Human Factors


  • MHF1 - How much worth is a 10 EUR eye tracker?


    Master Thesis (25 EC)

    YET (is Your Eyetracker) is a project to develop a cheap, yet functional, eye tracking device for teaching and student-driven research. The result of the project is YET0, a device that can be built for 10 EUR. First results show that YET0 can be used for certain research purposes. However, a systematic evaluation of the eye trackers capabilities has not yet been undertaken.

    In your project you will create a test suite for eye tracking and use it to compare YET0 against one or more commercially available eye tracking systems. Based on your results you make recommendations on what purposes YET0 is adequate for and make specific recommendations on how to improve performance of YET0.

    Basic skills in Python and R are prerequisites for this assignment.

  • MHF2 - Investigating Experience Related Performance Differences using a Driving Training Paradigm


    Master Thesis (25 EC)

    Driving safety is affected by drivers’ judgment of dangerous situations and sensitivity. Training leads to more automated decisions and increases driving performance which is referred as the “quantity training”. This project will investigate the relationship between quantity training – the more one drives the better they would become at driving – and quality training (Nyberg, Gregersen, and Wiklund, 2007) by examining the effect of experience on task performances in driving. Both long term and short-term experience effects will be investigated by comparing non-/driver groups or between-session differences of the non-driver group respectively.

    32 participants will take place in the study where half of the participants hold a driving license and the other half do not. First, both group of subjects will be trained to operate the driving simulator for tasks such as speeding, keeping safe distance and turn. Then, they will complete more complex scenarios including curve driving, overtaking on a motorway, and negotiating a roundabout and an intersection.

    The independent variables will consist of driving experience (with or without driving licence) and session number (for no-licence group) while the dependent variables will consist of frequency of violation of rules (priority failure, failure to obey to traffic sign or signal, speeding), time to completion, task based performance metrics (distance with other vehicles, speed adaptation, lane change, overtaking and turns).

    Personal performance thresholds will be determined using either psychometric curve fitting or regression. Statistical comparison of performance metrics between groups will reveal the effect of long term experience while comparison between sessions for no-license group will reveal short term training effects.


    Nyberg, A., Gregersen, N. P., & Wiklund, M. (2007). Practicing in relation to the outcome of the driving test. Accident Analysis & Prevention, 39(1), 159-168.\

  • MHF3 - Cognitive Workload and Eye Movement Based Evaluation of Novice and Experienced Drivers


    Master Thesis (25 EC)

    Experienced drivers grow attentional habits through training such as becoming more insensitive to attentional resources and demands. On the other hand, failing to show these attentional habits might be a possible cause of not attending to potential dangers for novice drivers (Johnson, 2013).

    Eye movements and pupil reaction are important indicators of visual information processing and can reveal attentional state of drivers. Previous studies showed that experienced drivers increase their visual scanning when complexity increased on roadways. However, no other eye and gaze behavior differences were found (Robins & Chapma, 2019). Most of the studies did not take cognitive workload into account while comparing eye behavior. In this study, cognitive workload determined by task complexity, task order (e.g. temporal adjacency of complex tasks) and subjective rating scales will be considered a main factor while examining gaze behavior and pupilometry outcomes of novice and experienced drivers. The threefold relationship between eye behaviour, driving experience and cognitive workload will be studied.

    32 participants will take place in the study where the first group holds a driving license and the second group do not. Training phase will take place in basic road structures without distractions. A subjective rating scale (see e.g., Verwey, W. B., & Veltman, H. A., 1996) will be used.

    Fixation duration, fixation number and pupil size change will be modelled using cognitive workload level and will be compared between two groups. The results will inform about the co-dependence of the cognitive workload and eye-movement strategies as well as pupil size change, if any. Also, physiological underpinnings of long-term training will be revealed as a results comparing eye movement behaviour between two groups.


     Johnson, Addie. (2013) Procedural memory and skill acquisition.

