I'm currently trying to interpret simple logistic regression with a categorical variable.
Description of variables:
"region" = the beneficiary’s residential area in the US; a factor with levels
"charges_cat"= which takes the value 0 (low) when charges are less than 10000 dollars and the value 1 (high) in all other cases.
> logm2<-glm(charges_cat~region, family=binomial) Coefficients: Estimate (Intercept) 0.01235 regionnorthwest -0.17888 regionsoutheast -0.13337 regionsouthwest -0.25351
My interpretation for b1 = regionnorthwest is:
> exp(-0.17888)  0.8362062
If the beneficiary is living in the northwest region of the US, the odds of charges being more than 10000 dollars is 16.38% lower than the odds of charges being more than 10000 dollars for a beneficiary who lives in the northeast region of the US, with southeast and southwest regions are fixed.
My question is; in a single logistic regression should I state the factor levels of the
"region" are fixed, such as "with southeast and southwest regions are fixed." or there is no need to state the dummy variables of the