Offloading under cognitive load: Humans are willing to offload parts of an attentionally demanding task to an algorithm
Open accessThe DOI matches its Crossref record; 28 of the 28 references it cites were found in a registry or an index.
Abstract
In the near future, humans will increasingly be required to offload tasks to artificial systems to facilitate daily as well as professional activities. Yet, research has shown that humans are often averse to offloading tasks to algorithms (so-called “algorithmic aversion”). In the present study, we asked whether this aversion is also present when humans act under high cognitive load. Participants performed an attentionally demanding task (a multiple object tracking (MOT) task), which required them to track a subset of moving targets among distractors on a computer screen. Participants first performed the MOT task alone (Solo condition) and were then given the option to offload an unlimited number of targets to a computer partner (Joint condition). We found that participants significantly offloaded some (but not all) targets to the computer partner, thereby improving their individual tracking accuracy (Experiment 1). A similar tendency for offloading was observed when participants were informed beforehand that the computer partner’s tracking accuracy was flawless (Experiment 2). The present findings show that humans are willing to (partially) offload task demands to an algorithm to reduce their own cognitive load. We suggest that the cognitive load of a task is an important factor to consider when evaluating human tendencies for offloading cognition onto artificial systems.
Overview
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
Integrity signals
Registry
Cites
Statistics
Figures






Verification
Claims the paper makes
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
The source backs part of the claim, not all of it — or backs it for a different population, condition or comparison.
In the citing paper
“Empirical evidence has shown that students who strategically use digital supports often report higher confidence in managing academic tasks, especially in complex learning environments (Wahn et al., 2023; Zou et al., 2025).”
From Cognitive offloading through digital tools and its relationship with critical thinking, task persistence, and learning depth (2026) · cites this paper as Wahn et al., 2023
Experimental — accuracy not yet measured
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
The self-check needs the paper’s full text.
Citations
Cites
Coefficient Alpha and the Internal Structure of Tests
Cronbach, L. J. (1951) · Psychometrika · doi:10.1007/bf02310555 · Cited by 43,931 · cited 1×
“For this purpose, we combined the data from our two experiments and first evaluated the reliability of all our questionnaire scales using Cronbach’s alpha [29].”
MethodStatistical power analyses using G∗Power 3.1: tests for correlation and regression analyses
Faul, F., Erdfelder, E., Buchner, A., & Lang, A.-G. (2009) · Behavior Research Methods · doi:10.3758/brm.41.4.1149 · Cited by 37,615 · cited 2×
“Sample size was determined using G*Power [24] (alpha = 0.”
MethodInfluentialThe Use of Cronbach’s Alpha When Developing and Reporting Research Instruments in Science Education
Taber, K. S. (2018) · Research in Science Education · doi:10.1007/s11165-016-9602-2 · Cited by 10,679
Algorithm aversion: People erroneously avoid algorithms after seeing them err.
Dietvorst, B. J., Simmons, J. P., & Massey, C. (2015) · Journal of Experimental Psychology: General · doi:10.1037/xge0000033 · Cited by 3,232 · cited 3×
“With regard to the second point, if people observe an algorithm making mistakes, their attitude can quickly change from appreciation to aversion [11].”
BackgroundTracking multiple independent targets: Evidence for a parallel tracking mechanism*
Pylyshyn, Z. W., & Storm, R. W. (1988) · Spatial Vision · doi:10.1163/156856888x00122 · Cited by 1,906 · cited 1×
“We addressed this question by conducting two behavioral experiments in which participants performed a multiple object tracking (MOT) task [15].”
MethodRisko, E. F., & Gilbert, S. J. (2016) · Trends in Cognitive Sciences · doi:10.1016/j.tics.2016.07.002 · Cited by 1,502 · cited 5×
“On the other hand, given that the present task is attentionally demanding, one could predict that humans will (partially) offload this task to an algorithm to reduce their own cognitive effort [1]–in line with the offloading tendencies humans show nowadays in certain daily activities (e.”
