Skip to main content
Add response to comment.
Source Link
Adam Bailey
  • 8.8k
  • 10
  • 10

An advantage of having EconOverflow SE would be that it would hopefully then become non-contentious that Economics SE could be for those who study, teach and apply economics and econometrics, with low-level questions accepted (subject to conditions such as effort shown). In other words, a structure similar to MathOverflow SE and Mathematics SE.

In response to the question in the comment by @EconJohn, I offer the following examples relating to a topic I happen to have studied:

Question suitable for Economics SE (but towards the high end) In the context of estimating the recreational value of a national park using the zonal travel cost method, what methods might be used to allow for heteroscedasticity in estimating the trip-generating function, and what are their respective pros and cons?

Question suitable for EconOverflow I am trying to estimate the recreational value of a national park and have zonal travel cost data for 15 zones. With so few data points, conventional methods for addressing heteroscedasticity in estimating the trip-generating function (eg weighted least squares with weightings estimated from the data, heteroscedasticity-consistent standard errors) seem unlikely to work very well. An alternative approach proposed in Christensen & Price (1982) looks promising as its assumptions are independent of the number of zones. Which method might be expected to produce results with the least bias?

An advantage of having EconOverflow SE would be that it would hopefully then become non-contentious that Economics SE could be for those who study, teach and apply economics and econometrics, with low-level questions accepted (subject to conditions such as effort shown). In other words, a structure similar to MathOverflow SE and Mathematics SE.

An advantage of having EconOverflow SE would be that it would hopefully then become non-contentious that Economics SE could be for those who study, teach and apply economics and econometrics, with low-level questions accepted (subject to conditions such as effort shown). In other words, a structure similar to MathOverflow SE and Mathematics SE.

In response to the question in the comment by @EconJohn, I offer the following examples relating to a topic I happen to have studied:

Question suitable for Economics SE (but towards the high end) In the context of estimating the recreational value of a national park using the zonal travel cost method, what methods might be used to allow for heteroscedasticity in estimating the trip-generating function, and what are their respective pros and cons?

Question suitable for EconOverflow I am trying to estimate the recreational value of a national park and have zonal travel cost data for 15 zones. With so few data points, conventional methods for addressing heteroscedasticity in estimating the trip-generating function (eg weighted least squares with weightings estimated from the data, heteroscedasticity-consistent standard errors) seem unlikely to work very well. An alternative approach proposed in Christensen & Price (1982) looks promising as its assumptions are independent of the number of zones. Which method might be expected to produce results with the least bias?

Source Link
Adam Bailey
  • 8.8k
  • 10
  • 10

An advantage of having EconOverflow SE would be that it would hopefully then become non-contentious that Economics SE could be for those who study, teach and apply economics and econometrics, with low-level questions accepted (subject to conditions such as effort shown). In other words, a structure similar to MathOverflow SE and Mathematics SE.