5

I would like to determine the factors affecting a persons decision to buy a new car. Upon inspecting the data, I have found that things such as:

  • Number of people in the household
  • Cost of tax
  • Income
  • Gender
  • Occupation

all are significant in decision making.

I would like to build a model to quantify how a decision is affected, if these parameters change. For example, how would a persons decision be affected if their income rose by £10,000 per annum, or they had a new baby?

I need to build some sort of linear model. I would like to end up with a list of coefficients for each significant variable, which will give me what I need, but wondered what approach I needed to take.

The data I am using has been split into two, and is binary (have bought a new car - 1, or have not bought a new car - 0)

sym246
  • 417
  • 4
  • 13

1 Answers1

1

You can model this with a multiple logistic regression. Standard logistic regression models the probability of a dependent binary variable as a function of one or more independent variables. The logistic tag has a number of relevant questions that may help you, such as this one.

mkt
  • 11,770
  • 9
  • 51
  • 125