So, what is Kitchen Life 2?
Kitchen Life 2 is a new social research project exploring real-life behaviours in kitchens using motion-sensitive video capture technology, as well as surveys, interviews, food diaries, and other types of observation (such as monitoring fridge temperatures) to understand behaviour in domestic and commercial kitchens.
Why are we doing this?
We hope to identify the key behaviours relating to food safety that occur in domestic and business kitchens; where, when, how often and with whom they occur alongside understanding the behavioural barriers and levers underlying food hygiene behaviours
Why is this important for the FSA?
It is important for the Food Standards Agency to understand how people behave in kitchens in real life (or as close as is possible) to avoid the biases often present in self-reported measures. This project will help us to accurately assess the risks to consumers with regards to foodborne disease and develop interventions that could reduce these risks.
How are we observing real life behaviours?
Over the past 18-months, we have put motion-sensitive cameras into a range of domestic and business kitchens. Cameras are in place for a week, with filming taking place for up to five days in each kitchen.
We are also using surveys and interviews alongside video footage, to understand how people make sense of their behaviours and to understand the intention-behaviour, or ‘say vs do’ gap. Other data, such as fridge photographs and fridge thermometers will also help us to understand the safety of food storage.
Data is categorised into a bespoke dashboard, allowing FSA analysts to search for behaviours of interest. For example, the dashboard allows us to quantify how many times hand washing occurred after raw meat preparation. The data will be analysed against the COM-B framework, which will allow us to create hypotheses around potential behaviour change interventions targeting specific unhygienic behaviours.
Do participants change their behaviour while being filmed?
We film 5 days of footage in total, but we only analyse 3 days at random. Having the camera in place for 5 days helps participants to feel more accustomed to having the camera in their kitchen. The video footage we have captured so far shows a wide variety of behaviours (including some which highlight food safety risks), so we really do believe that we are seeing people behave naturally in front of the camera.
What If you see something you were not meant to see?
As you can imagine, when filming people in domestic or business kitchens there are a lot of ethical and data protection issues to consider. We sought out ethics advice from our Advisory Committee on Social Science, as well as data protection advice from our Knowledge Information Management Security team to discuss, and develop risk assessments, on the following:
- Blurring filmed faces
- Who will have access to the footage
- Legal protocols if we see anything illegal
- Signage to let people in the business kitchens know that filming is occurring
- Compensating participants for their time
- The consent processes, particularly in food businesses where we wanted to make sure employees did not feel pressured by their employers to take part in the study. We made sure they could opt out without going through their employer.
Our advice to you when using a novel method
From commissioning the project, to fieldwork, we have come across a few bumps in the road! However, like all good challenges, we have learnt valuable lessons! Below are a few top tips when using a novel method.
- Seek advice from experts from inception – At the FSA we have an Advisory Committee for Social Science (ACSS). Kitchen Life 2 has benefited from its own Working Group made of academic experts who have inputted at various stages of the project to help shape it. The literature review was led by Dr Gulbanu Kaptan from Leeds University Business School. Dr Kaptan and our Working Group members helped us discuss the emerging findings from our literature review and shape the study design.
- Use multiple data sources – capturing observational data is rich but analysing the data using different methods (triangulation) really adds a lot of value. Triangulation can reduce biases that comes from using one single method and increase validity. We often saw a mismatch between observed behaviour and what participants thought they did. We were able to use what we saw in the footage to probe participants’ perceived behaviours during their interviews. The say-do gap seems largely down to lack of awareness (habitual behaviours), rather than social desirability bias.
- Have a backup plan before something goes wrong – When relying heavily on technology, and dealing with human beings, you never know when something will go wrong. For example, after filming a food business for a week, we realised the camera was not angled correctly. Another unexpected challenge we faced included food businesses not having adequate WiFi strength to stream the footage back to our contractors who were analysing the data. Having contingency plans in place enabled us to swiftly resolve these issues, and not lose businesses (who are hard to recruit) from our sample.
What have we found out so far?
Fieldwork is still in progress, so it is difficult for us to make conclusions at this stage. However, initial footage from domestic kitchens has shown concerning behaviours in relation to:
- The safe storage of leftovers
- Adherence to use-by dates
- Handwashing in domestic kitchens.
- Mobile phone use in food businesses kitchens.
Interestingly, many concerning behaviours seem to be done without awareness, or habitually, with participants sometimes not realising what they have been doing until probed in the follow-up interviews. This presents interesting challenges for the development of behavioural interventions, where subconscious cues may be more important than education or raising conscious awareness.
Next steps…
Detailed analysis on the dataset is taking place to answer queries on specific risky behaviours, to feed into the FSA’s risk assessment models.
Once all data is collected and analysed using the COM-B framework, we will bring together policy colleagues and academics to hold a workshop to generate potential behavioural interventions for food safety behaviours. This could cover anything from cooking to cleaning, depending on what people are doing in their kitchen, how risky it is, and how easy it could be to change behaviour.
We are co-supervising a PhD with Leeds University Business School, to further explore the “say-do” gap, by analysing the data from this project alongside our more traditional survey data. We recently have funding to carry out secondary analysis on our data. The FSA five-year strategy (2022-2027) was published in March 2022, which reflects the FSA’s greater concern around climate change. Consistent with our extended remit, the secondary analysis will have a sustainability focus around food waste behaviour.
Want to know more?
If you’d like to hear more about the project, the methodologies used or the data protection considerations we’ve had to make, please get in touch with either [email protected] or [email protected]. If you want to keep an eye out on results, head over to our website.
Author biography
Helen Heard is a Senior Social Science Research Officer and has worked in the FSA for two years. Previously Helen worked as a statistician in NHS England and in the Department for Health and Social Care.
Alice Rayner is a Principle Social Science Research Officer and has worked in the FSA for 4 years. Previously Alice worked in HMRC while on the Civil Service Fast Stream.
Anya Mohideen is a Social Science Research Officer and has worked in the FSA for a year. Previously Anya worked in an Obesity Clinic.
Acknowledgments
Huge thank you to the team at Basis Social, Lifestream, Dr Gülbanu Kaptan and Dr Joshua Weller from Leeds University Business School and our Advisory Social Science Committee for all their work on this project so far!