9th November 2020, 10:00-14:30 Understanding food in a digital world An FSA ESRC festival of Social Science Event, in partnership with the University of Sheffield Understanding food in a digital world Opening remarks from Professor Robin May and Julie Pierce (10:05-10:20) Audience link 1 Session 1. Digitalisation of food (10:20-11:05)
Session 2. Innovative data collection: Online survey methods (11:05-11:35)
Break (11:35-11:45) Audience link 2 Session 3. Innovative data collection: Other methods (11:45-12:30)
Lunch (12:30-13:10)
Session 4. The Potential of Social Media Analysis (13:10-13:55) Audience link 3 Session 5. Public engagement with the National Food Strategy (13:55-14:25)
Closing Remarks from Professor Rick Mumford (14:25-14:30) Opening Remarks:
Professor Robin May FSA Chief Scientific Advisor Opening Remarks:
Julie Pierce FSA Director of Openness, Data and Digital, and Wales Session 1. Digitalisation of food (10:20-11:05)
10:20 Michelle Patel (Head of Social Science, FSA) The impact of Covid-19 on consumers and the food system in the UK
10:30 Lucy Hart (Analyst, Foundry4) Understanding online platforms and their impact on the food landscape
10:40 Dr Jose A Bolanos (Post-Doctoral Research Officer, FSA & London School of Economics) Food in the Platform Economy: Leapfrogging risk regulation in the context of the platform economy
10:50 Session 1 Panel Discussion The impact of coronavirus on consumers of food Why is the FSA looking at this?
It’s our job to regulate the food system, guided by the interests of consumers so we need to understand what those are now, and what they might be in the future We use a range of tools – • operational analytics • data science • the latest economic, market and social research • intelligence from the food industry and other government departments,
As well as our own primary research, including qualitative research, consumer surveys and social media listening. Before Covid-19
Women twice as likely to do the food shop and the cooking
The wealthiest 10% spend four times more than the lowest income decile on food.
The average UK household spends £3,224 on groceries and £1,581 on restaurants and takeaways every year.
Most people shop for food in large supermarkets Before Covid-19
Most people eat twice as much sugar and more saturated fat than is recommended.
Most adults say they eat 4+ portions of fruit/veg a day, with teenagers eating 2.7.
About a quarter of people pay attention to nutritional information (27%) and ingredients (26%) when shopping.
40% say they follow a healthy and balanced diet consistently. 10% rarely or never eat anything they consider healthy. Before Covid-19
Most people enjoy cooking and preparing food (67%) and have time to do so (68%).
However only 5% are willing to spend an hour cooking a weeknight meal.
Consumers say they are concerned about food waste (57%), the amount of sugar in food (53%), animal welfare (50%) and food prices (44%).
Over time, consumption of red meat and cooked meats has gradually fallen. What’s different now…
Consumers shop less frequently but spend more per trip.
Online ordering and delivery increased by 75% and remain up 76% compared to 2019. Local shopping has increased.
Frozen food sales doubled in the first three months of lockdown.
18% have skipped meals because they couldn’t afford food in the past month. What’s different now…
44% are trying to eat more healthily.
In the average shopping basket, saturated fats increased by 4.7% while sodium increased by 6.2%.
59% bought a takeaway in the last month, with 14% buying one a week.
(Pre-lockdown #2) 21% say they won't eat out for the foreseeable future, What’s different now…
34% of those working at home say their lunches are worse now than when they were in the office.
44% of people are enjoying cooking more. 80% ate with family in the last month.
57% say they value food more than pre-pandemic. Data indicates a 30% reduction in household food waste
12% agree that the pandemic has made plant-based diets more appealing - twice as high among young people and Londoners. Let’s look more closely at food online -
Compared to the period before lockdown, have you done any of the following May 2020 June 2020 July 2020 more or less often?
