Intervention description {11a}

In the intervention group, the computer screen will be set up in a split-screen format. On the left side of the screen, the participant will receive instructions on the writing prompt, writing requirements, time requirements, grading feedback, and the grading rubric. The instructions will also highlight to the participant that they can use ChatGPT in any way they like to assist their writing, and there is no penalty in their writing score for how ChatGPT is used. The right side of the screen will display a blank ChatGPT interface where the participant can prompt questions and receive answers.

Explanation for the choice of comparators {6b}

In the control group, as in the intervention group, the computer screen will be set up in a split-screen format. On the left side of the screen, the participant will receive the same instructions on the writing prompt, writing requirements, time requirements, grading feedback, and the grading rubric. Additionally, the instructions will highlight to the participant that they can use a text editor in any way they like to assist their writing. On the right side, instead of ChatGPT, a basic text editor interface will be displayed. In summary, this comparator will keep the split-screen format consistent between the two groups and ensure that participants in the control group will complete the writing task with minimal support.

Criteria for discontinuing or modifying allocated interventions {11b}

This study is of minimal risk, and we do not anticipate needing to discontinue or modify the allocated interventions during the experiment. Participants can withdraw from the study at any time.

Strategies to improve adherence to interventions {11c}

Adherence to the interventions will be high because the procedures are straightforward and will be clearly explained in the step-by-step instructions on the computer screen. The participant will be alone in a noise-canceling room during the entire experiment. The participant can reach out to the experimenter through an intercom if they need any clarification.

Relevant concomitant care permitted or prohibited during the trial {11d}

Not applicable. This is not a clinical study.

Provisions for post-trial care {30}

Not applicable. This is a minimal-risk study.

Outcomes {12}

The study has two primary outcomes. First, we will measure participants’ writing performance scores on the analytical writing task. The task is adapted from the Analytical Writing section in the GRE, a worldwide standardized computer-based exam developed by the Educational Testing Service (ETS) [27]. Participants’ writing performance will be scored using the GRE 0–6 rubric and by an automatic essay-scoring platform called ScoreItNow!, which is powered by ETS’s e-rater engine [32, 33]. We chose to adapt from the GRE writing materials for two reasons. First, their writing task and grading rubrics were established writing materials designed to measure critical thinking and analytical writing skills and have been used in research as practice materials for writing (e.g. [34]). Second, OpenAI’s technical report shows that ChatGPT (GPT-4) can score 4 out of 6 (~ 54th percentile) on the GRE analytical writing task [31]. This gives us a benchmark for assessing the potential increase in writing performance when individuals collaborate with generative AI.

Second, we will measure participants’ cognitive effort during the writing process. Participants’ cognitive effort will be measured using a psychophysiological proxy—i.e., changes in pupil size [35, 36]. Pupil diameter and gaze data will be collected using the Tobii Pro Fusion eye tracker at a sampling rate of 120 Hz. During the preparation stage of the study, the room light will be adjusted so that the illuminance at the participants’ eyes is at a constant value of 320 LUX. Baseline pupil diameters will be recorded during a resting task in the experiment preparation stage that asks the participant to stare at a cross that will appear for 10 s each on the left, center, and right sections of the computer screen. Pupil diameters and gaze data will be recorded throughout the writing process.

The study has several secondary outcomes. First, to identify the neural substrates of cognitive effort during the writing process, we developed an additional psychophysiological proxy, changes in the cortical hemodynamic activity in the frontal lobe of the brain. Specifically, we will examine hemodynamic changes in oxyhemoglobin (HbO). Brain activity will be recorded throughout the writing process using the NIRSport 2 fNIRS device and the Aurora software with a predefined montage (Fig. 2). The montage consists of eight sources, eight detectors, and eight short-distance detectors. The 18 long-distance channels (source-detector distance of 30 mm) and eight short-distance channels (source-detector distance of 8 mm) are located over the prefrontal cortex (PFC) and supplementary motor area (SMA) (Fig. 2). The PFC is often involved in executive function (e.g., cognitive control, cognitive efforts, inhibition) [37, 38]. The SMA is associated with cognitive effort [39, 40]. The sampling rate of the fNIRS is 10.2 Hz. Available fNIRS cap sizes are 54 cm, 56 cm, and 58 cm. The cap size selected will always be rounded down to the nearest available size based on the participant’s head measurement. The cap is placed on the center of the participant’s head based on the Cz point from the 10–20 system.

