Pith. sign in

REVIEW 4 major objections 6 minor 57 references

Predicting Situation Awareness from Physiological Signals

T0 review · 4 major / 6 minor · reviewed 2026-08-07 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

Pith's one-line read This paper claims that multimodal physiological signals carry enough information to predict an operator's situation awareness at all three levels along a continuum, with EEG plus eye tracking capturing most of the signal.

desk verdict A substantial multimodal SA-prediction study whose headline numbers are inflated by a centered future-including target and within-participant cross-validation; the core signal likely survives, but the real-time claim needs revision. read the letter →

arxiv 2506.07930 v1 pith:5LJRHD5V submitted 2025-06-09 cs.HC

Kieran J. Smith , Tristan C. Endsley , Torin K. Clark This is my paper
classification cs.HC
keywords situationawarenessphysiologicalsignalsEEGeyetrackingfNIRSmachinelearninghuman-automationinteractionfreeze-probeassessment
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

Situation awareness has three levels—perceiving what matters, understanding what it means, and projecting what happens next—and today's best measures interrupt operators to ask. This paper tries to show that passive physiological recordings can stand in for those interruptions by predicting all three levels, and total SA, as continuous scores rather than binary high/low labels. Using data from 31 people flying a multitasking aircraft simulator, the authors report that models built from six physiological streams beat models trained on shuffled labels, with cross-validated $Q^2$ values of 0.14, 0.00, 0.26, and 0.36 for levels 1–3 and total, rising to 0.21, 0.29, 0.34, and 0.41 when per-sensor predictions are fused. EEG and eye tracking carry most of the signal, and a reduced EEG-plus-eye-tracking fusion reaches $Q^2=0.48$ for total SA. If the claim holds, adaptive systems could track operator SA continuously and adjust task demands without freezing the task.

What carries the argument

The load-bearing machinery is a supervised learning pipeline: physiological streams are cleaned and reduced to per-trial features, such as spectral EEG powers, fixation and pupil metrics, heart-rate variability, respiration measures, electrodermal responses, and fNIRS hemoglobin dynamics. A relaxed LASSO procedure—a two-stage shrinkage-and-selection step followed by ordinary least squares—selects a sparse linear model for each SA target; a sensor-fusion variant builds one model per sensor and combines their predictions weighted by in-sample fit. Predictions are evaluated by five-fold cross-validation in which every participant appears in every fold, and shuffled-label null models set the chance baseline, while ablation and single-sensor comparisons isolate each sensor's contribution.

What would settle it

Run the identical feature and model pipeline with leave-one-participant-out cross-validation and a purely retrospective SA label; if total-SA $Q^2$ falls to zero or below, the reported predictive signal is within-participant and partly future-informed, not a general real-time estimate.

Watch

Extended reading notes

Core claim

On the paper's own terms, the central discovery is that physiological signals contain usable, cross-validated information about all three levels of situation awareness, not just about overall task engagement. Direct multiple-regression models trained on trial-level features from EEG, eye tracking, fNIRS, ECG, respiration, and EDA predict perception, comprehension, projection, and total SA with $Q^2$ = 0.14, 0.00, 0.26, and 0.36, and none of 50 shuffled-label models matched them. Projection (level 3) was the most predictable and comprehension (level 2) the hardest, but a sensor-fusion approach that weights each sensor's separate predictions lifted level-2 $Q^2$ to 0.29 and total SA to 0.41. EEG and eye-tracking emerged as the most informative sensors, and a two-sensor fusion of just those streams produced total-SA $Q^2$ of 0.48, slightly above the full six-sensor fusion.

Load-bearing premise

The load-bearing premise is that the physiological patterns learned from a given participant's earlier trials will predict that participant's later SA scores and then transfer to operators the model has never met, even though the reported five-fold cross-validation never holds out an entire participant and the smoothed SA label includes future trials.

Editorial extensions

If this is right

  • Continuous, non-disruptive SA estimation becomes feasible: an adaptive system could infer rising or falling SA from physiology and adjust automation, alerts, or display content in real time.
  • A reduced sensor suite is enough in practice: EEG plus eye tracking can match or exceed the full six-sensor total-SA model, lowering the burden of deployment.
  • Level-specific adaptation is possible: because projection is the most predictable target, systems might prioritize forecasting support, while comprehension would need multi-sensor fusion.
  • The modest $Q^2$ ceiling, at most 0.48 in the authors' best model, means physiological SA estimates are probabilistic; engineering decisions will need application-specific accuracy thresholds.
  • Null-model comparison establishes that the predictive signal is real relative to chance, so future work can build on these features rather than starting from scratch.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • The paper leaves untested whether the within-participant results transfer to unseen operators; a leave-one-participant-out evaluation would settle that and is a direct next step.
  • Because the smoothed three-trial label is centered, the models were not trained on a strictly causal target; recalculating with past-only labels would test whether the reported $Q^2$ reflects real-time predictability.
  • The sensor-fusion gain for comprehension suggests that no single stream indexes understanding, and that fusing streams approximates a latent cognitive state that nonlinear models might capture better than linear regressions.
  • The relative value of EEG and eye tracking over other sensors may shift with task domain; rerunning the ablation in driving, process control, or medical monitoring would test how general the sensor-reduction guidance is.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