Had an online food delivery from a supermarket More 20% 24% 26%
About the same 17% 18% 19% Less 18% 15% 12% I never do this 45% 42% 42% Purchased food from a takeaway, either direct or online More 16% 18% 19%
About the same 25% 25% 25% Less 34% 34% 32% I never do this 25% 23% 24% Purchased food from a local supplier (i.e. farm shops, veg box), either direct or More 24% 24% 21% online About the same 25% 26% 24% Less 10% 11% 11% I never do this 41% 40% 43% Had a food delivery from an online food ordering company (e.g. Deliveroo, Just More 14% 17% 17% Eat, Uber Eats)
About the same 18% 19% 20% Less 19% 20% 15% I never do this 49% 45% 48% And more broadly at how the pandemic is making consumers think differently about the food system….
Some reported only perceived negative Risk evaluations disruption to their Awareness of food Logistical pressures relied on often household food systems issues were had caused shifts in implicit calculations habits. Others found more prominent than suppliers for many. of risk versus reward. that lockdown in past research provided a break from routine
Trust in the food industry Widespread fears of Some had become strengthened under food-borne virus much more health pandemic, driven transmission conscious, others primarily by a subsided as time responded by perceived robust went on. snacking ‘return to normal’ after early disruptions Key trends we’re tracking
Food shopping, dining and eating outside the home
Storing, cooking and eating at home
The impact of economic insecurity
The impact of societal changes on diet
Consumer risk perceptions, including acceptance of new foods New data coming out soon
https://www.food.gov.uk/research/research- projects/the-covid-19-consumer-research …or search ‘FSA covid research’
Sources – FSA consumer research including the Covid Consumer Tracker, Public Attitudes Tracker, and Food and You. PHE published research, ONS published research, The Food Foundation, Kantar market data. Session 1. Digitalisation of food (10:20-11:05)
10:20 Michelle Patel (Head of Social Science, FSA) The impact of Covid-19 on consumers and the food system in the UK
10:30 Lucy Hart (Analyst, Foundry4) Understanding online platforms and their impact on the food landscape
10:40 Dr Jose A Bolanos (Post-Doctoral Research Officer, FSA & London School of Economics) Food in the Platform Economy: Leapfrogging risk regulation in the context of the platform economy
10:50 Session 1 Panel Discussion Understanding online platforms & their impact on the food landscape
Lucy Hart, 9th November 2020
1 Lucy Hart [email protected]
/2 0 1 2 3 4
What is an What kinds of How big is the What doesthis online food platform exist? market? mean forthe platform? food landscape?
/2 1 What is an online food platform?
/2 2 Pre 1990s
/2 3 Today
/2 4 /2 5 How does a platform work?
Owns stock Sells >1 vendors’ stock Definition
Vendor Yes No Sells goods to consumers
Retailer Yes Yes Sells goods from multiple vendors to consumers
Platform No Yes Facilitates the sale (or/2 exchange) of goods between vendor(s) and consumer(s) 6 How do platforms describe themselves?
Facebook: "We build technologies that help Uber: “We ignite opportunity by setting the people connect with friends and family, find world in motion.” communities, and grow businesses" Airbnb: “An economic empowerment Deliveroo: aiming to be “the definitive food engine” company” /2 Amazon.com: “aims to be Earth's most Just Eat: “creating the world's greatest food customer-centric company”7 community” Why do vendors & consumers usethem?
/2 8 An online (digital) platform is…
A digital platform is a customer-centric technology business, over and above its involvement in the foodindustry
It facilitates, or brokers, the sale of goods between vendor(s) and consumer(s)
It does not own the stock being sold or exchanged
It instigates an indirect network effect – the more vendors, the more choice for consumers; the more consumers, the more potential revenue for vendors/2
Most platforms make money by charging a commission on each transaction9 What kinds of platform exist?
/3 0 1 2 3 4
What goodsor What’s the What’s the What’s the services are delivery route from nature of the exchanged? timescale? vendor to transaction? consumer?
