Substitution elasticities between GHG-polluting and nonpolluting inputs in agricultural production: A meta-regression

Determining water use efficiency of wheat and cotton: A meta-regression analysis

“…The WLS is also reported to be robust in the presence of potentially omitted variables (Liu and Richard Shumway, 2016).…”

“…SQRT(n) is a measure based on the number of observations reported. Both measures are continuous and can be used to account for publication bias and deal with heteroscedasticity (Liu and Richard Shumway, 2016;Stanley et al, 2008).…”

“…To achieve efficient, consistent, unbiased and robust estimates, we need to examine multiple assumptions before estimating the regression models (Florax, 2002;Liu and Richard Shumway, 2016;Nelson and Kennedy, 2009). The studies included in the meta-analysis were distinct and independent.…”

“…However, independence of the cases may be a problem for cotton as only one mean WUE value was used in all but three studies that reported WUE using both irrigation methods. In this case, failure to account for correlation in the same study may cause underestimation of standard errors (Liu and Richard Shumway, 2016). The Lagrange multiplier (LM) test can evaluate this issue and has been used in meta-analysis (Florax et al, 2005).…”

“…The White test or Breusch-Pagan test can be used to detect this problem. To deal with heteroskedasticity, a WLS approach is commonly employed in meta-regression analysis (Liu and Richard Shumway, 2016;Stanley et al, 2008). Two kinds of weights are the inverse of the standard error and the square root of sample size in each empirical study (Florax et al, 2005;Stanley et al, 2008;Stern, 2012).…”