4 major / 6 minor

Summary. The paper presents a lab study (N=31) in which participants performed a modified MATB-II multi-tasking simulation while wearing six physiological sensor streams (EEG, fNIRS, eye tracking, ECG, EDA, respiration). The authors extract features per trial and fit relaxed-LASSO linear regression models to predict continuous, difficulty-adjusted, three-trial moving-averaged SA scores at levels 1, 2, and 3, plus total SA. They report direct-model Q2 values of 0.14, 0.00, 0.26, and 0.36 for levels 1-3 and total, and show that these exceed shuffled-label null models. Sensor-fusion and ablation analyses identify EEG and eye tracking as most useful, and a two-sensor fusion model achieves Q2 of 0.48 for total SA. The central claim is that multimodal physiological signals carry useful information for non-disruptive, continuous prediction of SA at all three levels.

Significance. If the predictive results are valid, the paper would be a useful contribution to passive situation-awareness estimation, combining a relatively rich sensor suite, all three SA levels along a continuum, and a realistic multi-tasking battery. I credit the authors for collecting a substantial dataset, comparing direct and sensor-fusion models against shuffled-label nulls, and performing ablation and sensor-weighting analyses to guide sensor selection. The descriptive feature tables and the explicit overfitting criterion are also useful. However, the validation pipeline has leaks that undermine the central claim in its current form: the three-trial moving-average target is centered and includes future SA scores, the final five-fold cross-validation places the same participants in every fold, and question-difficulty adjustment is computed on the full dataset. These issues are load-bearing for the claim that physiology predicts current or real-time SA, and they require re-analysis before the results can be accepted.

major comments (4)
  1. [Section III.B] The three-trial moving-average target is centered: the authors state that first and last trials are removed to avoid edge effects, which implies the label for trial t is the average of SA_{t-1}, SA_t, and SA_{t+1}. The physiology features are extracted from trial t only, so the target contains a future SA score relative to the predictors. A model can therefore achieve positive Q2 by exploiting correlation with future freeze-probe outcomes, not by tracking the current SA state. Because the shuffled-label null preserves this same target construction, beating the null does not isolate information about current SA. Please report results with a causal (trailing) moving average or with per-trial un-smoothed labels, or explicitly justify that the centered target is the quantity a real-time system would predict. As written, the abstract's 'real-time, non-disruptive' claim is not supported by the reported analysis.
  2. [Section III.C] The final 5-fold cross-validation is explicitly described as having each participant represented in each fold. This means training data contain other trials from the same test participant. Physiological signals contain stable individual differences, and SA scores are temporally autocorrelated, so a model can achieve positive Q2 by recognizing participants or slow trends without tracking the current SA state. The paper's internal model-selection step uses leave-one-participant-out, but the performance estimates reported in Tables 1 and 4 come from within-participant folds. Please report leave-one-participant-out (or participant-exclusive fold) performance as the primary estimate, or clearly separate within-participant prediction from between-participant generalization. The conclusion that physiology can support real-time SA prediction for operators who were not in the training set is not supported by the present cross-validation.
  3. [Section III.B] The question-difficulty adjustment is computed using the percentage of correct responses for each question across all participants, before any train/test split. This means the SA labels in the test folds depend on information from the training folds. The subsequent standardization is also described as being applied before model-building, which likely shares statistics across folds. This label-construction leakage can inflate predictive performance and is not controlled by the shuffled-label null, because the null models use the same leaked labels. Please nest the difficulty estimation and standardization inside each training fold (or show that the adjustment has negligible effect).
  4. [Abstract and Table 1] The abstract claims that the results 'demonstrate that multimodal physiological signals provide useful information in predicting all SA levels.' However, the level-2 direct model has Q2 = 0.00, which by standard definition means the model is no better than predicting the mean of the test set. While the level-2 sensor-fusion model reaches Q2 = 0.29, the direct-model result at level 2 does not support the unqualified 'all levels' claim. Please either soften the claim to acknowledge that level-2 comprehension was not directly predictable, or base the claim on the sensor-fusion results and state the level-2 direct-model result as a failure to find direct predictive signal.
minor comments (6)
  1. [Section III.C] The sentence defining the overfit criterion reads 'Models with internal Q2 metrics within 0.2 of that models’ R2 were considered not overfit' and later 'define as not being overfit.' Please clean up the grammar and clarify whether the criterion is |Q2_internal - R2| <= 0.2 or Q2_internal >= R2 - 0.2.
  2. [Section IV.A] The text says 'our model-selection method may may manage to return overfit models'; there is a duplicated 'may.'
  3. [Section V] The conclusion contains 'we found find that our models pick up on genuine signals'; remove the duplicate 'find.'
  4. [Tables 8-11] Many coefficients in the supplementary tables are reported as exactly 0.00, which likely reflects rounding. Please report coefficients with meaningful precision or add a footnote stating that values below 0.005 are displayed as 0.00.
  5. [Section III.C] The paper mentions internal leave-one-trial-out and leave-one-participant-out cross-validation for model selection, but only the final five-fold results are reported. Please report the internal metric values or state that they are available in supplementary material, so readers can see how the overfit criterion behaves.
  6. [Figures] Figures 2, 3, 5, and 6 are referenced by number, but in the manuscript text some of these are placed in the supplementary appendices. Please make the placement explicit in the captions and main text to avoid reader confusion.