/3 1 Types of 1. On-demand food delivery platforms enable the on-demand purchase and delivery of goods platform (normally hot meals)
2. Scheduled food delivery platforms enable the scheduled delivery of goods (normally prepackaged or raw goods) 3. Marketplaces facilitate exchanges mostly between individuals and are likely to be fairly informal On-demand delivery platforms
/3 3
The BigThree 1 2 3 4
Restaurants Virtual Home cooks Retailers restaurants What is a virtual restaurant? What is a virtual restaurant? Scheduled food delivery platforms Scheduled food delivery platforms
Food-specific Generalist Marketplaces Marketplaces How big is the market? How big is the market?
Data from Beambox,2019 /44 Kbox
Labels itself: “a risk-free What it does: Kbox does not own kitchens; it solution to increase food sales, finds restaurants, hotels, clubs, pubs, leisure maximise efficiency, whilst centres and other spaces with kitchen space developing an exciting, on (and staff) that is not being fully utilised trend, multi-brand It then identifies the most appropriate virtual Delivery opportunity” restaurant for that location, taking into account the existing offering and the area's Founded: 2019, London demographics Funding: £5 million UK footprint: 30brands, The virtual restaurant is then 'licensed' to the unknown kitchens kitchens to produce food with its recipes and branding for delivery
/45 28 What does this mean for the food landscape?
/46 28 Changes tothe 1. Tech businesses entering food world 2. Supply chain rearranged food landscape 3. Online-as-default businesses 4. Lots of information on supplychain 5. Increased interconnectedness 6. Ever-adapting business models 7. Bundling & unbundling ofsteps 8. Faster emergence & change 9. Lower visibility of business identity 10. Private standard setting
/47 28 Thank you.
Foundry4.com
30 Session 1. Digitalisation of food (10:20-11:05) 10:20 Michelle Patel (Head of Social Science, FSA) The impact of Covid-19 on consumers and the food system in the UK
10:30 Lucy Hart (Analyst, Foundry4) Understanding online platforms and their impact on the food landscape
10:40 Dr Jose A Bolanos (Post-Doctoral Research Officer, FSA & London School of Economics) Food in the Platform Economy: Leapfrogging risk regulation in the context of the platform economy
10:50 Session 1 Panel Discussion FOOD IN THE PLATFORM ECONOMY
LEAPFROGGING RISK REGULATION IN THE CONTEXT OF THE PLATFORM ECONOMY
Dr Jose A Bolanos https://www.josebolanos.xyz
Contact me at [email protected] [email protected]
Or follow me on www.LinkedIn.com/in/jbolns. www.Instagram.com/jbolns. The regulation of the platform economy has been receiving increasing attention for a while already, and COVID-19 highlighted the importance of having an online food ecosystem where people can trust that the food they buy online is safe and what it says it is. THE NEED The digital nature of the platform economy does, however, imply regulatory challenges. For instance, there is currently no systematic way to assess the food safety and integrity risks posed by online food businesses. Fortunately, the FSA has first-class risk assessment resources, and it has undertaken exploratory research about food in the THE platform economy. OPPORTUNITY This motivated this report to examine if existing FSA resources could be joined into a tool to assess the food safety and integrity risks* posed by online food businesses.
* Additional/different to those posed by food businesses not operating online (i.e. challenges related to the digital nature of activities by online food businesses as opposed to the physical preparation of food). Good regulation emanates from strong foundations. This project is grounded on a discussion of how complexity (complication + reflexivity + uncertainty) calls for… THE 1. An explicitly pluralist modular design. ✓ Deliverable made of modules that altogether add to a risk matrix AND can independently FOUNDATION contribute/learn to/from other efforts. 2. Risk assessment that helps both the regulator and online food businesses. ✓ Deliverable enables both assessment forms for use by inspectors or auditor and self-assessments for use by online food businesses. An IKEA-like risk ranking matrix: modular components that add into a full product.
… But way easier to put together than an IKEA desk, I promise.
That said, a TRL 3 product (aka. a proof of concept). MODULE OVERVIEW Typology Splits online food businesses into categories.
Risk ranking matrix List of Hazards Can enable (1) a ranked Lists challenges faced by online food businesses that could lead assessment of the risks posed to food safety/integrity incidents. by online food businesses (or types thereof) and (2) accompanying business guidance.
