Intra-Individual Variability from a Lifespan Perspective: A Comparison of Latency and Accuracy Measures.

Delphine Fagot,Thierry Lecerf,Paolo Ghisletta,Erika Borella,Anik De Ribaupierre,Nathalie Mella

doi:10.3390/jintelligence6010016

Abstract

Within-task variability across trials (intra-individual variability (IIV)) has been mainly studied using latency measures but rarely with accuracy measures. The aim of the Geneva Variability Study was to examine IIV in both latency and accuracy measures of cognitive performance across the lifespan, administering the same tasks to children, younger adults, and older adults. Six processing speed tasks (Response Time (RT) tasks, 8 conditions) and two working memory tasks scored in terms of the number of correct responses (Working Memory (WM)—verbal and visuo-spatial, 6 conditions), as well as control tasks, were administered to over 500 individuals distributed across the three age periods. The main questions were whether age differences in IIV would vary throughout the lifespan according (i) to the type of measure used (RTs vs. accuracy); and (ii) to task complexity. The objective of this paper was to present the general experimental design and to provide an essentially descriptive picture of the results. For all experimental tasks, IIV was estimated using intra-individual standard deviation (iSDr), controlling for the individual level (mean) of performance and for potential practice effects. As concerns RTs, and in conformity with a majority of the literature, younger adults were less variable than both children and older adults, and the young children were often the most variable. In contrast, IIV in the WM accuracy scores pointed to different age trends—age effects were either not observed or, when found, they indicated that younger adults were the more variable group. Overall, the findings suggest that IIV provides complementary information to that based on a mean performance, and that the relation of IIV to cognitive development depends on the type of measure used.

Highlights

Almost all research in cognitive psychology, including cognitive developmental psychology, has focused on mean performance
Descriptive statistics for the different response time (RT) tasks used are presented in Table 3; Table A2 reports the statistics for the five age groups
Analyses on the coefficient of variation (CV) have been conducted, the results on the reaction times (RTs) score are substantially the same as those obtained with the standard deviations

Summary

Introduction

Almost all research in cognitive psychology, including cognitive developmental psychology, has focused on mean performance. Individual variability should not be neglected, whether inter- or intra-individual. Many authors have proposed to consider that individual variability should be taken into consideration when proposing general laws of behavior (e.g., Reference [1]). General psychology should be able to integrate the bulk of individual variability, just as much as differential psychology. Several types of individual variabilities should be distinguished (e.g., Reference [3]). Inter-individual variability has been often acknowledged, for instance when comparing groups or individuals from an applied perspective, or when standard deviations are reported. That standard deviations only provide information about quantitative differences among individuals, as if they differed solely in how distant from their group mean they are; they do not provide any indication on qualitative differences

Objectives

Methods

Results

Discussion

Conclusion