Back to History Current Version

The effect of cognitive load on time perception and time preferences

Last registered on March 08, 2018

Pre-Trial

Trial Information

General Information

Title
The effect of cognitive load on time perception and time preferences
RCT ID
AEARCTR-0002772
Initial registration date
March 08, 2018

Initial registration date is when the trial was registered.

It corresponds to when the registration was submitted to the Registry to be reviewed for publication.

First published
March 08, 2018, 12:38 PM EST

First published corresponds to when the trial was first made public on the Registry after being reviewed.

Locations

Region

Primary Investigator

Affiliation
Lund University

Other Primary Investigator(s)

Additional Trial Information

Status
In development
Start date
2018-03-22
End date
2019-01-15
Secondary IDs
Abstract
This paper studies whether cognitive load affects time preferences through the channel of time perception. Experimental studies on the effect cognitive load has on time preferences show mixed results. According to dual system theory being under cognitive load while making decisions reduces the efficiency of the logical but slow system 2 in the brain, leading to a poorer system 2 decisions and more frequent use of the impulsive and fast system 1. This should lead to more impatient decisions, both because imagining the far future requires more cognitive power than imagining periods of time closer to the present, and because decisions governed by system 1 are probably highly present biased. However, digging deeper into the mechanisms coupled to cognitive load in the psychology literature, it has been shown that being under cognitive load is related to the shrinking of subjective time related to objective time. Given the assumption that time preferences are formed over subjective time, the shrinking of subjective time under cognitive load results in more patience in objective terms, assuming that time preferences over subjective time are fix. Hence, the effect of altered time perception under cognitive load goes in the opposite direction to the dual system theory effect, which might explain the mixed result in the experimental literature on the subject.

This study tests whether the effect of cognitive load on time preferences goes via the two proposed mechanisms. We conduct an experimental study where (i) the prospective time perception (ii) cognitive performance and (iii) time preferences are tested, all under varying degree of cognitive load induced by Stroop tasks. Our hypothesis is that increasing cognitive load decreases the ratio of subjective time to objective time, impairs performance on the cognitive performance task and that the sum of these two effects determines the total effect of cognitive load on time preferences.


In order to ensure identification and strengthen the temporal link between prospective time perception over a very short period of time and time preferences measured over a long period of time we perform a second experiment where changes in cognitive performance, time perception, time preferences as well as in future perspective (estimated by a scrambled sentence task) are induced with varying degree of arousal induced by external tempo.
External Link(s)

Registration Citation

Citation
Hardardottir, Hjördis. 2018. "The effect of cognitive load on time perception and time preferences." AEA RCT Registry. March 08. https://doi.org/10.1257/rct.2772-1.0
Former Citation
Hardardottir, Hjördis. 2018. "The effect of cognitive load on time perception and time preferences." AEA RCT Registry. March 08. https://www.socialscienceregistry.org/trials/2772/history/26399
Experimental Details

Interventions

Intervention(s)
Intervention Start Date
2018-03-22
Intervention End Date
2018-06-15

Primary Outcomes

Primary Outcomes (end points)
Treatment 1: Changes (within subject) in estimated time preferences, prospective time perception and performance on a numeracy task when the level of competing cognitive load increases. Treatment 2: Changes (within subject) in estimated time preferences, prospective time perception, performance on a numeracy task and in a future time perspective task when the level of arousal induced by external tempo (metronome) increases.
Primary Outcomes (explanation)
Time preferences are measured using a double multiple price list.

Time perception is measured as reported prospective time perception. That is, subjects are told that when they press a button a time interval begins and that after the interval they will be asked to estimate the length of the interval (with monetary incentives to be close to the correct time).

Numeracy is measured using multiplication tasks of increasing complexity. Subjects get paid depending on how many correct answers they get and how quickly, compared to other participants in the session, they answer the tasks.

For the future time perspective index, subjects unscramble sentences that can be made either future oriented or present oriented. (E.g. I usually/rarely think about the future). Their future time perspective index is the share of future oriented sentences constructed.