Circularity Check

1 steps flagged · score 2.0 of 10

No significant circularity: cross-validated predictions against an in-paper shuffled-label null carry the central claim; only a minor self-citation defines the smoothed SA target.

  1. self citation load bearing [Section III.B (Situation Awareness), supported by reference [50] and repeated in Section V.D (Limitations).]
    "Lastly, since freeze-probe SA scores face challenges with single-trial sensitivity, we computed a moving-average of 3 SA scores [50] within each participant. To avoid edge effects, we removed the first and last trials of each participant."

    Every Q2 result in the paper is a prediction of one dependent variable: the within-participant centered moving average of three SA scores. That target is introduced with the sole citation [50], which is the authors' own manuscript under review (K. J. Smith, Torin K. Clark, Tristan C. Endsley, 'Balancing Temporal Dynamics with Measurement Noise in Real-Time Situation Awareness Prediction,' Ergonomics, under review). The target's validity is likewise supported only by the same self-citation in the Limitations section: 'The moving average scores used here ...

full rationale

The derivation chain is: per-trial physiological features -> relaxed-LASSO + OLS models -> predictions on 20% held-out folds -> Q2 compared against a 50-permutation shuffled-label null. SA labels come exclusively from freeze-probe answers (Section III.B) and never from physiological features or model outputs, so no step is self-definitional: the target is not defined in terms of the predictors, and no fitted parameter is renamed as a prediction. The 5-fold predictions are genuinely out-of-sample for trials, the null is computed in-paper ('SA scores were randomly permuted 50 different times ... models were generated in the exact same fashion'), and direct models beat all 50 shuffled baselines in Q2 and MAE (Table 3), so the central empirical claim that paired physiology outperforms dissociated physiology has independent content. The one self-reference that carries weight is citation [50] (same first and senior authors, under review), which motivates the centered three-trial moving-average target and its claim to track task performance; because every Q2 number is a prediction of this self-defined target, the evaluation benchmark is partly internal to the authors' own citation chain, but this neither constructs labels from physiology nor forces the predictive comparison. Validity risks, which I weigh but do not score as circularity because they are leakage rather than reduction-to-input, are disclosed in the manuscript's own method and limitations: the centered three-trial window includes SA from trial t+1 ('we computed a moving-average of 3 SA scores ... within each participant'), each cross-validation fold contains every participant ('Each participant was represented in each fold'), and question-difficulty statistics are computed on the full dataset before splitting, so part of the measured Q2 may reflect temporal autocorrelation, participant identity, or label-construction leakage rather than current-SA-specific signal. Section V.D likewise concedes the granularity of the six-question-per-level measure and the reliance on the moving average. These are correctness concerns, not circular steps, and they do not make the prediction equivalent to its input. Verdict: score 2, one minor self-citation that is bounded and leaves the predictive result with independent empirical content.

Assumptions & free parameters 5 free parameters · 4 assumptions · 0 invented entities

The study introduces no new theoretical entities. The free parameters are modeling choices and data-adaptive adjustments. The key axioms are domain assumptions about SA measurement and generalization that are not independently validated.

free parameters (5)
  • Question difficulty adjustment percentages = per-question p(correct) computed across all participants
    Used to adjust SA scores; computed on the full dataset, which may introduce mild leakage in cross-validation.
  • Moving-average window length = 3 trials (centered)
    Chosen by authors to reduce single-trial noise; centered window uses future trials in the target.
  • Relaxed LASSO tuning parameters (lambda, feature count cap) = lambda at 1 standard error or minimum; predictor cap of 40
    Selection via internal cross-validation; standard but involves researcher choices.
  • Overfit criterion (Q2-R2 tolerance) = 0.2
    Threshold for considering a model non-overfit; arbitrary but stated.
  • Sensor fusion weights = based on training R2 values
    Weights are fit to training data within each fold.
assumptions (4)
  • domain assumption Freeze-probe SA scores measure true situation awareness
    The entire prediction target rests on this; no independent construct validation is provided.
  • domain assumption Physiological features carry SA-specific information beyond task load and arousal
    Task load was manipulated across trials; models may be predicting load rather than SA.
  • ad hoc to paper Centered three-trial moving average is an appropriate target for real-time prediction
    Not justified for real-time use; uses future observations.
  • ad hoc to paper Within-participant cross-validation performance approximates between-participant generalization
    Folds include the same participants in training and test; no participant-independent final evaluation is reported.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Predicting Situation Awareness from Physiological Signals." pith.science (2026). https://pith.science/paper/5LJRHD5V

