Applied Effort Influence on Mental Workload Measures

Denys Bulikhov, Purdue University

Abstract

Some of the variability found in measures of mental workload (see e.g. Singleton et al. 1973; Wierwille and Connor 1983; Steelman-Allen et al. 2011; Casner and Gore 2010) may be due to the effort applied to the task by participants, rather than by the independent variable of interest. If true, capturing and removing the variation due to ‘applied effort’ could improve the ability of studies to detect effects of interest. While introducing participants to two sub-tasks derived from Multi-Attribute Task Battery II (Santiago-Espada et al. 2011), the study investigated the influence of applied effort on MATB-II performance measures of mental workload while holding other effects constant. Two groups of participants each completed easy and hard trials of MATB-II-derived sub-tasks. Treatment group of participants was offered an additional reward if they achieved a sufficiently high performance. The treatment group performed better by just under 4% in both easy and hard trials which provides a suggestion about the size of the effect of applied effort in this study. Additionally pilot error analysis was performed using Tracking task results. Error probability distributions did not fit known distributions and did not show consistent difference between treatment and control groups. Novel method of distribution “tails” comparison showed significant difference in extreme error durations, extents and delays between treatment and control groups. Measuring or controlling for applied effort can improve the ability of researchers to determine the effects of interventions on workload measures by reducing the amount of variability that is captured as error. Also, “tails” method seems to be a viable tool in comparing probability distributions which otherwise are similar and/or do not fit any known distributions.

Degree

Ph.D.

Advisors

Landry, Purdue University.

Subject Area

Physiology|Endocrinology|Mathematics|Neurosciences

COinS