    Robbins, C., & Chapman, P. (2019). How does drivers’ visual search change as a function of experience? A systematic review and meta-analysis. Accident Analysis & Prevention132, 105266.

    Verwey, W. B., & Veltman, H. A. (1996) Detecting short periods of elevated workload: A comparison of nine workload assessment techniques. Journal of Experimental Psychology-Applied, 2(3), 270-285

  • MHF4 - Sensory Cueing to Enhance Performance during a Sustained Attention Task


    Master Thesis (25 EC)

    Sustained attention is a type of selective attention which allows us to ignore irrelevant events around us, while focusing on a specific event. This paradigm helps us stay concentrated on a specific event or a task for a period of time. Previous studies have looked into the effect of background sound on sorting (Burleson et al., 1989) and sustained attention tasks (Kiss and Linnell, 2021) while results indicated that background sound can alter task performance positively depending on the type of the task.

    When semantically congruent auditory cues are introduced in tasks that involve visual spatial perception, attention increased (Ho & Spence, 2005; Spence & Soto-Faraco, 2020; but also see Ahveningen, 2019). If sensory cueing can improve performance during high cognitive load periods, training algorithms can benefit from this phenomenon such as systems for operating complex devices or machinery.

    This thesis study will investigate if auditory cues can increase performance in a sustained attention (SA) task that require high attentional load. For the SA task, participants will be presented with different letters in a temporal order, interleaved with fixation marks. As an example, they will press the ‘green button’ only when they see the letter X that follows the letter A and they will press the ‘red button’ for other combinations (see: Pulopulos, Allaert, Vanderhasselt, Sanchez-Lopez, De Witte, Baeken and De Raedt, 2020).

     The auditory cues will be presented when participant is more likely to make an error. This will correspond between 50-75% personal discrimination threshold which will be determined during pre-testing sessions. The performance threshold will be calculated by using psychometric fitting over varying stimulus presentation durations. During the SA task sessions, participants will either hear auditory cues in possible high-error moments as described above, or in random time points (sham session). In total 30 participants will be tested. A between session comparison of reaction times and performance will reveal any possible enhancing or distracting effects of auditory cuing. Future implementations of these findings can advance high level cognitive training in more realistic tasks.


    -          Ho, C., & Spence, C. (2005). Assessing the Effectiveness of Various Auditory Cues in Capturing a Driver's Visual Attention. Journal of Experimental Psychology: Applied, 11(3), 157–174.

    -          Spence, C., & Soto-Faraco, S. (2020). Crossmodal Attention Applied: Lessons for Driving (Elements in Perception). Cambridge: Cambridge University Press. doi:10.1017/9781108919951

    -          Ahveninen, Jyrki, et al. "Peripheral visual localization is degraded by globally incongruent auditory-spatial attention cues." Experimental brain research 237.9 (2019): 2137-2143.

    -          Pulopulos, M. M., Schmausser, M., De Smet, S., Vanderhasselt, M. A., Baliyan, S., Venero, C., ... & De Raedt, R. (2020). The effect of HF-rTMS over the left DLPFC on stress regulation as measured by cortisol and heart rate variability. Hormones and Behavior, 124, 104803.




    Conversational agents, such as chatbots and voice interfaces, can be used for multiple purposes e.g., support customer experience with services etc. These new tools are growing and more and more integrated into systems such as websites, social networks, cars. Smart and AI-based conversational agents are shaping the future of human-computer interaction however little is known about how to assess people reaction and satisfaction after the use of these systems.


    Advance previous work done on a new scale to assess satisfaction with chatbots. Your experimental work will focus on the evaluation of conversational agents to further streamline the reliability and validity of the scale.

    Your work will consist of testing with a remote usability test different chatbots with a set of tools, including the new scale to perform a confirmatory factorial analysis. You should be aware of statistical methods regarding factorial analysis and be able to use R. The target is to involve at least 100 participants working (potentially) in a team.

    Key references

    •             Coperich, K., Cudney, E., & Nembhard, H. Continuous Improvement Study of Chatbot Technologies using a Human Factors Methodology.