@misc{pith2026250607930,
  author       = {Pith},
  title        = {Pith review of: Predicting Situation Awareness from Physiological Signals},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/5LJRHD5V}},
  note         = {Machine review of arXiv:2506.07930}
}
abstract

Situation awareness (SA)--comprising the ability to 1) perceive critical elements in the environment, 2) comprehend their meanings, and 3) project their future states--is critical for human operator performance. Due to the disruptive nature of gold-standard SA measures, researchers have sought physiological indicators to provide real-time information about SA. We extend prior work by using a multimodal suite of neurophysiological, psychophysiological, and behavioral signals, predicting all three levels of SA along a continuum, and predicting a comprehensive measure of SA in a complex multi-tasking simulation. We present a lab study in which 31 participants controlled an aircraft simulator task battery while wearing physiological sensors and responding to SA 'freeze-probe' assessments. We demonstrate the validity of task and assessment for measuring SA. Multimodal physiological models predict SA with greater predictive performance ($Q^2$ for levels 1-3 and total, respectively: 0.14, 0.00, 0.26, and 0.36) than models built with shuffled labels, demonstrating that multimodal physiological signals provide useful information in predicting all SA levels. Level 3 SA (projection) was best predicted, and level 2 SA comprehension) was the most challenging to predict. Ablation analysis and single sensor models found EEG and eye-tracking signals to be particularly useful to predictions of level 3 and total SA. A reduced sensor fusion model showed that predictive performance can be maintained with a subset of sensors. This first rigorous cross-validation assessment of predictive performance demonstrates the utility of multimodal physiological signals for inferring complex, holistic, objective measures of SA at all levels, non-disruptively, and along a continuum.

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

57 extracted references · 41 canonical work pages

  1. [1]

    Department of the Air Force Role in Joint All-Domain Operations,

    W. G. Holt and S. N. Bratton, “Department of the Air Force Role in Joint All-Domain Operations,” Maxwell AFB, AL, ADFP 3-99, Nov. 2021

    work page 2021

  2. [2]

    Human–Autonomy Teaming: Definitions, Debates, and Directions,

    J. B. Lyons, K. Sycara, M. Lewis, and A. Capiola, “Human–Autonomy Teaming: Definitions, Debates, and Directions,” Front. Psychol., vol. 12, May 2021, doi: 10.3389/fpsyg.2021.589585

    work page arXiv 2021

  3. [3]

    State of California Vehicle Code Section 38750. 2025. Accessed: Jan. 12, 2025. [Online]. Available: https://leginfo.legislature.ca.gov/faces/codes_displaySec tion.xhtml?sectionNum=38750&lawCode=VEH

    work page 2025

  4. [4]

    Effects of adaptive cruise control and highly automated driving on workload and situation awareness: A review of the empirical evidence,

    J. C. F. de Winter, R. Happee, M. H. Martens, and N. A. Stanton, “Effects of adaptive cruise control and highly automated driving on workload and situation awareness: A review of the empirical evidence,” Transportation Research Part F: Traffic Psychology and Behaviour, vol. 27, pp. 196–217, Nov. 2014, doi: 10.1016/j.trf.2014.06.016

    doi:10.1016/j.trf.2014.06.016 2014

  5. [5]

    M. R. Endsley, Designing for Situation Awareness: An Approach to User-Centered Design, Second Edition, 2nd ed. Boca Raton: CRC Press, 2012. doi: 10.1201/b11371

    doi:10.1201/b11371 2012

  6. [6]

    The Out-of-the-Loop Performance Problem and Level of Control in Automation,

    M. R. Endsley and E. O. Kiris, “The Out-of-the-Loop Performance Problem and Level of Control in Automation,” Hum Factors, vol. 37, no. 2, pp. 381–394, Jun. 1995, doi: 10.1518/001872095779064555

    doi:10.1518/001872095779064555 1995

  7. [7]

    Task specific effects of automation and automation failure on performance, workload and situational awareness,

    M. A. Carmody and J. P. Gluckman, “Task specific effects of automation and automation failure on performance, workload and situational awareness,” presented at the Adaptive Function Allocation for Intelligent Cockpits (AFAIC), 1993, pp. 13–18

    work page 1993

  8. [8]

