2
$\begingroup$

I have conducted a systematic review and lme model of 50 studies attempting to economically value ecological services (previously un-valued). These studies do not use data samples or observations, instead they are using methods such as transferring the cost of artificial carbon production, or potential benefits from tourism to provide one or more $/hectare/year values per study (a total of 120 estimates). The overall goal is to establish whether method of valuation, or environmental factors explain the variance in these values.

The literature I have found on publication bias recommend regressing the t-ratio of each estimate against the remaining variables (e.g. Card and Krueger, 1995), but this seems reliant on the original studies using measured observations. I looked at the following question on this site, but no clear answer was given. Publication bias in meta-analysis of single case studies

Would forest and funnel plots be adequate in this situation if publication is the final goal?

$\endgroup$
  • $\begingroup$ Whenever you have a point effect estimate and a variability estimate (or sample size estimate), it is possible to appraise small study effects (a term which is preferred in general to publication bias). It is not clear whether you have both. Please, provide more data, and possibly the R data frame as well. $\endgroup$ – Joe_74 Mar 28 '16 at 23:30
  • $\begingroup$ I am voting to close this as unclear since the OP did not come back to clarify as had been requested in a comment. $\endgroup$ – mdewey Aug 15 '18 at 16:57