    •             Duijst, D. (2017). Can we Improve the User Experience of Chatbots with Personalisation? MSc Information Studie, Amsterdam. 

    •             Følstad, A., & Brandtzæg, P. B. (2017). Chatbots and the new world of HCI. interactions, 24(4), 38-42.

    •             Hill, J., Ford, W. R., & Farreras, I. G. (2015). Real conversations with artificial intelligence: A comparison between human–human online conversations and human–chatbot conversations. Computers in Human Behavior, 49, 245-250.




    Internship (10EC) + Thesis (25 EC)

    Preliminary setting up and piloting of an experiment for workload, awareness and engagement assessment of Level 2 and Level 3 autonomous driving.

    Internship (mandatory for the project)

    -          Internship at the BMS LAB (VR Section) with remote collaboration with APPLUS IDIADA (


    This project is part of an ongoing collaboration between APPLUS IDIADA and the UT group Human factors – CODE (Cogntion, Data & Education) regarding how to monitor human behaviour during driving with Level 2 (L2, Partial automation) or Level 3 (L3, Conditional Automation) autonomous cars. Currently, L3 is available (to a certain extent) at the BMS LAB, while Level 2 is yet to be defined.     


    Three main goals:

    i)                    Support the development of Autonomous driving simulation (L2 and L3) under the supervision of the BMS LAB staff and in coordination with APPLUS IDIADA.

    ii)                   in collaboration with APPLUS IDIADA and by a systematic literature review (using the PRISMA framework) you will identify scales to assess driver workload and awareness.

    iii)                 Set up and perform a pilot experiment in the simulator to compare different self-reported measurements of workload and awareness at least with people “driving” in L3 setting.   


    -          You are interested in Virtual Reality and Simulators and human cognition

    -          You have some knowledge about EEG and its relation to cognitive processes (EEG will be part of the ARM course in 1B)

    -          (Non-mandatory) You have previous experience in systematic literature review processes

    -          (Non-mandatory) you have some knowledge about programming in Python or Unity

    Key scales and methods in the literature for a starting point

    ·         Workload: The Integrated Workload Scale (IWS), NASA Task Load Index (NASATLX), Rating Scale Mental Effort (RSME)

    ·         Awareness: Situation Awareness Rating Technique (SART), situation present assessment method (SPAM)

    ·         Engagement (Fatigue): Karolinska sleepiness scale (KSS)

    Selection process

    For this project, you will be asked to send a brief letter of interest in which you present yourself and explain how your expertise fits the profile, and your motivation to apply for this project. Based on the letter we will invite you for a brief interview.

    Send your letter by email to

    Deadline 1 October

  • MHF7 - Research on items of the new Dutch Hazard Prediction Test for prospective drivers


    Internship (10EC) + Master Thesis (25EC)

    •             Setting up cognitive lab studies into accessibility, usability and construct coverage of new developed test item versions for the Hazard prediction Test, that will be mandatory for Dutch category B driver’s license candidates in close collaboration with CBR.

    •             Development and testing of item difficulty and item discrimination models for the first collection of newly developed HP items, using pilot test data.


    Dr. Erik Roelofs (UT, CODE)

    Dr. Simone Borsci (UT, CODE)

    Dr. Daniel Heicoop (CBR)

    Internship (mandatory for the project)

    Internship at CBR, the Dutch Institute for driving exams, Rijswijk, in close cooperation with University of Twente (CODE), and RoyalHaskonigDHV Amersfoort. Regular physical presence at CBR office is desirable and to be determined by mutual agreement.

    Internship allowance: CBR offers an internship allowance. The amount will be determined by mutual agreement.


    The CBR, responsible for driving exams in the Netherlands, is currently developing a test for hazard prediction. In the near future, candidates for the B driving license will have to take this test as part of the total examination program for the B driving license. In the spring of 2023 a pilot and try-out is to be carried out to enable further item and test development.