    Washington, D.C.: National Academies Press, 2021, p

    Committee on Human-System Integration Research Topics for the 711th Human Performance Wing of the Air Force Research Laboratory, Board on Human- Systems Integration, Division of Behavioral and Social Sciences and Education, and National Academies of Sciences, Engineering, and Medicine, Human-AI Teaming: State of the Art and Research Needs. Washington, D.C...

    doi:10.17226/26355 2021

Show all 57 references
  1. [9]

    Zacharias, Autonomous horizons: the way forward

    G. Zacharias, Autonomous horizons: the way forward. Maxwell Air Force Base, Alabama: Air University Press; Curtis E. LeMay Center for Doctrine Development and Education, 2019

    2019

  2. [10]

    Situation awareness global assessment technique (SAGAT),

    M. R. Endsley, “Situation awareness global assessment technique (SAGAT),” in Proceedings of the IEEE 1988 National Aerospace and Electronics Conference, May 1988, pp. 789–795 vol.3. doi: 10.1109/NAECON.1988.195097

    1988

  3. [11]

    Quantitative Analysis of Situational Awareness (QUASA): Applying Signal Detection Theory to True/False Probes and Self-Ratings,

    B. McGuinness, “Quantitative Analysis of Situational Awareness (QUASA): Applying Signal Detection Theory to True/False Probes and Self-Ratings,” Jun

  4. [12]

    J. A. Stern, L. Wang, and D. Schroeder, “Physiological Measurement Techniques: What the Heart and Eye Can 10 > REPLACE THIS LINE WITH YOUR MANUSCRIPT ID NUMBER (DOUBLE -CLICK HERE TO EDIT) < Tell Us About Aspects of Situational Awareness,” in Proceedings of the International C...

    1995

  5. [13]

    Role of Volitional Effort in the Application of Psychophysiological Measures to Situation Awareness,

    E. A. Byrne, “Role of Volitional Effort in the Application of Psychophysiological Measures to Situation Awareness,” in Proceedings of the International Conference on Experimental Analysis and Measurement of Situation Awareness, Daytona Beach, FL: Embry-Riddle Aeronautical Univ...

    1995

  6. [14]

    Investigating the relationship between eye movements and situation awareness in weather forecasting,

    E. M. Argyle, J. J. Gourley, Z. Kang, and R. L. Shehab, “Investigating the relationship between eye movements and situation awareness in weather forecasting,” Applied Ergonomics, vol. 85, p. 103071, May 2020, doi: 10.1016/j.apergo.2020.103071

    2020

  7. [15]

    Correlation between Situational Awareness and EEG signals,

    J. L. Kästle, B. Anvari, J. Krol, and H. A. Wurdemann, “Correlation between Situational Awareness and EEG signals,” Neurocomputing, vol. 432, pp. 70–79, Apr. 2021, doi: 10.1016/j.neucom.2020.12.026

    2021 doi

  8. [16]

    Using Eye-Tracking Data to Predict Situation Awareness in Real Time During Takeover Transitions in Conditionally Automated Driving,

    F. Zhou, X. J. Yang, and J. C. F. de Winter, “Using Eye-Tracking Data to Predict Situation Awareness in Real Time During Takeover Transitions in Conditionally Automated Driving,” IEEE Transactions on Intelligent Transportation Systems, vol. 23, no. 3, pp. 2284–2295, Mar. 2022,...

    2022

  9. [17]

    Assessing Situation Awareness (SA) Using Single- or Dual-Location functional Near Infrared Spectroscopy (fNIRS):,

    B. Bracken, A. Winder, B. Hager, M. Endsley, and E. Festa, “Assessing Situation Awareness (SA) Using Single- or Dual-Location functional Near Infrared Spectroscopy (fNIRS):,” in Proceedings of the 16th International Joint Conference on Biomedical Engineering Systems and Techno...

    2023 doi

  10. [18]

    Multimodal Sensing and Computational Intelligence for Situation Awareness Classification in Autonomous Driving,

    J. Yang et al., “Multimodal Sensing and Computational Intelligence for Situation Awareness Classification in Autonomous Driving,” IEEE Transactions on Human- Machine Systems, pp. 1–12, 2023, doi: 10.1109/THMS.2023.3234429

    2023

  11. [20]

    Examining the Relationship between Construction Workers’ Visual Attention and Situation Awareness under Fall and Tripping Hazard Conditions: Using Mobile Eye Tracking,

    S. Hasanzadeh, B. Esmaeili, and M. D. Dodd, “Examining the Relationship between Construction Workers’ Visual Attention and Situation Awareness under Fall and Tripping Hazard Conditions: Using Mobile Eye Tracking,” J. Constr. Eng. Manage., vol. 144, no. 7, p. 04018060, Jul. 201...

