Hi,
I intend to run a probit selection model (likelihood of selecting into a sample) – to indicate that intention for future use of a product is only observed if endogenous variable (heard of product) is equal to one. However, in survey the respondents were asked if they had heard of a given product category, and then I generated a new variable which indicate whether a respondent had heard of at least one product. It looks like I have selection bias for each product category. How do I address this self selection bias? Thank you
I intend to run a probit selection model (likelihood of selecting into a sample) – to indicate that intention for future use of a product is only observed if endogenous variable (heard of product) is equal to one. However, in survey the respondents were asked if they had heard of a given product category, and then I generated a new variable which indicate whether a respondent had heard of at least one product. It looks like I have selection bias for each product category. How do I address this self selection bias? Thank you