School of Languages, Literatures, Linguistics and Media, Bangor University, Bangor, UK.
Kent Business School, University of Kent, Kent, UK.
Behav Res Methods. 2022 Jun;54(3):1319-1345. doi: 10.3758/s13428-021-01658-7. Epub 2021 Sep 10.
This paper introduces a novel approach to evaluate performance in the executive functioning skills of bilingual and monolingual children. This approach targets method- and analysis-specific issues in the field, which has reached an impasse (Antoniou et al., 2021). This study moves beyond the traditional approach towards bilingualism by using an array of executive functioning tasks and frontier methodologies, which allow us to jointly consider multiple tasks and metrics in a new measure; technical efficiency (TE). We use a data envelopment analysis technique to estimate TE for a sample of 32 Greek-English bilingual and 38 Greek monolingual children. In a second stage, we compare the TE of the groups using an ANCOVA, a bootstrap regression, and a k-means nearest-neighbour technique, while controlling for a range of background variables. Results show that bilinguals have superior TE compared to their monolingual counterparts, being around 6.5% more efficient. Robustness tests reveal that TE yields similar results to the more complex conventional MANCOVA analyses, while utilising information in a more efficient way. By using the TE approach on a relevant existing dataset, we further highlight TE's advantages compared to conventional analyses; not only does TE use a single measure, instead of two principal components, but it also allows more group observations as it accounts for differences between the groups by construction.
本文介绍了一种评估双语和单语儿童执行功能技能表现的新方法。该方法针对该领域中特定于方法和分析的问题,这些问题已经陷入僵局(Antoniou 等人,2021 年)。本研究通过使用一系列执行功能任务和前沿方法超越了传统的双语方法,这些方法使我们能够在新的衡量标准中共同考虑多个任务和指标;技术效率(TE)。我们使用数据包络分析技术对 32 名希腊语-英语双语和 38 名希腊语单语儿童的样本进行 TE 估计。在第二阶段,我们使用协方差分析(ANCOVA)、引导回归和 K 均值最近邻技术来比较两组的 TE,同时控制一系列背景变量。结果表明,双语者的 TE 优于单语者,效率高出约 6.5%。稳健性测试表明,TE 产生的结果与更复杂的传统 MANCOVA 分析相似,但以更有效的方式利用信息。通过在相关的现有数据集上使用 TE 方法,我们进一步强调了 TE 与传统分析相比的优势;TE 不仅使用单个度量标准,而不是两个主成分,而且由于它通过构建考虑到组之间的差异,因此允许更多的组观察。