Third, we will measure participants’ subjective perceptions of the writing task by self-reported survey measures in the post-survey (Table 1). We will measure participants’ subjective perceptions of the two primary outcomes—that is, their self-perceived writing performance and self-perceived cognitive effort. Self-perceived writing performance will be measured with a one-item scale using the same grading rubric described in the instructions for their writing task and used in the scoring tool. Self-perceived cognitive effort will be measured using a one-item scale adapted from the National Aeronautics and Space Administration-task load index (NASA-TLX) [41, 42]. We will also measure participants’ subjective perceptions of several mental health and learning-related outcomes, including stress, challenge, and self-efficacy in writing. Self-perceived stress will be measured using a one-item scale adapted from the Primary Appraisal Secondary Appraisal scale (PASA) [43, 44]. Self-perceived challenge will be measured using a one-item sub-scale adapted from the Primary Appraisal Secondary Appraisal scale (PASA) [43, 44]. Self-efficacy in writing will be measured using a 16-item scale that measures three dimensions of writing self-efficacy: ideation, convention, and self-regulation [45]. Furthermore, we will measure participants’ situational interest in analytical writing using a four-item Likert scale adapted from the situational interest scale [46]. Additionally, we will measure participants’ behavioral intention to use ChatGPT in the future for essay writing tasks [47].

Participant timeline {13}

The time schedule is provided via the schematic diagram below (Fig. 3). The entire experiment will last for approximately 1–1.5 h for each participant.

Schedule of enrollment, interventions, and assessments of the study

Sample size {14}

To estimate the required sample size, we conducted a simulation analysis on the intervention effect on writing performance using ordinary least squares (OLS) regression. Recent empirical evidence suggests that the effect size of generative AI on writing tasks ranges around Cohen’s d = 0.4–0.5, such as [1, 48]. In our simulation analysis, the simulated data assumes normally distributed data, equal and standardized standard deviations between the two conditions, and an anticipated effect size of Cohen’s d = 0.45. In the end, our analysis indicated that recruiting a minimum of 160 participants would be necessary to achieve a statistical power greater than 0.8 under an alpha level of 0.05. The simulation was implemented in R, and the corresponding code is available at the Open Science Framework (OSF) via https://osf.io/9jgme/.

We opt to base our sample size estimation on writing performance, but not on the other primary outcome, cognitive effort, for two reasons. First, the effect of generative AI on performance outcomes has been studied [1, 48], but we did not find prior evidence on the effect size of generative AI on cognitive effort using physiological measures. Second, our physiological measure of cognitive effort may likely be powered once the sample size satisfies our behavioral measure of writing performance. Pupillometry studies on cognitive efforts, such as the N-back test, typically recruit 20–50 participants in short, repeated, within-subject trials (e.g., [49]). These studies provide a general estimation of participants needed. Although our study design (i.e., a between-subject RCT) differs from common pupillometry studies, cognitive effort is still a repeated outcome measure using time series pupil data throughout the entire writing process. Repeated outcome measures generally can enhance statistical power by taking into account within-subject variability [50].

Recruitment {15}

The recruitment will follow a convenience sampling strategy. To aim for a student population with diverse academic backgrounds, participants will be recruited broadly through social media platforms, email lists, and flyers at the research university where the experiment will be conducted. Given that the experiment will start during the summer, the research team can recruit summer school students as participants. Thus, the study sample will not be limited to the students presently at the university. The recruitment materials include a brief description of the study, the eligibility criteria for participation, and the compensation for participation. Individuals who are interested in participation can sign up on a calendar by selecting available time slots provided by the experimenters. Participants will receive 30 euros in compensation upon completion of the experiment. Participants who withdraw in the middle of the experiment will receive partial compensation, prorated based on the amount of time they spend in the experiment.

Insider Brief

• A Canadian-led research team from TRIUMF and the Perimeter Institute has developed a quantum-assisted AI model to simulate particle collisions more efficiently, addressing global computational challenges.
• The study demonstrates that combining deep learning with quantum computing—using technology from D-Wave—can significantly reduce the time and cost of high-energy physics simulations.
• Published in npj Quantum Information, the work supports future upgrades to CERN’s Large Hadron Collider and underscores Canada’s growing leadership in quantum and AI-driven scientific research.

PRESS RELEASE — In a landmark achievement for Canadian science, a team of scientists led by TRIUMF and the Perimeter Institute for Theoretical Physics have unveiled transformative research that – for the first time – merges quantum computing techniques with advanced AI to model complex simulations in a fast, accurate and energy-efficient way.