    2018

  12. [21]

    The Influence of HUD Information Visual Coding on Pilot’s Situational Awareness,

    G. Sun, X. Wanyan, X. Wu, and D. Zhuang, “The Influence of HUD Information Visual Coding on Pilot’s Situational Awareness,” in 2017 9th International Conference on Intelligent Human-Machine Systems and Cybernetics (IHMSC), Hangzhou: IEEE, Aug. 2017, pp. 139–143. doi: 10.1109/I...

    2017 doi

  13. [22]

    Operator situation awareness and physiological states during offshore well control scenarios,

    R. K. Mehta et al., “Operator situation awareness and physiological states during offshore well control scenarios,” Journal of Loss Prevention in the Process Industries, vol. 55, pp. 332–337, Sep. 2018, doi: 10.1016/j.jlp.2018.07.010

    2018 doi

  14. [23]

    Situation awareness based on eye movements in relation to the task environment,

    J. C. F. de Winter, Y. B. Eisma, C. D. D. Cabrall, P. A. Hancock, and N. A. Stanton, “Situation awareness based on eye movements in relation to the task environment,” Cogn Tech Work, vol. 21, no. 1, pp. 99– 111, Feb. 2019, doi: 10.1007/s10111-018-0527-6

    2019 doi

  15. [24]

    Mapping Brain Activity During Loss of Situation Awareness: An EEG Investigation of a Basis for Top-Down Influence on Perception,

    D. Catherwood et al., “Mapping Brain Activity During Loss of Situation Awareness: An EEG Investigation of a Basis for Top-Down Influence on Perception,” Hum Factors, vol. 56, no. 8, pp. 1428–1452, Dec. 2014, doi: 10.1177/0018720814537070

    2014 doi

  16. [25]

    Effects of Neural Mechanisms of Pretask Resting EEG Alpha Information on Situational Awareness: A Functional Connectivity Approach,

    A. Kaur, R. Chaujar, and V. Chinnadurai, “Effects of Neural Mechanisms of Pretask Resting EEG Alpha Information on Situational Awareness: A Functional Connectivity Approach,” Hum Factors, vol. 62, no. 7, pp. 1150–1170, Nov. 2020, doi: 10.1177/0018720819869129

    2020 doi

  17. [26]

    Recognising situation awareness associated with different workloads using EEG and eye-tracking features in air traffic control tasks,

    Q. Li, K. K. H. Ng, S. C. M. Yu, C. Y. Yiu, and M. Lyu, “Recognising situation awareness associated with different workloads using EEG and eye-tracking features in air traffic control tasks,” Knowledge-Based Systems, vol. 260, p. 110179, Jan. 2023, doi: 10.1016/j.knosys.2022.110179

    2023

  18. [27]

    EEG-based neural activity for decoding situation awareness at different levels,

    Q. Wu, N. Chen, and P.-L. P. Rau, “EEG-based neural activity for decoding situation awareness at different levels,” Ergonomics, vol. 0, no. 0, pp. 1–13, doi: 10.1080/00140139.2024.2449570

    2024

  19. [28]

    Physiological Measurements of Situation Awareness: A Systematic Review,

    T. Zhang et al., “Physiological Measurements of Situation Awareness: A Systematic Review,” Hum Factors, vol. 65, no. 5, pp. 737–758, Aug. 2023, doi: 10.1177/0018720820969071

    2023 doi

  20. [29]

    Toward a Characterization of Adaptive Systems: A Framework for Researchers and System Designers,

    K. M. Feigh, M. C. Dorneich, and C. C. Hayes, “Toward a Characterization of Adaptive Systems: A Framework for Researchers and System Designers,” Hum Factors, vol. 54, no. 6, pp. 1008–1024, Dec. 2012, doi: 10.1177/0018720812443983

    2012 doi

  21. [30]

    Situation Awareness, Mental Workload, and Trust in Automation: Viable, Empirically Supported Cognitive Engineering Constructs,

    R. Parasuraman, T. B. Sheridan, and C. D. Wickens, “Situation Awareness, Mental Workload, and Trust in Automation: Viable, Empirically Supported Cognitive Engineering Constructs,” Journal of Cognitive Engineering and Decision Making, vol. 2, no. 2, pp. 140–160, Jun. 2008, doi:...

    2008 doi

  22. [31]

    Performance consequences of automation-induced ‘complacency.,’

    R. Parasuraman, R. Molloy, and I. L. Singh, “Performance consequences of automation-induced ‘complacency.,’” The International Journal of Aviation Psychology, vol. 3, no. 1, pp. 1–23, 1993, doi: 10.1207/s15327108ijap0301_1

    1993 doi

  23. [32]

    Complacency and Bias in Human Use of Automation: An Attentional 11 > REPLACE THIS LINE WITH YOUR MANUSCRIPT ID NUMBER (DOUBLE -CLICK HERE TO EDIT) < Integration,

