EATING
The forgotten snack
In a 2015 study of 40 adults, looking back through camera photos helped people recall 265 foods they had left out of their diet interviews, often snacks.
See the numbersWHAT THE STUDIES FOUND
The camera that
remembers for you.
We are poor witnesses to our own day.
CLICK every 10 seconds
≈ 3,600 a day Photos taken on a typical day
- Memory leaves gaps
- Photos fill them in
- Consent comes first
- An independent guide to the research
Asked what we ate, how long we sat or how long the walk took, we guess. Researchers tested those guesses against small cameras worn on a lanyard that quietly take a photo every few seconds.
THE SHORT ANSWER
Our memory edits
the day. Photos don't.
Researchers compared what people reported
with what the camera saw.
SITTING
Hidden sitting time
Coding photos for posture gave a sitting estimate that differed by nearly 30 minutes a day from a motion sensor alone.
Compare the studies
PRIVACY
A lens points both ways
Automatic cameras also photograph other people. An Oxford-led ethics framework puts consent, privacy and confidentiality at the centre.
How studies handle itFOUR STUDIES, SIDE BY SIDE
What did the
camera add?
Pick a study to see how far the camera's record differed from what people reported or remembered. The bar shows how big the gap was.
Summaries of published, peer-reviewed studies. This is general information about research, not medical or dietary advice.
HOW STUDIES PROTECT PEOPLE
Filming a day,
fairly.
A camera that clicks every few seconds captures the wearer's family, colleagues and strangers too. These are the safeguards the published studies and ethics framework describe.
Ask before the lens goes on
The Oxford-led ethics framework puts respect for autonomy first, through proper informed consent for wearers.
Think about the people nearby
Third parties may feel their privacy has been affected. The framework calls for privacy and confidentiality controls for everyone captured.
Give the wearer an off switch
In the school-travel study, teenagers could remove the camera at any time and pause recording for seven minutes with a privacy button.
Let the wearer review first
In the diet study, participants viewed their own photos and chose what to add to their account of the day.
TWO NUMBERS WORTH KNOWING
Thousands of photos,
a few honest gaps.
One figure on how much a worn camera records,
one on what it caught that people forgot.
≈ 3,600 a day
Photos taken on a typical day
The research camera used in an Oxford school-travel study weighed 175 g and took approximately 3,600 first-person photos on a typical day, with no action from the wearer.
IMAGES · OXFORD STUDY
265 foods
Foods recalled only after seeing photos
Forty adults wore a camera before three diet interviews. Viewing the photos added 265 unreported foods, often snacks, and brought reported intake closer to a lab measure.
DIET · BRIT J NUTR 2015
Even with the camera, the same diet study found intake was still under-reported by 9% for men and 7% for women: photos narrow the gap rather than close it.
WHERE CAMERAS STRUGGLE
Not a perfect
witness either.
In the school-travel study, 15 of 150 journey stages were lost: 8 because the camera was not worn and 7 because the lens was covered at the start or end of the trip.
The researchers also noted that very low light made journeys hard to see, clothing could block the lens, some places (such as a friend's house) made people uncomfortable wearing it, and a photo every 10 seconds adds a small timing error.
Questions
Questions people ask
What is a wearable camera in research?
A small camera worn on the body, often on a lanyard at chest height, that takes photos automatically. The one used in an Oxford school-travel study weighed 175 g and took about 3,600 photos on a typical day.
Why not just ask people what they did?
Self-reports drift. In a 2015 diet study, adults under-reported energy intake by 13 to 17% in standard interviews; reviewing camera photos reduced that gap by 6 to 8%.
Did the camera help people remember?
In a 2007 case study of one woman with severe memory loss, she recalled about 80% of recent events after viewing camera photos, compared with around 49% using a written diary. It was a single case, not a general result.
What about the people who get photographed?
An ethics framework from Oxford-led researchers says third parties may feel their privacy is affected, and calls for informed consent and privacy and confidentiality controls to protect everyone.
Is the camera always right?
No. Studies lost images when the lens was covered or the camera was not worn, and low light made some scenes unreadable. Researchers treat photos as one record among several.
Where the facts come from
Updated
- Gemming et al. (2015) Wearable cameras can reduce dietary under-reporting, British Journal of Nutrition
- Kelly et al. (2012) Evaluating the feasibility of measuring travel to school using a wearable camera, Am J Prev Med
- Kerr et al. (2013) Using a wearable camera to improve classifications of sedentary behavior, Am J Prev Med
- Kelly et al. (2013) An ethical framework for automated, wearable cameras in health behavior research, Am J Prev Med
- Berry et al. (2007) A wearable camera as a pictorial diary to improve autobiographical memory, Neuropsychological Rehabilitation