Do All Switches Cost the Same? Reliability of Language Switching and Mixing Costs.

Dorit Segal,Anat Prior,Tamar H Gollan

doi:10.5334/joc.140

Dorit Segal, Anat Prior + Show 1 more

Open Access

https://doi.org/10.5334/joc.140

Copy DOI

Abstract

The current study examined the reliability and consistency of switching and mixing costs in the language and the color-shape tasks in three pre-existing data sets, to assess whether they are equally well suited for the study of individual differences. Specifically, we considered if the language task is as reliable as the color-shape task – an important question given the wide use of language switching tasks but little information available to address this question. Switching costs had low to moderate reliability and internal consistency, and these were similar for the language and the color-shape tasks. Mixing costs were more reliable in the language task than in the color-shape task when tested twice on the same day and trended in the same direction when tested a week apart. In addition, mixing costs were larger and more consistent than switching costs in all data sets and they were also were more reliable than switching costs in the language task when tested on the same day. These results reveal the language task to be as good as the color-shape task for measuring switching and mixing ability. Low variability of switching costs may decrease their reliability and consistency, in turn interfering with the chance of detecting cross task correlations. We advocate for exploring procedures to increase the variability of switching costs, which might increase reliability and consistency of these measures, and improve the ability to determine if bilingual language use relies on cognitive mechanisms that overlap with those underlying nonlinguistic multi-tasking.

Highlights

One of the most extraordinary aspects of the human mind is the ability to execute two concurrent tasks – people can walk and talk at the same time or listen to music while reading a book
The current study examined test-retest reliability and internal consistency of the commonly used language switching task and compared it to the reliability and consistency of the color-shape task
Mixing costs were more consistent than switching costs in all studies and in both tasks

Summary

Introduction

One of the most extraordinary aspects of the human mind is the ability to execute two concurrent tasks – people can walk and talk at the same time or listen to music while reading a book. Weissberger, Wierenga, Bondi, and Gollan (2012) found that older bilinguals who could not complete the color-shape task exhibited larger language-switching costs than matched bilinguals who were able to do both tasks While these findings imply that language and task switching tap the same mechanism, supporting the existence of a domaingeneral switching mechanism, other studies report quite different results. There is evidence of age-related decline in non-linguistic task switching but not in language switching tasks, and switching costs for a linguistic and a non-linguistic task were not correlated in young (Timmer, Calabria, & Costa, 2019), middle-aged or older bilingual adults (Calabria, Branzi, Marne, Hernandez, & Costa, 2015) These findings directly conflict with those presented above, and imply the opposite conclusion, i.e., that language switching is supported by language-specific switching mechanisms (See de Bruin, Treccani, & Della Sala, 2015; Paap & Greenberg, 2013 and Paap, Johnson, & Sawi, 2015 for similar conclusions). Reliability affects mostly correlational studies because a correlation between measures will always be lower than the reliability of the measures (Cohen, Cohen, West & Aiken, 2003)

Methods

Results

Conclusion