    Three main goals:

    ·         Determine conditions of task design and delivery that create optimal usability of the test-user interaction, including E.g. length of route, length and connectedness of clips, optimal number of (pictorial) response.

    ·         Determine the necessary features of item scenarios in order to create construct relevant task load and minimize task loads that may cause usability limitations.

    ·         Develop a set of task features for a Hazard Prediction item model, that enable variation in task complexity and warrants systematic and equivalent test versions.


    ·                     You are interested in test design, human cognition, traffic psychology, psychometrics either via a master Educational Science and technology or a master Human Factors & Engineering Psychology.

    ·                     You have some knowledge and experience in the use of eye trackers in order to monitor task test-takers’ task solution processes.

    ·                     (Desirable) You have some knowledge and experience in the use of R or SPSS in order to carry out item calibrations and item difficulty/discrimination prediction models.

    ·                     (Desirable) Possession of driver’s license B.

    ·                     (Non-mandatory) You have previous experience in systematic literature review processes.

    Selection process

    For this project, you will be asked to send a brief letter of interest in which you present yourself and explain how your expertise fits the profile, and your motivation to apply for this project. Based on the letter we will invite you for a brief interview.

    Send your letter by email to
    Deadline 1 October 2022

    Relevant literature to start with:

    Crundall, D., 2016. Hazard prediction discriminates between novice and experienced drivers. Accident Analysis & Prevention. 86, 47–58.

    Crundall, D., van Loon, E., Baguley, T., Kroll, V., 2021. A novel driving assessment combining hazard perception, hazard prediction and theory questions. Accident Analysis & Prevention, 149, 105847.

    Endriulaitienė, A., Šeibokaitė, L., Markšaitytė, R., Slavinskienė, J., Crundall, D.,Ventsislavova, P. (2022). Correlations among self-report, static image, and video-based hazard perception assessments: The validity of a new Lithuanian hazard prediction test . Accident Analysis & Prevention, 173. 106716.

    Leighton, J. (2017). Using think-aloud interviews and cognitive labs in educational research. New York, NY: Oxford University Press.

    Roelofs, E. (2019). A Framework for Improving the Accessibility of Assessment Tasks. In B. P. Veldkamp, & C. Sluijter (Eds.), Theoretical and Practical Advances in Computer-based Educational Measurement (pp. 21–45). Cham: Springer International Publishing. doi:10.1007/978-3-030-18480-3_2

    Roelofs, E. C., Emons, W. H., & Verschoor, A. J. (2021). Exploring task features that predict psychometric quality of test items: the case for the Dutch driving theory exam. International Journal of Testing, 21, 80–104. doi:10.1080/15305058.2021.1916506

  • MCP8 - Adapting automated vehicle behaviour to user trust: A driving simulator study


    Internship (10EC) + Master Thesis (25EC)

    Trust is one of the main factors slowing down the adoption of automated driving technology.

    Trust is one of the main factors slowing down the uptake of automated driving technology. Recent findings suggest that trust may be improved by tailoring automated vehicle behaviour to each user: can adaptive automation lead to the safe use of automated vehicles? Researchers from Leiden University and  University of Twente are interested in answering this question.

    Please not that this internship and thesis are performed in combination.


    By joining this internship, you will work in close contact with the BMS Lab, fine tuning the driving simulator of the University of Twente. Furthermore, you will perform a thorough literature review, mapping findings linked to the use of adaptive automation in the automotive domain. Finally, you will work in Leiden University’s Driving Simulator Lab. Here, you will help with the development of a lab protocol, to be followed during one of Leiden’s upcoming studies.


    -          Collaborate with the BMS Lab, fine-tuning the driving simulator

    -          Perform a thorough literature review focused on the theme “adaptive automation”

    -          Develop a lab protocol, to be followed in Leiden University’s Driving Simulator Lab


    -          This study aims to improve trust by tailoring automated vehicle (AV) behaviour to each user. In the driving simulator of the UT, while being driven by an AV, participants will continuously report how much they trust its behaviour. They will indicate their trust through a slider, with “0” indicating “No trust” and 100 indicating “Full trust”. An alert will be sent to the experimenter if the participant’s trust will be 10% below or above the previously reported value. The speed of the AV will decrease or increase every time an alert is received. We hypothesize that, when compared to two control groups (Control1: speed stays constant; Control2: speed randomly changes), adapting vehicle’s speed to the user’s trust will lead to higher trust levels in individuals that tend to distrust automation.


