do not always fall within the range of 0 and 1.
C) LPM models require us to assume errors are normally distributed.
D) LPM models are complicated to interpret.

Question

Quizplus · Accepted Answer

Linear Probability Models are a simple way to estimate the relationship between a binary dependent variable and a set of independent variables. However, one of the key limitations of LPMs is that the predicted probabilities can fall outside the 0 to 1 range, which does not make sense for probabilities. This is because LPMs use ordinary least squares (OLS) regression, which does not inherently constrain the fitted values to be between 0 and 1. This can lead to nonsensical predictions, such as a negative probability of an event occurring or a probability greater than one. Alternative models like logistic regression or probit regression are often used to handle binary dependent variables because they produce predicted probabilities that are constrained to the 0 to 1 range.

Which of the Following Is a Drawback of Using LPM