Secondary Outcomes

Secondary Outcomes (end points)
Secondary Outcomes (explanation)

Experimental Design

Experimental Design
Treatment 1:
Cognitive load is induced throughout the treatment with Stroop tasks. There are three difficulty levels, baseline, congruent and dissimilar. In the baseline Stroop tasks, subject are presented with a name of a color, written in letters of that same color. To the right of the word there is a bullet where the name of the color, again, is written. Subject should simply choose the bullet to successfully complete a baseline Stroop task. In the congruent Stroop tasks, subjects are presented with a name of a color, written in letters of that same color. To the right of the word there is a list of six bullets, each representing a color. The subject should choose the bullet that corresponds to the word on the left. In the dissimilar Stroop tasks, subjects are presented with a name of a color, written in letters of a different color. To the right of the word there is a list of six colors. The subject should choose the color that corresponds to the color of the letters of the word on the left.

All subjects perform all tasks, time perception, time preferences elicitation and numeracy tasks with all three types of Stroop tasks (in a group of 2, 3 or 4 tasks at the time) generating a disturbing cognitive load.

In the time perception task, Stroop tasks are solved during the interval that is to be estimated. In the time preferences elicitation and the numeracy tasks, the screen is divided in two. The multiple price list/multiplication task appears on the left side while on the right side Stroop tasks (2, 3 or 4 at the time) pop up randomly (arrival determined by a poisson process).

The design is within-subject. Hence all subject do all tasks with all three levels of cognitive load.

Time perception is measured over three intervals. A random number between 7 and 12 seconds, a random number between 13 and 18 seconds and a random number between 19 and 24 seconds, all intervals with all three levels of cognitive load induced by Stroop tasks.

Time preferences are measured using a double multiple price list. There are 4 different versions of the MPL in terms of time: today vs. in one week, today vs. in three weeks, in one week vs. in three weeks and in three weeks vs. in six weeks. For each of the four time schemes, there are two versions of the MPL, one with a larger range than the other. In order to control for changes in concavity of the utility function as cognitive load increases, risk preferences are measured using MPL in the baseline Stroop condition and the dissimilar Stroop (high cognitive load). Here also, are two lists of different range presented in each of the two Stroop conditions.


Treatment 2:
Time preferences, time perception, performance on numeracy tasks and future time perspective are measured under two levels of arousal induced by external tempo generated by a metronome (in headphones).

The method for measuring time and risk preferences and numeracy are the same as in treatment 1, except that there is now no Stroop tasks popping up, instead, subjects listen two tempo of either 120 Bpm or 165 Bpm in headphones while solving the tasks.

Again, the design is within-subject, hence all subjects perform all tasks with both high and low tempo.

In order to avoid counting time during the time perception task, subjects solve a series of baseline Stroop tasks during the time intervals that are to be estimated.

The future time perspective is measured using a scrambled sentence task. Under time pressure the subject have to unscramble words and form sentences. In all tasks, one can either construct a future oriented sentence or a present oriented sentence. The share of future oriented sentences is the subjects measure of future time perspective. All subjects do two versions of the future time perspective task, one with 120 bpm in the headphones, the other with 165 bpm in the headphones. Which version is coupled with which beat is randomized.

For both treatment 1 and treatment 2, the order of tasks is randomized for each session.

Finally, at the end of the session, subjects are asked to estimate how long the experiment has taken.

Background questions that will be collected are individual information (age, gender, income, family status), self-reported attitudes towards risk preferences, time preferences and future time perception.
Experimental Design Details
Randomization Method
The design is within subject and treatment 1 and treatment 2 are independent. Hence, there will be no randomization into treatments.

The order of the tasks is randomized (by a computer) for each session. For the time preference and risk preference elicitation, there are three versions of the MPL in treatment 1 and two versions in treatment 2. Which of the lists is matched with which version is randomized by a computer for each subject.

The length of the intervals to be estimated in the time perception tasks is always in the intervals [8,12], [13,17] and [18,24], but the exact number of second is randomly drawn by a computer for each subject.
Randomization Unit
Individual
Was the treatment clustered?
No

Experiment Characteristics

Sample size: planned number of clusters
300 individuals
Sample size: planned number of observations
300 individuals
Sample size (or number of clusters) by treatment arms
150 in each treatment
Minimum detectable effect size for main outcomes (accounting for sample design and clustering)
IRB

Institutional Review Boards (IRBs)

IRB Name
IRB Approval Date
IRB Approval Number

Post-Trial

Post Trial Information

Study Withdrawal

There is information in this trial unavailable to the public. Use the button below to request access.

Request Information

Intervention

Is the intervention completed?
No
Data Collection Complete
Data Publication

Data Publication

Is public data available?
No

Program Files

Program Files
Reports, Papers & Other Materials

Relevant Paper(s)

Reports & Other Materials