“This is a uniquely Canadian success story,” said Wojciech Fedorko, Deputy Department Head, Scientific Computing at TRIUMF. “Uniting the expertise from our country’s research institutions and industry leaders has not only advanced our ability to carry out fundamental research, but also demonstrated Canada’s ability to lead the world in quantum and AI innovation.”

Traditional simulations of particle collisions are already both time-consuming and costly, often running on massive supercomputers for weeks or months. By leveraging quantum processes and technology made possible by California-based D-Wave Quantum Inc., the researchers were able to create a new “quantum-assisted” generative model capable of running simulations and open new opportunities to cost-effectively analyze rapidly growing data sets.

The research, published today in npj Quantum Information, is part of a worldwide effort to create the tools needed to accommodate upgrades to CERN’s particle accelerator, the Large Hadron Collider (LHC), and alleviate a computational bottleneck that would impact researchers all over the world.

“Our method shows that quantum and AI technologies developed here in Canada can solve real-world scientific bottlenecks,” said Javier Toledo-Marín, joint appointee at TRIUMF and Perimeter Institute. “By combining deep learning with quantum technology, we are forging a new path for both theoretical experimentation and technological application.”

In addition to TRIUMF and Perimeter, contributions to the published research came from the National Research Council of Canada (NRC), the University of British Columbia and the University of Virginia, showcasing not only the wealth of research talent and scientific ingenuity across the country, but also the international collaboration that places Canada at the forefront of worldwide scientific innovation.

Apple researchers, in collaboration with the University of Southern California, have developed a new artificial intelligence (AI) model that tracks behavioural data over sensor signals. The new research builds on prior work by the Apple Heart and Movement Study (AHMS) and was aimed at understanding if behavioural data, such as sleep pattern and step count, can be a better determinant of a person’s health compared to traditional indices such as heart rate and blood oxygen level. As per the paper, the AI model performed surprisingly well, even if with some caveats.

New Apple Study Shows Benefits of Moving Beyond Traditional Health Data

The study, titled “Beyond Sensor Data: Foundation Models of Behavioral Data from Wearables Improve Health Predictions” was published in the pre-print journal arXiv and is yet to be peer reviewed. The researchers set out to develop an AI model, dubbed Wearable Behaviour Model (WBM), that relies on processed behavioural data from wearables such as how long a person sleeps and their REM cycles, daily steps taken and gait, and how their activity pattern changes over the week.

Traditionally, to predict or assess someone’s health, wearable health research has typically focused on raw sensor readings such as continuous heart rate monitoring, blood oxygen levels, and body temperature. The study believes that while this data can be useful at times, it also lacks the full context about the individual and can have inconsistencies.

Regardless, so far, behavioural data, which is also something most wearables process, has not been used in systems as a reliable indicator of a person’s health. There are two main reasons for it, according to the study. First, this data is much more voluminous compared to sensor data, and as a result, it can also be very noisy. Second, creating algorithms and systems that can collect and analyse this data and reliably make health predictions is very challenging.

This is where a large language model (LLM) comes in and solves the analysis problem. To solve the noise in data, researchers fed the model with structured and processed data. The data itself comes from more than 1,62,000 Apple Watch users who participated in the AHMS research, totalling more than 2.5 billion hours of wearable data.

Once trained, the AI model used 27 different behavioural metrics, which were grouped into categories such as activity, cardiovascular health, sleep, and mobility. It was then tested across 57 different health-related tasks, such as finding out if someone had a particular medical condition (diabetes or heart disease) and tracking temporary health changes (recovery from injury or infection). Compared to the baseline accuracy, researchers claimed that WMB outperformed in 39 out of 47 outcomes.

Comparison between performance of the WBM model the test model and the combination of both
Photo Credit: Apple

The findings from the model were then compared with another test model that was only fed raw heart data, also known as photoplethysmogram (PPG) data. Interestingly, when individually compared, there was no clear winner. However, when researchers combined the two models, the accuracy of prediction and health analysis was measured to be higher.

Researchers believe combining traditional sensor data with behavioural data could improve the accuracy in the prediction of health conditions. The study stated that behavioural data metrics are easier of interpret, align better with real-life health outcomes, and are less affected by technical errors.

Notably, the study also highlighted several key limitations. The data was taken from Apple Watch users in the US, and the broader global population was not represented in this. Additionally, due to the high price of wearable devices that accurately collect and store behavioural data, accessibility of preventive healthcare also becomes a challenge.