BackgroundResultInfluentialBonaccio, S., & Dalal, R. S. (2006) · Organizational Behavior and Human Decision Processes · doi:10.1016/j.obhdp.2006.07.001 · Cited by 1,011
Capacity limits of information processing in the brain
Marois, R., & Ivanoff, J. (2005) · Trends in Cognitive Sciences · doi:10.1016/j.tics.2005.04.010 · Cited by 928 · cited 1×
“As such, offloading can help humans to overcome cognitive capacity limitations [2, 3], enabling them to attain goals that could not have been attained (as quickly, easily, or efficiently) otherwise.”
BackgroundFranke, T., Attig, C., & Wessel, D. (2019) · International Journal of Human–Computer Interaction · doi:10.1080/10447318.2018.1456150 · Cited by 828 · cited 2×
“The set consisted of the Desirability of Control Scale (20 items [25]), the Trust in Automation Scale (3-items-subset [26]), and the Affinity for Technological Systems Scale (9 items [27]).”
MethodResultBurger, J. M., & Cooper, H. M. (1979) · Motivation and Emotion · doi:10.1007/bf00994052 · Cited by 806 · cited 2×
“The set consisted of the Desirability of Control Scale (20 items [25]), the Trust in Automation Scale (3-items-subset [26]), and the Affinity for Technological Systems Scale (9 items [27]).”
MethodResultPeople are averse to machines making moral decisions.
Bigman, Y. E., & Gray, K. (2018) · Cognition · doi:10.1016/j.cognition.2018.08.003 · Cited by 738 · cited 6×
“For instance, in the latter case, participants were asked in one study [10] whether it is more permissible that a medical decision (i.”
InfluentialThe Spatial Resolution of Visual Attention
Intriligator, J., & Cavanagh, P. (2001) · Cognitive Psychology · doi:10.1006/cogp.2001.0755 · Cited by 733 · cited 1×
“Studies showed that humans are able to track a limited number of objects (four objects [20]; but also see: [21]) and that tracking is an effortful task that requires sustained attention over prolonged periods of time (e.”
BackgroundHow many objects can you track?: Evidence for a resource-limited attentive tracking mechanism
Alvarez, G. A., & Franconeri, S. L. (2007) · Journal of Vision · doi:10.1167/7.13.14 · Cited by 493 · cited 2×
“Note that in the Solo condition, participants tracked between 3 and 4 targets on average, indicating that they tracked (close to) the maximum number of four targets humans are typically capable of tracking [21].”
BackgroundAlnaes, D., Sneve, M. H., Espeseth, T., Endestad, T., van de Pavert, S. H. P., & Laeng, B. (2014) · Journal of Vision · doi:10.1167/14.4.1 · Cited by 483
Theoretical Considerations and Development of a Questionnaire to Measure Trust in Automation
Körber, M. (2019) · Advances in Intelligent Systems and Computing · doi:10.1007/978-3-319-96074-6_2 · Cited by 317 · cited 3×
“The set consisted of the Desirability of Control Scale (20 items [25]), the Trust in Automation Scale (3-items-subset [26]), and the Affinity for Technological Systems Scale (9 items [27]).”
MethodResultIs Attentional Resource Allocation Across Sensory Modalities Task-Dependent?
Wahn, B., & König, P. (2017) · Advances in Cognitive Psychology · doi:10.5709/acp-0209-2 · Cited by 164 · cited 1×
“As such, offloading can help humans to overcome cognitive capacity limitations [2, 3], enabling them to attain goals that could not have been attained (as quickly, easily, or efficiently) otherwise.”
BackgroundSelf-citationJago, A. S. (2019) · Academy of Management Discoveries · doi:10.5465/amd.2017.0002 · Cited by 160 · cited 2×
“Another study [12] investigated how humans judge the authenticity of works (e.”
BackgroundWhat Have We Learned about Attention from Multiple-Object Tracking (and Vice Versa)?
