Re: [Gretl-users] Is there a possibility to estimate a dynamic panel probit model in gretl?
Am 17.10.2014 um 12:55 schrieb Jan Tille:
Well, sorry, I thought I am, because I want to know whether being
above the median is followed by being above the median and that is
binary (i.e. an LDV), isn't it? Jan
Well even if it's called binary-dependent, I'd say it's not sufficient
to have (a) a binary variable, and (b) consider it as the dependent
variable. Typically the variable should also be observed as-is on your
unit of observation. In a latent-variable interpretation you typically
have an exogenous and pre-specified threshold that determines your
binary observable outcome.
Instead what you have is that in order to determine the outcome of the
binary variable you need to account for other units of observations as
well, because the median is of course a sample-dependent thing. That
looks non-standard to me.
hth,
sven