Risk maps Rationales to gauge risk in a way that is concomitant with how the FSA assesses risk elsewhere. GOODS (raw foods) SERVICES (finished foods) Groceries Events/experiences Takeaway/cooked meals Artisanal SMEs Industrial Artisanal SMEs Industrial Artisanal SMEs Industrial People, SMEs selling their own Online supermarkets People, households, or SMEs selling their own Large-scale catering People, SMEs selling their own cooked Large restaurant chains households, or ready-to-cook ingredient selling their own raw micro-businesses mobile food experiences service providers selling households, or meals online. selling their own cooked micro-businesses boxes online. E.g. Hello food. E.g. Ocado, Asda trading food online. E.g. La Paella their own services online. micro-businesses E.g. Sushi Express, 26 Grains. meals online. E.g. Domino's, trading raw food Fresh, Gousto. Online. experiences online. Paella, Cally Munchy. E.g. Eden Caterers, trading cooked McDonalds. online. E.g. Xia, E.g. Ximena, chef. Jaspers. meals online. E.g.
local producer. Xavier, baker. Vendor
Artisanal (blogs) Listings Social media Artisanal (blogs) Listings Social media Artisanal (blogs) Listings Social media
n/a. Online forums/listings of Social media platforms n/a. Online forums/listings of Social media platforms for n/a. Listings of takeaway, Social media platforms for actors trading raw food for marketing and food experience marketing and restaurant, and cooked meal marketing and (no booking). E.g. Food communication. E.g. providers (no booking). communication. E.g. operators (no booking). E.g. communication. E.g. Tidbit & Drink Surplus Facebook, Twitter. E.g. Vegan directory. Instagram. Three Best Rated, Foodism. Social. Informative Network.
Artisanal (affil.) Ordering & logistics Social marketplaces Artisanal (affil.) Ordering & logistics Social marketplaces Artisanal (affil.) Ordering & logistics Social marketplaces n/a. Platforms aggregating Town-square-like n/a. Platforms aggregating Town-square-like platforms n/a. Platforms aggregating cooked Town-square-like platforms raw food items from platforms enabling social food experiences/events enabling social interaction meals from external vendors enabling social interaction external vendors and interaction and (optional) from external vendors and (optional) and sometimes facilitating and (optional) intermediation sometimes facilitating intermediation by and sometimes intermediation by external delivery. E.g. Just Eat, by external vendors and delivery. E.g. Olio, external vendors and facilitating event vendors and potential Deliveroo, Too Good to Go, potential buyers. E.g. Farmdrop, Amazon potential buyers. E.g. management. E.g. buyers. E.g. Feast It Karma, BookATable. Facebook Marketplace. Fresh. Facebook Marketplace. Caterwings, EatWith, (Corporate). Intermediary CityPantry. Title (describes the types of business to which this module applies)
Hazard Self-assessment Diagnosis Impact* (potential, health) Impact* (potential, non-health)
TBD by experts/committees TBD by experts/committees
Title. This column This column presents the Safety impact. This column Integrity impact. This Explanation. presents the hazard hazard as questions for provides a dropdown menu column provides a dropdown as a question for external assessment, in a for experts (or committees) menu for experts (or self-reflection by the non-descript manner that to define the potential health committees) to define the business. can be adjusted to match impact that could derive potential non-health impact assessments at different from this hazard. that could derive from this levels. hazard.
* Values in red are exclusively to demonstrate file functionality. They have been selected arbitrarily and without even looking at the question next to them, as opposed to corresponding to any type of implicit or explicit guesstimate or assessment. Impact
Low Medium High
Mild illness: not usually life-threatening, usually no Moderate illness: incapacitating but not usually Severe illness: causing life-threatening or sequelae, normally of short duration, self-limiting life-threatening, sequelae rare, moderate substantial sequelae or illness of long symptoms (e.g. transient diarrhoea). duration (e.g. diarrhoea req. hospitalisation). duration (e.g. chronic hepatitis).
