The Tload-Dback task is a dual memory task that is time based. It is built on the Time based Resource Sharing Model from Camos and Barrouillet. The task consists of an alternation between a number parity task and a 1-back letter task. Participants are first trained on each task individually and then trained to perform the task in alternation, letter then number, etc. The task is devised to be conducted on 2 different occasions, a calibration phase and a cognitive load phase. The idea is that cognitive load is the result of placing a time constraint on information processing and reducing the time available to process ongoing information for continuous period of time. By applying a method of task switching with time pressure, the shared cognitive resources come under strain. This approach can be used to equalize the cognitive load of this task across participants to account for interindividual differences. This task is often used for cognitive fatigue induction. The benefit is that the high cognitive load condition is adaptable to meet the specific demands of a participant. This allows for creating a condition that does not underwhelm nor exceed the competencies of an individual and should allow for better interindividual comparisons. The calibration phase, carried out in 1 session, is done in a step-wise block format. Each block is incrementally faster than the prevoius block and this continues until a weighted accuracy performance drops below 85%. The performance score is percentage weighted as each task implies more or less cognitive difficulty. For this reason, number parity performance is weighted at 35% of the total score while the 1-back letter performance is weighted at 65%. The fastest or shortest presentation speed is considered to be the individual's maximum processing time for this task. However, by performing a working memory task, attentional resources are implicated. This could be problematic given that the task is carried out on two different days. As attentional resources are known to fluctuate through time due to a myriad of reasons, it is reasonable to ask whether a single calibration is sufficient for determining a maximum processing time. Since the tasks inception in 2017, it has been used in a variety of studies. No one, however, has tested whether or not a single calibration phase is sufficient or even reliable. This project aims to determine if 1 calibration phase is adequate or not, and whether the score obtained after 1 calibration is a reliable measure.