    -          A motivated Master student in psychology (Cognitive or Human Factors and Engineering)

    -          Good communication, organizational and writing skills

    -          Available for 3 or 4 days a week

    -          A true interest in the future of transportation is a plus

    If you want to know more about this internship position, please send an email to Francesco:

  • MHF9 - Internship + Master thesis @ Deep Blue (Rome, Italy)

    Internship (10EC) + Thesis (25 EC)

    Deep Blue is a Human Factors and Safety research and consultancy agency providing solutions for operating in safety critical sectors.

    • Deep Blue was founded in 2001 in Rome, Italy.
    • Deep Blue employs 55 qualified and young staff members (half of which have a PhD).
    • Deep Blue owns 70% equity of an innovative start up focused on safety of drone operations.
    • Over 100 EU funded R&D projects since 2001 ( 40 ongoing).
    • Established supplier of HF services to major organisations in the aviation sector.

    We are active in the following sectors:

    • Transport (ATM, aviation, railway, multi modal, space, maritime)
    • Manufacturing
    • Secure Societies
    • Environment
    • Healthcare

    Our main activities are:

    • Research for EU Commission and Single European Sky
    • Consultancy for large companies and international organisations like Eurocontrol and World Food Program
    • Training courses for Eurocontrol, International Air Transport
    • Association, Joint Aviation Authorities, European Space Agency

    Available projects:

    Project 1: HAIKU Project


    • Pave the way for human centric AI via the exploration of interactive AI prototypes in a wide range of aviation contexts
    • Deliver truly human centric Digital Assistants, capable to ‘ the way humans work


    • Deliver prototypes of AI assistants, demonstrated in the different use cases
    • Design human machine teaming for the different aviation applications
    • Define characteristics and strategy for AI explainability
    • Define and test human in the loop learning strategies
    • Determine how human role will evolve
    • Develop new Human Factors design guidance and methods for AI
    • Develop new safety and validation assurance methods for Digital Assistants

    Internship goal
    Support the development of the project vision, scenarios and frameworks

    Internship tasks

    • Support the definition of the HAIKU foundation (vision and principles), through interactive   
      activities with the Project Consortium
    • Contribute to the development of the HAIKU future scenarios 2030 and 2050
    • Support the development of models concerning human AI teaming
    • Contribute to the first step of the societal impact analysis and to the development of a Framework concerning Societal Acceptance of AI

    Project 2: Industry 4.0 – H-AI interaction

    Internship goal: to identify, analyse and consolidate the needs of the current workforce working with innovative machines using AI in order to inform, from a holistic/UX point of view, the re-design of the systems.

    Expected Outcomes: users’ needs and recommendations to be identified through user research activities, concepts development for H-AI teaming.

    Global project with challenges related to cultural and language differences.

    Methodological challenges related to COVID-19.

    Expected Contribution:

    • Helping in definition of User Research methods for remote interaction
    • Data collection (e.g. interviews, focus groups…)
    • Data analysis and report creation


    Start in February or March with a duration of 6 months (please take into account that we close for two weeks in August).

    This internship + thesis requires some adaptions to the planning/timeline of your curriculum, if interested please contact your internship coordinator to discuss practicalities.

    • The internship takes place at our office in Rome
    • Compensation of €500 a month and the possibility of applying for an Erasmus+ grant (+/-€420 a month)
    • One available spot (but you won’t be the only intern as we are planning on hiring other interns from different universities/study programmes)
    • To apply, please send us your CV and your motivation. It is also possible to propose an internship topic yourself. Mail questions & applications to