Very high 6 12 18 Events occur almost certainly.
High 5 10 15 Occurs very often (>99%).
Medium 4 8 12 Occurs regularly (10-99%).
Low 3 6 9
Rare but does occur (1-10%). Likelihood Very low Very rare but cannot be excluded 2 4 6 (<1%).
Negligible Extremely rare, does not merit 1 2 3 consideration.
Uncertainty
Low Medium High Solid and complete data available; strong evidence Some but no complete data available; Scarce or no data; evidence in unpublished in multiple references; authors report similar evidence in small number of references; reports, observations, or personal conclusions. authors’ conclusions vary. communication; authors’ conclusions vary.
1 2 3 VIDEO OF HOW IT ALL WORKS TOGETHER In its original form, this presentation includes a video of how the tool joins all modules, but the video is too large for email sharing. Please email me for any further details.
Self-assessment form Assessment form Self-assessment (vendors) Assessment form
CLICK HERE TO RESET FORM CLICK HERE TO RESET FORM
INTERMEDIARIES(S) INTERMEDIARIES(S) Step 1: Do you, your staff, or contracted undertake food delivery VENDOR(S) THAT DO VENDOR(S) THAT THAT DO NOT THAT DO NOT (irrespective of food being raw or cooked) and/or event/experience NOT DELIVER NOR DELIVER AND/OR FACILITATE DELIVERY FACILITATE DELIVERY NO YES MANAGE FOOD MANAGE FOOD NOR NOR management (incl. but not limited to supper clubs and catering EXPERIENCES EXPERIENCES EXPERIENCE/EVENT EXPERIENCE/EVENT services)? MANAGEMENT MANAGEMENT
Step 2: Are you an artisanal operator (e.g. home kitchens, micro local CLICK ONLY IF THE PLATFORM IS A ARTISANAL SME INDUSTRIAL producer, etc.), a small/medium business (e.g. takeaway, restaurant, SOCIAL MARKETPLACE catering, etc), or an industrial (e.g. restaurant chain, large catering ARTISANAL SME INDUSTRIAL service chain, online supermarket) operator? FORMAT (BUSINESS-LEVEL) CLICK HERE FOR FINAL FORMATTING FORMAT (TYPE-LEVEL)
Answer the questions. The coloured boxes will provide you a rough estimate of the likelihood of facing a Total risk score: 30 related food safety/integrity issue (which may harm others and lead to losses and penalties for you and/or your Likelihood business). Do note that this form is entirely for you to self-assess your activities and self-reflect on them. If you of a related incident Answer the questions. The coloured boxes will provide you a rough estimate of the likelihood of facing a answer incorrectly, you'd be lying to yourself. related food safety/integrity issue (which may harm others and lead to losses and penalties for you and/or your Composite Risk business). Do note that this form is entirely for you to self-assess your activities and self-reflect on them. If you Score . ANSWER HERE DO NOT MODIFY answer incorrectly, you'd be lying to yourself. Have you been operating online for many years? n/a ? . ANSWER HERE DO NOT MODIFY Are you a registered food business? n/a ? . n/a 0 Have you been operating online for many years? MINIMALLY (Barely) HIGH YES (or, 100% of this type of Do you ensure that the working environment is clean and suitable for REGULARLY (Yes, but 'things' Is this vendor a registered food business? vendors/platforms) 6 safe food production? happen) LOW MINIMALLY (or, < 1% of this type of Has this vendor operated online for many years? vendors/platforms) 20 Do you have a food safety management system (FSMS) in place to ALMOST ALWAYS (or, > 99% of this cover all of your food business activities? YES (Absolutely, no exceptions) LOWEST Does this vendor prioritise cleanliness? type of vendors/platforms) 4 Have you considered how online activities may call for adjusting your Does this vendor have an FSMS in place? n/a 0 FSMS? n/a ? Has this vendor considered how online activities may call for adjusting its FSMS? n/a 0 Have you received food safety training? n/a ? NEXT STEPS
