Cognitive Effort Discounting Task (COGED)

Background

The Cognitive Effort Discounting (COG-ED) Task is a behavioral paradigm that measures how mental effort reduces the subjective value of a reward. The research was spearheaded by Andrew Westbrook and Todd Braver in 2013 who sought to show how cognitive effort reduces the "utility" of a reward in a measurable way, similarly to what has been shown with Delay Discounting and studies looking at physical effort.

COG-ED procedures are usually divided into three phases:

1. Familiarization Phase: Participants get familiarized with the cognitive task and its various levels of cognitive exertion.
2. Choice Phase: The discounting procedure that pins the lower reward for performing an easy level of the cognitive task against a higher reward for a more strenuous level of the cognitive task and systematically adjusts the values up or down (based on choices made) to establish the amount at which a participant becomes indifferent ('Indifference Point') to the higher reward and views the smaller reward for an easier task just as desirable.
3. Reward Phase: Participants are asked to perform the cognitive task at the level and for the promised reward as indicated by random sample of their choices made during the choice phase.

The COG-ED procedure has been successfully used to establish COG-ED indifference points for rewards (gains) as well as for losses.

The most popular cognitive task used in the COG-ED paradigm is the strenuous n-back task assessing working memory. During an n-back procedure, participants see a constant stream of stimuli (e.g. letters), one stimulus after the other. Depending on the level of N, participants are asked to make a response anytime the current stimulus matches the stimulus presented N-positions before it. The higher N is, the more difficult the task becomes. For example, for N=1, the task is still pretty easy for most people. Participants are asked to make a response anytime the current stimulus matches the one preceding it (aka 'one position back'). For N=2, the task already gets much harder: Participants are asked to make a response anytime the current stimulus matches the one that was presented two positions before it (aka 'two positions back').

While the n-Back procedure is frequently used in COG-ED studies, other tasks such as the 'Speech task' by Drew McLaughlin, that varies the ease with which speech can be understood, are also run.

Task Procedure

After participants learn about the purpose of the study, they start working on the different levels of the cognitive task, e.g. the n-back task. Once all levels of the n-back task have been completed, the Choice ('Discounting') Procedure begins, followed by the Reward Procedure during which participants can earn a promised reward based on their actual choices.

Cognitive Task (Letter N-Back): Familiarization

Participants learn about the letter n-back procedure and its various levels (the default Millisecond implementation runs level 1-4). Each level of the letter n-back task is linked in the instructions to a specific color. For example, the BLACK task is linked to level 1 (N=1) and the RED task is linked to level 2 (N=2). The letter stimuli are then presented in these level-specific colors.

Each level runs one block of 20 trials (+ N start trials) with 5 target trials each. A new letter appears every 3.5s and is presented for 2s. Participants are asked to press the left 'S' key for each target and the 'K' key for any other letters. They have the entire 3.5s to respond.

Participants start with level 1 and proceed to level 4. At the end of each level they receive performance feedback. Additionally, at the end of each level, participants are asked to provide self-reports on the 6 rating scales of the NASA Task Load Index.

Choice Task (Discounting Procedure)

Instructions introduce the choice task in which participants have to decide between performing the Black task (N=1) for less money vs. performing one of the higher-level N tasks (e.g. the Red task) for more money. The lower amount of money for the Black task is continuously adjusted based on the participant's choice. Participants are told that one of their choice will be randomly sampled for the Reward Phase during which they then must perform the selected task for the promised money.

The choice task runs one practice session with 6 trials pinning the Black task (N=1) against the Red task (N=2). The Red task always promises a reward for $2; the amount for the Black task is continuously updated based on the participants' choices. Participants get 9s to make each choice. If no choice is made during that time frame, the black task is automatically selected. After 6 choices, practice ends. Choices made during the practice session don't qualify for the Reward Phase.

The test session of the choice task runs 54 trials, pinning the Black task (N=1) against level 2, level 3, and level 4 n-back tasks at three different reward points ($2, $3, $4) for 6 trials each.

Cognitive Task (Letter N-Back): Reward Phase

One of the choices made during the Choice Task is randomly selected. Participants need to perform five rounds of the selected level n-back task to earn the promised money. The Millisecond implementation of the COG-ED does not provide any behavioral measure of actual 'effort' during this phase.

What it Measures

COG-ED Tasks measure the subjective cost of cognitive effort

Psychological domains

• Working Memory: Ability to hold information and work with information in short term memory storage
• Need For Cognition: Personality trait that values cognitive stimulation
• Cognitive Motivation: General willingness to engage in cognitive work

Main Performance Metrics

• Indifference Points: Estimated indifference points for different levels of cognitive efforts

Psychiatric Conditions

COG-ED measures tend to be higher or lower in the following patient groups:

• Schizophrenia
• Major Depressive Disorder
• Attention Deficit Hyperactivity Disorder (ADHD)

Test Variations

References

Search Google Scholar for peer-reviewed, published research using the Inquisit Cognitive Effort Discounting Task (COGED).

---

Adjustments to reported z-scores of extreme hit and false alarm rates as described by: Gregg, A. & Sedikides, C. (2010). Narcissistic Fragility: Rethinking Its Links to Explicit and Implicit Self-esteem, Self and Identity, 9:2, 142-161 (p.148)

Westbrook A, Kester D, Braver TS (2013). What Is the Subjective Cost of Cognitive Effort? Load, Trait, and Aging Effects Revealed by Economic Preference. Pessiglione M, ed. PLoS One 981 8:e68210.

McLaughlin, D.J, Todd S. Braver,T.S., Peelle, J.E. (2020). Measuring the Subjective Cost of Listening Effort Using a Discounting Task.