Scholl, B. J. (2009) · The MIT Press eBooks · doi:10.7551/mitpress/8135.003.0005 · Cited by 123
Hou, Y. T.-Y., & Jung, M. F. (2021) · Proceedings of the ACM on Human-Computer Interaction · doi:10.1145/3479864 · Cited by 105 · cited 1×
“Further factors include the prior description of an algorithm (see recent review [13]) and various kinds of affective influences such as whether a robot has human facial features (like iCub) or not.”
BackgroundPupil Sizes Scale with Attentional Load and Task Experience in a Multiple Object Tracking Task
Wahn, B., Ferris, D. P., Hairston, W. D., & Peter König, B. (2016) · PLoS ONE · doi:10.1371/journal.pone.0168087 · Cited by 96
Self-citationIt’s a Match: Task Assignment in Human–Robot Collaboration Depends on Mind Perception
Wiese, E., Weis, P. P., Bigman, Y., Kapsaskis, K., & Gray, K. J. (2021) · International Journal of Social Robotics · doi:10.1007/s12369-021-00771-z · Cited by 69 · cited 4×
“aversion may depend on the (ascribed) reliability [5] and expertise of an algorithm [7, 8, 10], we also explored whether the expertise factor would affect participants’ behavior in the present study.”
BackgroundInfluentialUsing tools to help us think: actual but also believed reliability modulates cognitive offloading
Weis, P. P., & Wiese, E. (2019) · Human Factors: The Journal of the Human Factors and Ergonomics Society · doi:10.1177/0018720818797553 · Cited by 49 · cited 4×
“aversion may depend on the (ascribed) reliability [5] and expertise of an algorithm [7, 8, 10], we also explored whether the expertise factor would affect participants’ behavior in the present study.”
BackgroundMethodInfluentialProblem Solvers Adjust Cognitive Offloading Based on Performance Goals
Weis, P. P., & Wiese, E. (2019) · Cognitive Science · doi:10.1111/cogs.12802 · Cited by 38
Know Your Cognitive Environment! Mental Models as Crucial Determinant of Offloading Preferences
Weis, P. P., & Wiese, E. (2022) · Human Factors: The Journal of the Human Factors and Ergonomics Society · doi:10.1177/0018720820956861 · Cited by 29 · cited 4×
“aversion may depend on the (ascribed) reliability [5] and expertise of an algorithm [7, 8, 10], we also explored whether the expertise factor would affect participants’ behavior in the present study.”
BackgroundInfluentialCan Limitations of Visuospatial Attention Be Circumvented? A Review
Wahn, B., & König, P. (2017) · Frontiers in Psychology · doi:10.3389/fpsyg.2017.01896 · Cited by 16 · cited 3×
“By sharing a task with an algorithm, humans could not only reduce cognitive effort and thereby increase individual accuracy but could also go beyond their own capacity limitations [14] and thus achieve an overall performance level that exceeds the level they would achieve if they performed the entire task alone.”
BackgroundInfluentialSelf-citationHumans share task load with a computer partner if (they believe that) it acts human-like
Wahn, B., & Kingstone, A. F. (2020) · Acta Psychologica · doi:10.1016/j.actpsy.2020.103205 · Cited by 14 · cited 1×
“We chose this task because it has been reliably used in earlier research to test the limitations of attentional processing [16] and to investigate human-human [17] as well as human-AI collaborations [18, 19].”
MethodSelf-citationCoordination effort in joint action is reflected in pupil size
Wahn, B., Ruuskanen, V., Kingstone, A. F., & Mathôt, S. (2021) · Acta Psychologica · doi:10.1016/j.actpsy.2021.103291 · Cited by 8
Self-citationInterpersonal coordination in joint multiple object tracking.
Wahn, B., König, P., & Kingstone, A. F. (2021) · Journal of Experimental Psychology Human Perception & Performance · doi:10.1037/xhp0000935 · Cited by 6 · cited 2×
“We addressed this possibility by drawing on data from an earlier study conducted by the first author ([17]; see “No Information” condition).”
