Working memory resource depletion effect in academic learning. An integrated approach – WM-RDE

Learning at school requires cognitive effort. Optimizing these efforts is a key to academic learning achievement. Several thousand published experiments, within the cognitive load theory framework, have identified the factors that contribute to this optimization. The temporal dimension of this optimization has been evoked very recently: certain academic learning tasks could exhaust students, resulting on learning impairment. The hypothesis of Sweller and other authors from cognitive load theory is that working memory resources would be depleted during a demanding learning task. But these authors do not say: What mechanisms govern this exhaustion? How can this depletion be measured? The working memory resource depletion effect (WM-RDE) project proposes to answer these questions. For that we use:
1. A working memory model, the Barrouillet & Camos’ Time-Based Resource Sharing model (TBRS), which allows us to make accurate hypotheses on the depletable resource in memory (controlled attention) and on the cause of this depletion (the temporal sharing of resources between processing and refreshing subtasks during a long period). Puma and Tricot have previously shown that TBRS model fits well with academic learning tasks and with cognitive load theory.
2. A model of cognitive fatigue recently proposed by André, Audiffren & Baumeister: cognitive fatigue corresponds to a progressive weakening of the connectivity within and between large-scale neuronal networks involved in effortful control. This phenomenon could be caused by an accumulation of adenosine in overloaded brain regions.
Cognitive load theory, TBRS and the new cognitive fatigue model will be taken into consideration in the present project by controlling different variables (e.g., motivation, cost/benefit of effortful control, time sharing between subtasks) and measuring other variables, during an academic learning task involving processing and refreshing information (e.g. listening and transcribing a foreign language sound file during 35 minutes).
3. A multi-measure approach to resource mobilization and depletion: we will use subjective, behavioral, physiological and neurophysiological measures. As our main hypothesis is that the depleted resources are WM resource, we will also perform a direct an original measure of depletion: the difference between a WM capacity test (n-back) before and just after the leaning task. Complementarity between all these indexes allows us to interpret the results.
The WM-RDE project involves six different sites: Montpellier, Poitiers, Toulouse-Albi, Bordeaux, Nimes and Cergy. Therefore, we will be able to run 20 experiments to test our hypotheses.
First, our aim is to consolidate the resource depletion effect at the behavioral level. We will run a six-site pre-registered study using the same protocol, to replicate the resource depletion effect (1 experiment, replicated in 6 different labs). Then, we will realize a multi-site study using the same protocol but with a different depleting task on each site (6 experiments, 1 per lab).
Second, we will look for physiological correlates of the WM resource depletion effect, based on a strict replication of the first experience (or one of the 6 others if better). The same experiment measuring physiological variables will be performed in the three sites already equipped with physiological measures tools (1 experiment, replicated in 3 different labs).
Third, we will replicate the experiment to evaluate how WM is solicited, by focusing on WM parameters (i.e. changing cognitive load in WM), and after that, by distinguishing between static and dynamic measures of WM (6 experiments, in 4 different labs).
Four, we will try to generalize our results by changing task, population (with pupils declaring fatigue), and level of engagement / motivation (6 experiments, 1 per lab).
Five, we will disseminate our results to the scientific community on the one hand and to education professionals on the other.