Modules offer standalone opportunities 1. The typology already contributed to organising thinking by other FSA efforts. ✓ Similar future usage can help to avoid a loss of perspective by projects focusing on specific types of businesses or sectors. 2. Considering the list of hazards in the design of future FSA projects. ✓ May help to avoid accidental oversights. 3. Revision/validation of the typology and list of hazards. ✓ Enhance contributions of each (e.g. including additional food integrity concerns/hazards). 4. Engagement and consultations (as part of larger effort, not one solely about this). ✓ Increase the FSA’s understanding of online food businesses and build bridges with the sector 5. Generating and communicating self-assessment guidelines. ✓ Contribute to fostering common thinking across the platform economy. Ultimately, however, the report’s primary suggestion is to 1. Continue the development of the matrix. ✓ What can enable a systematic assessment of the food safety and integrity risks by online food businesses. Q&A
Dr Jose A Bolanos https://www.josebolanos.xyz
Contact me at [email protected] [email protected]
Or follow me on www.LinkedIn.com/in/jbolns. www.Instagram.com/jbolns. Session 1. Digitalisation of food (10:20-11:05)
10:20 Michelle Patel (Head of Social Science, FSA) The impact of Covid-19 on consumers and the food system in the UK
10:30 Lucy Hart (Analyst, Foundry4) Understanding online platforms and their impact on the food landscape
10:40 Dr Jose A Bolanos (Post-Doctoral Research Officer, FSA & London School of Economics) Food in the Platform Economy: Leapfrogging risk regulation in the context of the platform economy
10:50 Session 1 Panel Discussion After the break, at 1145, please rejoin the event using Audience link 2. Session 2. Innovative data collection: Online Survey methods (11:05-11:35)
11:05 Using a web-push methodology for FSAs flagship survey, Food and You Dr Patten Smith (Research Director, IPSOS Mori, and Visiting Professor at Surrey University)
11:15 US FDA Food Safety and Nutrition Survey - Moving from Phone to Address-based Sampling Amy M. Lando (Center for Food Safety and Applied Nutrition, U.S. Food and Drug Administration)
11:25 Session 2 panel discussion Using a web-push methodology for FSAs flagship survey ‘Food and You’ Patten Smith Ipsos MORI
ESRC FESTIVAL OF SOCIAL SCIENCE UNDERSTANDING FOOD IN A DIGITAL WORLD
9 November 2020 © 2016 Ipsos. All rights reserved. Contains Ipsos' Confidential and Proprietary information and may not be disclosed or reproduced without the prior written consent of Ipsos.
Document Name Here | Month 2016 | Version 1 | Public | Internal Use Only | Confidential | Strictly Confidential (DELETE CLASSIFICATION) 65 The Food and You survey series
▪ Food and You - FSA’s flagship consumer survey, which measures: – self-reported attitudes – food safety behaviour and knowledge – other food-related issues ▪ Until 2018/19 biennial face to face random probability survey covering general population of England, Wales and NI ; overall n=c. 3,000 ▪ 2018 review: moved survey to web-push methodology; overall n = 4,000 ▪ Ipsos MORI - just completed first wave fieldwork with the new methodology
Document Name Here | Month 2016 | Version 1 | Public | Internal Use Only | Confidential | Strictly Confidential (DELETE CLASSIFICATION) 66 What is web push?