MethodSelf-citation
Research works printed but not linked: 1
- E Jussupow, 2020, Why are we averse toward algorithms? A comprehensive literature review on algorithm aversion., In Proceedings of the 28th European Conference on Information Systems (ECIS), An Online AIS Conference
Cited by
Chen, A., Xiang, M., Zhou, J., Jia, J., Shang, J., Li, X., Gašević, D., & Fan, Y. (2025) · Computers & Education · doi:10.1016/j.compedu.2024.105198 · Cited by 115
The Cognitive Architecture of Digital Externalization
Skulmowski, A. (2023) · Educational Psychology Review · doi:10.1007/s10648-023-09818-1 · Cited by 76
“Current research demonstrates people’s willingness to offload (parts of) a task to an algorithm-based partner in case a high level of cognitive load is involved in a task (Wahn et al., 2023).”
BackgroundPlacebo or Assistant? Generative AI Between Externalization and Anthropomorphization
Skulmowski, A. (2024) · Educational Psychology Review · doi:10.1007/s10648-024-09894-x · Cited by 43
“…of studies suggest that humans prefer to share their workload with technological aids under certain circumstances (e.g., Wahn & Kingstone, 2021; Wahn et al., 2023), while there are factors determining whether an AI is found to be trust-worthy and acceptable or whether users feel uneasy because…”
Skulmowski, A. (2024) · Educational Psychology Review · doi:10.1007/s10648-024-09869-y · Cited by 110
“…focusing on touchscreen activities and gestures) and using artificial intelligence to externalize demanding aspects of complex activities (e.g., Wahn et al., 2023; for an overview of digital externalization, see Skulmowski, 2023) to leave more cognitive resources for the actual learning…”
Wu, H., Ni, H., Wenfu, L., & Wu, T. (2026) · Frontiers in Psychology · doi:10.3389/fpsyg.2025.1725393 · Cited by 59
Metadata
- Authors
- Wahn, B., Schmitz, L., Gerster, F. N., & Weiss, M.
- Affiliations
- University Medical Center Hamburg-Eppendorf (DE)
- Ruhr University Bochum (DE)
- Radboud University Nijmegen (NL)
- Corresponding author
- Wahn, Basil
- Venue
- PLOS ONE
- Publisher
- Public Library of Science (PLoS)
- Published
- 2023-05-19
- Volume
- 18
- Issue
- 5
- Pages
- e0286102
- ISSN
- 1932-6203 (ISSN-L)
- PMC
- PMC10198496
- OpenAlex
- W4377115947
- Cited by
- 49
- References
- 28
- Citations by year
- 2026: 25 · 2025: 13 · 2024: 9 · 2023: 2
- FWCI
- 14.47
- Type
- journal article
- Language
- en
- Open access
- gold
- doi.org
- Licence
- Locations
- PLoS ONE (publishedVersion, cc-by)
- PubMed Central (publishedVersion, cc-by)
- PubMed (publishedVersion)
- Radboud Repository (Radboud University) (submittedVersion)
- RePEc: Research Papers in Economics (submittedVersion)
- DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion, cc-by-sa)
- Radboud Repository (Radboud University) (submittedVersion, other-oa)
- Funders
- Ministry of Culture and Science of the State of North Rhine-Westphalia
- Topics
- Social Sciences › Human Factors and Ergonomics › Human-Automation Interaction and Safety
- Social Sciences › Psychology › Social Psychology › Social Robot Interaction and HRI
- Life Sciences › Neuroscience › Cognitive Neuroscience › Neural and Behavioral Psychology Studies
- Keywords
- algorithmic aversion, task offloading, cognitive load, multiple object tracking, human-computer collaboration, human-automation interaction, attention, trust in algorithms
- MeSH terms
- Affect, Algorithms, Cognition, Computers, Humans
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.