Minimum R2 in Linear regression

[daisystat]

Is there a minimum R2 that needs to be deemed to be acceptable before looking towards any implications of the regression equation?
Many thanks in advance for any advice.

 

[Miner]

The answer varies depending on the field/domain of interest and the purpose of the regression.

In some domains the R^2 may be quite low, or if your interest is primarily understanding what IVs are important and whether they are positive or negative slopes. In other domains, the requirements are greater, or if your interest is in predictive models. I have been involved in embedded software algorithms where an R^2 of 0.99 was not good enough.

 

[Dason]

And there are fields where an R^2 of 0.2 is groundbreaking

 

[daisystat]

The answer varies depending on the field/domain of interest and the purpose of the regression.

In some domains the R^2 may be quite low, or if your interest is primarily understanding what IVs are important and whether they are positive or negative slopes. In other domains, the requirements are greater, or if your interest is in predictive models. I have been involved in embedded software algorithms where an R^2 of 0.99 was not good enough.

The answer varies depending on the field/domain of interest and the purpose of the regression.

In some domains the R^2 may be quite low, or if your interest is primarily understanding what IVs are important and whether they are positive or negative slopes. In other domains, the requirements are greater, or if your interest is in predictive models. I have been involved in embedded software algorithms where an R^2 of 0.99 was not good enough.

Thank you for your comments - really helpful.