    R. Parasuraman and D. H. Manzey, “Complacency and Bias in Human Use of Automation: An Attentional 11 > REPLACE THIS LINE WITH YOUR MANUSCRIPT ID NUMBER (DOUBLE -CLICK HERE TO EDIT) < Integration,” Hum Factors, vol. 52, no. 3, pp. 381–410, Jun. 2010, doi: 10.1177/0018720810376055

    2010 doi

  24. [33]

    Boucsein, Electrodermal Activity

    W. Boucsein, Electrodermal Activity. Boston, MA: Springer US, 2012. doi: 10.1007/978-1-4614-1126-0

    2012 doi

  25. [34]

    Comparison of Electrodermal Activity from Multiple Body Locations Based on Standard EDA Indices’ Quality and Robustness against Motion Artifact,

    M.-B. Hossain, Y. Kong, H. F. Posada-Quintero, and K. H. Chon, “Comparison of Electrodermal Activity from Multiple Body Locations Based on Standard EDA Indices’ Quality and Robustness against Motion Artifact,” Sensors (Basel), vol. 22, no. 9, p. 3177, Apr. 2022, doi: 10.3390/s22093177

    2022 doi

  26. [35]

    EDA Guide,

    BIOPAC Systems, Inc., “EDA Guide,” BIOPAC Systems, Inc

  27. [36]

    LabStreamingLayer submodules · GitHub

    “LabStreamingLayer submodules · GitHub.” Accessed: Feb. 21, 2023. https://github.com/labstreaminglayer

    2023

  28. [37]

    The Multi-Attribute Task Battery II (MATB-II) Software for Human Performance and Workload Research: A User’s Guide,

    Y. Santiago-Espada, R. R. Myer, K. A. Latorella, and J. R. Comstock, “The Multi-Attribute Task Battery II (MATB-II) Software for Human Performance and Workload Research: A User’s Guide,” p. 269, Jul. 2011

    2011

  29. [38]

    Measuring Situational Awareness Aptitude Using Functional Near-Infrared Spectroscopy,

    L. Hirshfield, M. Costa, D. Bandara, and S. Bratt, “Measuring Situational Awareness Aptitude Using Functional Near-Infrared Spectroscopy,” in Foundations of Augmented Cognition, D. D. Schmorrow and C. M. Fidopiastis, Eds., in Lecture Notes in Computer Science. Cham: Springer I...

    2015 doi

  30. [39]

    Situation awareness for socio technical systems: a simulation gaming study in intermodal transport operations,

    S. Kurapati, “Situation awareness for socio technical systems: a simulation gaming study in intermodal transport operations,” TU Delft, Delft, Netherlands, 2017

    2017

  31. [40]

    Study on Situation Awareness under Different Mental Workloads,

    C. Feng, X. Wanyan, S. Liu, H. Chen, and D. Zhuang, “Study on Situation Awareness under Different Mental Workloads,” JNWPU, vol. 38, no. 3, Art. no. 3, Jun. 2020, doi: 10.1051/jnwpu/20203830610

    2020

  32. [41]

    Eye-Tracking and Situation Awareness: Predicting The Malfunctions in Dynamic Control Tasks with a Mathematical Model for Different Age Groups and Interactive Devices,

    C. Shi, “Eye-Tracking and Situation Awareness: Predicting The Malfunctions in Dynamic Control Tasks with a Mathematical Model for Different Age Groups and Interactive Devices,” Ph.D., The Pennsylvania State University, United States -- Pennsylvania, 2020

    2020

  33. [42]

    Situation Awareness Oriented Design: From User’s Cognitive Requirements to Creating Effective Supporting Technologies,

    M. R. Endsley, C. A. Bolstad, D. G. Jones, and J. M. Riley, “Situation Awareness Oriented Design: From User’s Cognitive Requirements to Creating Effective Supporting Technologies,” Proceedings of the Human Factors and Ergonomics Society Annual Meeting, vol. 47, no. 3, pp. 268–...

    2003 doi

  34. [43]

    The NIRS Brain AnalyzIR Toolbox,

    H. Santosa, X. Zhai, F. Fishburn, and T. Huppert, “The NIRS Brain AnalyzIR Toolbox,” Algorithms, vol. 11, no. 5, Art. no. 5, May 2018, doi: 10.3390/a11050073

    2018 doi

  35. [44]

    EEGLAB: an open source toolbox for analysis of single-trial EEG dynamics including independent component analysis,

    A. Delorme and S. Makeig, “EEGLAB: an open source toolbox for analysis of single-trial EEG dynamics including independent component analysis,” Journal of Neuroscience Methods, vol. 134, no. 1, pp. 9–21, Mar. 2004, doi: 10.1016/j.jneumeth.2003.10.009

    2004 doi

  36. [45]

    Use of EEG Workload Indices for Diagnostic Monitoring of Vigilance Decrement,

    A. T. Kamzanova, A. M. Kustubayeva, and G. Matthews, “Use of EEG Workload Indices for Diagnostic Monitoring of Vigilance Decrement,” Hum Factors, vol. 56, no. 6, pp. 1136–1149, Sep. 2014, doi: 10.1177/0018720814526617

