【摘要】Bootstrapping methods have so far been rarely used to evaluate spatial datasets. Basedon an extensive Monte Carlo study we find that also for spatial, cross-sectional data, the wild bootstraptest proposed by Davidson and Flachaire (2008) based on restricted residuals clearly outperformsasymptotic as well as competing bootstrap tests, like the pairs bootstrap.

【文献来源】Klarl T.Journal of Regional Science.2014（2）