In order to improve the accuracy of meta-analysis for benefit transfer, this research develops meta-regression analysis models to identify, measure and correct for publication selection bias in the recreation valuation literature, in particular, estimates of economic value and demand elasticities. These models are used to test several hypotheses across a variety of publication bias sources and to correct for publication bias. Monte Carlo simulations are used to evaluate how alternative corrective methods perform under realistic research conditions.
The three project objectives are to:
1. test for publication bias by publication type (e.g., journal vs. report, or publications that make methodological contributions vs. those that provide new estimates of values);
2. test for publication bias of recreation demand elasticities, including
i. on the distribution of estimated elasticities,
ii. on the generalizability of elasticity estimates,
iii. to measure the effect of moderator variables,
iv. and provide publication bias corrected elasticity estimates;
3. and evaluate and improve methods for publication bias detection and correction.
This research will expand an existing recreation valuation database through an extensive search of the literature and will update it with current studies and by extracting additional data from all studies. A variety of multivariate meta-regression analysis models will be developed to test the hypotheses associated with the defined research objectives, including equality of mean effect sizes across publication types, activities, regions, estimation methods, etc.; symmetry of effect size distributions; presence of a ‘true’ effect size; and moderator effects (study characteristics including activity type, region, estimation methods, etc.). Monte Carlo simulations will be used to test the efficiency, robustness, and power of existing publication bias correction methods and suggest improvements.
This research will provide a means for deriving valid and reliable estimates of recreation values from the literature using meta-regression analysis benefit transfers. Indirectly, this research will affect the entire domain of environmental valuation by providing improved methods of publication bias detection and correction that are transferable to all areas of resource valuation and empirical research.
methodological improvements, extent of the market, data collection methods, efficient bid designs, demand function, benefit function, empirical evidence, applied research, research synthesis, public policy, decision making, non-market valuation, contingent valuation, travel cost method, preferences, public good, socio-economic, willingness-to-pay, environmental assets, social science, modeling, measurement methods, econometric analysis, , Economic, Social, & Behavioral Science Research Program, Scientific Discipline, RFA, Social Science, decision-making, Economics & Decision Making, Ecology and Ecosystems, econometric analysis, demand elasticities, ecosystem valuation, public policy, decision analysis, decision making, bias, recreational value, benefits transfer, meta-regression analysis models, Monte Carlo study, public values, valuation