▪ Web-push surveys have three main characteristics: 1. High quality surveys with random probability sampling 2. Predominantly online data collection methods 3. Offline contact / recruitment methods
Document Name Here | Month 2016 | Version 1 | Public | Internal Use Only | Confidential | Strictly Confidential (DELETE CLASSIFICATION) 67 Why web-push? Pressures to move to online data collection – Face-to-face survey declining response rates – Increasing face-to-face survey costs coupled with flatlining survey budgets – Reduced viability of RDD alternatives – response rates and mobile only – Govt. digital transformation – Increasing online population coverage – now c. 95% – Increasing public expectation of online contact – Lower cost of online contact – Methodological evidence that low response rates can be less damaging than feared
Document Name Here | Month 2016 | Version 1 | Public | Internal Use Only | Confidential | Strictly Confidential (DELETE CLASSIFICATION) 68 Why do we have to push people to web? ▪ Good random probability surveys require high coverage sample frames ▪ In UK, good sample frames lack email addresses, with consequence that… ▪ …we have to make contact via postal address and persuade (‘push’) respondents to respond online ▪ Thus was born the push-to-web survey defined as use of offline contact modes to encourage people to go online and complete a web questionnaire
Document Name Here | Month 2016 | Version 1 | Public | Internal Use Only | Confidential | Strictly Confidential (DELETE CLASSIFICATION) 69 The new Food and You Survey Summarised
▪ Ipsos MORI at forefront of UK web-push methods, and new Food and You incorporates our latest thinking ▪ PAF sample of addresses – England, Wales and N. Ireland ▪ Each address sent letter with URL and 2 logins to online questionnaire ▪ Ask any two adults (16+ ) in household to participate ▪ 3 reminder mailings ▪ Second reminder includes paper version of questionnaire ▪ £10 conditional incentive for participation
Document Name Here | Month 2016 | Version 1 | Public | Internal Use Only | Confidential | Strictly Confidential (DELETE CLASSIFICATION) 70 Invite letter
Document Name Here | Month 2016 | Version 1 | Public | Internal Use Only | Confidential | Strictly Confidential (DELETE CLASSIFICATION) 71 Why no random selection of individuals?
▪ In the 85% of UK households containing 1 or 2 adults, so if take 2 adults, no selection needed ▪ Why not random selection in remaining 15%? ▪ Any 2 adults in household – overall 94% of selections would have been selected using random procedures ▪ Because households often don’t comply, including random sampling instructions would only have increased correct selections from 94% to 96%! ▪ And ‘any 2’ method – should deliver higher response rate
Document Name Here | Month 2016 | Version 1 | Public | Internal Use Only | Confidential | Strictly Confidential (DELETE CLASSIFICATION) 72 Why a paper questionnaire?
▪ Non-coverage - not everyone is online (c. 5%) ▪ Non-response - not everyone is comfortable completing online questionnaires ▪ Older and poorer people less likely to be included in online than offline surveys ▪ Remedy – including paper questionnaire after initial efforts to push people to web improves sample profile
Document Name Here | Month 2016 | Version 1 | Public | Internal Use Only | Confidential | Strictly Confidential (DELETE CLASSIFICATION) 73 Where are we now?
▪ Completed first wave of new food and you - completing data processing - over 9,000 responses ▪ Higher response rate than usual for UK web-push surveys (30% v. c. 20%) – see next slide ▪ 63% questionnaires completed online ▪ Second wave of new Food and You launched later this week. ▪ Overall, a thoroughgoing success!
Document Name Here | Month 2016 | Version 1 | Public | Internal Use Only | Confidential | Strictly Confidential (DELETE CLASSIFICATION) 74 First wave repose figures
Country Issued Number of Number of Address level Number of addresses returns addresses response rate returns per overall taking part participating address
England 10,526 5,050 3,377 32.1% 1.50
Northern Ireland 5,263 2,054 1,375 26.1% 1.49
Wales 5,263 2,067 1,554 29.5% 1.33
Overall 21,053 9,171 6,306 30.0% 1.52
Document Name Here | Month 2016 | Version 1 | Public | Internal Use Only | Confidential | Strictly Confidential (DELETE CLASSIFICATION) 75 Thank you.
Document Name Here | Month 2016 | Version 1 | Public | Internal Use Only | Confidential | Strictly Confidential (DELETE CLASSIFICATION) 76 Session 2. Innovative data collection: Online Survey methods (11:05-11:35) 11:05 Using a web-push methodology for FSAs flagship survey, Food and You Dr Patten Smith (Research Director, IPSOS Mori, and Visiting Professor at Surrey University)
11:15 US FDA Food Safety and Nutrition Survey - Moving from Phone to Address-based Sampling Amy M. