Publications
What Your Data Helped Us Discover
Publications from the Neureka citizen science project
Every study below exists because thousands of people gave their time to answer questions and play brain games on their phones. Each one is written in plain language first, with the full citation underneath
9 studies published or in preprint
13,000 + participant contributions analysed
7 peer-reviewed journals
Peer Reviewed Publications
Social disadvantage outweighs knowledge in dementia risk.
Public health campaigns often assume that if people know more about dementia, they will act to lower their risk. Looking at 1,730 adults, we found that education level and socioeconomic circumstances were far stronger predictors of a person’s dementia risk profile than how much they actually knew about dementia. This suggests awareness alone will not close the gap — reducing dementia risk across a population also means addressing social disadvantage.
Baez, S., Rodriguez Marquez, S., Rosická, A. M., & Gillan, C. M. (in press). Social disadvantage outweighs knowledge in dementia risk. Alzheimer’s & Dementia.
ALZHEIMER'S & DEMENTIA
Digital questionnaire response time (DQRT): a ubiquitous and low-cost digital assay of cognitive status.
Every time you answer a question in an app, we can quietly record how long it took you. Using data from 2,977 Neureka users, we tested whether this simple timing measure says anything meaningful about cognition — and found it tracks age, education and performance on proper tests of processing speed. It has limitations, but it means we can measure processing speed at enormous scale, repeatedly, without asking anyone to do an extra task.
BEHAVIOUR RESEARCH METHODS
Teckentrup, V., Rosická, A. M., Donegan, K. R., Gallagher, E., Hanlon, A., & Gillan, C. (2025). Digital questionnaire response time (DQRT): a ubiquitous and low-cost digital assay of cognitive status. Behavior Research Methods, 57(7), 1–20.
Modifiable dementia risk factors associated with objective and subjective cognition.
We looked at how changeable dementia risk factors — depression, hearing difficulty, loneliness, smoking, exercise, diabetes, and others — relate to two different things: how people perform on cognitive tests, and how people rate their own memory. Almost all of these risk factors showed up more clearly in people’s own reports about their memory than in their test scores, and this held even after accounting for depression. How your memory feels to you may be an early and sensitive signal worth taking seriously.
ALZHEIMER'S & DEMENTIA
Rosická, A. M., Teckentrup, V., Fittipaldi, S., Ibanez, A., Pringle, A., Gallagher, E., … & Gillan, C. M. (2024). Modifiable dementia risk factors associated with objective and subjective cognition. Alzheimer’s & Dementia.
Reliable, rapid, and remote measurement of metacognitive bias.
We built a smartphone game called Meta Mind to measure how much confidence people have in their decisions. Across 3,410 citizen scientists, we validated this test and found that people with higher anxiety and depression tended to be less confident than their performance warranted, while people with more compulsive behaviours and intrusive thoughts tended to be more confident.
Fox, C. A., McDonogh, A., Donegan, K. R., Teckentrup, V., Crossen, R. J., Hanlon, A. K., … & Gillan, C. M. (2024). Reliable, rapid, and remote measurement of metacognitive bias. Scientific Reports.
SCIENTIFIC REPORTS
Confidence biases in problem gambling.
Building on the confidence work above, we asked whether people with gambling problems judge their own decisions differently. Problem gamblers were considerably more confident than other participants, not actually doing any better on the task. This fits with a ‘transdiagnostic’ view of mental health, where gambling disorder shares some features with OCD and other addictions that have a ‘compulsive’ feature.
Friedemann, M., Fox, C. A., Hanlon, A. K., Tighe, D., Yeung, N., & Gillan, C. M. (2024). Confidence biases in problem gambling. Journal of Behavioral Addictions.
JOURNAL OF BEHAVIOURAL ADDICTIONS
Using smartphones to optimise and scale-up the assessment of model-based planning.
Planning ahead is thought to protect us from running on autopilot and getting stuck in habits. We have found previously that people who plan less tend to be more compulsive, but the effect is small and the lab tests are long. We rebuilt the standard test as a diamond-shooting smartphone game and found we could recover these same effects at scale. This is important because if we can measure this in a game, it means we can start to study how planning changes as symptoms change over time and start to understand causality
Donegan, K. R., Brown, V., Price, R. B., Pringle, A., Gallagher, E., Hanlon, A., & Gillan, C. (2023). Using smartphones to optimise and scale-up the assessment of model-based planning. Communications Psychology.
COMMUNICATIONS PSYCHOLOGY
Elevated emotion network connectivity leaves people vulnerable to fluctuations in depression.
In some people, emotions move together. When sadness rises, guilt and irritability rise with it. We asked 349 people to rate their emotions twice a day for eight weeks and to complete a weekly depression questionnaire. We used a network analysis and found that having tightly linked emotions predicted how much a person’s depression swung up and down over time — pointing to instability rather than severity as the key risk. We replicated our findings in other datasets too, confirming that data from our app is similar to data gathered using other more labour intensive methods. Yay - citizen science!
Kelley, S., Fisher, A. J., Lee, C. T., Gallagher, E., Hanlon, A., Robertson, I., & Gillan, C. (2023). Elevated emotion network connectivity leaves people vulnerable to fluctuations in depression. Proceedings of the National Academy of Sciences.
PNAS
Preprints
Preprints are papers we have shared publicly before peer review is complete, so that the findings are available to everyone straight away.
Understanding the habitual properties of compulsivity through experience sampling
Are compulsions driven by distressing thoughts and anxiety, or are they closer to habits? We asked 180 people, some with OCD-type compulsions such as checking and cleaning, some who binge eat, some who compulsively check their phones, to report on their day three times daily for six weeks. Compulsive behaviour tracked urges roughly three to four times more closely than it tracked obsessive thoughts or anxiety, and occurred more often when self-control was running low, supporting a habit-like mechanism shared across very different compulsions.
Teckentrup, V., Fox, C. A., Donegan, K. R., Gallagher, E., Hanlon, A., Fineberg, N. A., … Gillan, C. (2025). Understanding the habitual properties of compulsivity through experience sampling.
Compulsivity is associated with accelerated stimulus-response habit learning.
A long-standing theory holds that compulsive behaviour comes from an overactive habit system, but until now the evidence was indirect. In a pre-registered study, we trained 1,000 people on a smartphone habit- learning task over seven days and confirmed that habits grew stronger with more practice and under time pressure. People higher in compulsivity formed these habits more readily and this was specific to compulsivity, not to anxiety, depression, tics, skin-picking or hair-pulling.