    2014 doi

  37. [46]

    Recurrence quantification analysis of eye movements,

    N. C. Anderson, W. F. Bischof, K. E. W. Laidlaw, E. F. Risko, and A. Kingstone, “Recurrence quantification analysis of eye movements,” Behav Res, vol. 45, no. 3, pp. 842–856, Sep. 2013, doi: 10.3758/s13428-012- 0299-5

    2013 doi

  38. [47]

    R-DECO: an open-source Matlab based graphical user interface for the detection and correction of R-peaks,

    J. Moeyersons, M. Amoni, S. V. Huffel, R. Willems, and C. Varon, “R-DECO: an open-source Matlab based graphical user interface for the detection and correction of R-peaks,” PeerJ Comput. Sci., vol. 5, p. e226, Oct. 2019, doi: 10.7717/peerj-cs.226

    2019 doi

  39. [48]

    A continuous measure of phasic electrodermal activity,

    M. Benedek and C. Kaernbach, “A continuous measure of phasic electrodermal activity,” Journal of Neuroscience Methods, vol. 190, no. 1, pp. 80–91, Jun. 2010, doi: 10.1016/j.jneumeth.2010.04.028

    2010 doi

  40. [49]

    Decomposition of skin conductance data by means of nonnegative deconvolution,

    M. Benedek and C. Kaernbach, “Decomposition of skin conductance data by means of nonnegative deconvolution,” Psychophysiology, vol. 47, no. 4, pp. 647–658, 2010, doi: 10.1111/j.1469- 8986.2009.00972.x

    2010

  41. [50]

    Balancing Temporal Dynamics with Measurement Noise in Real-Time Situation Awareness Prediction,

    K. J. Smith, Clark, Torin K., and Endsley, Tristan C., “Balancing Temporal Dynamics with Measurement Noise in Real-Time Situation Awareness Prediction,” Ergonomics, under review

  42. [51]

    Assessing Physiological Signal Utility and Sensor Burden in Estimating Trust, Situation Awareness, and Mental Workload,

    S. L. Buchner, J. R. Kintz, J. Y. Zhang, N. T. Banerjee, T. K. Clark, and A. P. Hayman, “Assessing Physiological Signal Utility and Sensor Burden in Estimating Trust, Situation Awareness, and Mental Workload,” Journal of Cognitive Engineering and Decision Making, p. 1555343424...

    2025 doi

  43. [52]

    Relaxed Lasso,

    N. Meinshausen, “Relaxed Lasso,” Computational Statistics & Data Analysis, vol. 52, no. 1, pp. 374–393, Sep. 2007, doi: 10.1016/j.csda.2006.12.019

    2007 doi

  44. [53]

    University of Colorado Boulder Research Computing. (2023). University of Colorado Boulder. https://doi.org/10.25811/k3w6-pk81

    2023 doi

  45. [54]

    Combining Electroencephalograph and Functional Near Infrared Spectroscopy to Explore Users’ Mental Workload,

    L. M. Hirshfield et al., “Combining Electroencephalograph and Functional Near Infrared Spectroscopy to Explore Users’ Mental Workload,” in Foundations of Augmented Cognition. Neuroergonomics and Operational Neuroscience, D. D. Schmorrow, I. V. Estabrooke, and M. Grootjen, Eds....

    2009 doi

  46. [55]

    M. R. Endsley, Situation awareness measurement: How to measure situation awareness in individuals and teams. 2021

    2021

  47. [56]

    Situation Awareness as a Predictor of Performance for En Route Air Traffic Controllers,

    F. T. Durso, C. A. Hackworth, T. R. Truitt, J. Crutchfield, D. Nikolic, and C. A. Manning, “Situation Awareness as a Predictor of Performance for En Route Air Traffic Controllers,” Air Traffic Control Quarterly, vol. 6, no. 1, pp. 1–20, 1998, doi: 10.2514/atcq.6.1.1

    1998 doi

  48. [57]

    A Systematic Review and Meta- Analysis of Direct Objective Measures of Situation Awareness: A Comparison of SAGAT and SPAM,

    M. R. Endsley, “A Systematic Review and Meta- Analysis of Direct Objective Measures of Situation Awareness: A Comparison of SAGAT and SPAM,” Hum Factors, vol. 63, no. 1, pp. 124–150, Feb. 2021, doi: 10.1177/0018720819875376. 12 > REPLACE THIS LINE WITH YOUR MANUSCRIPT ID NUMBE...

    2021 doi

  49. [2004]

    16, 2023

    Accessed: Nov. 16, 2023. [Online]. Available: https://apps.dtic.mil/sti/citations/ADA465817

    2023

Pith tools

Reviewed August 7, 2026 · model on record in the stance chip above.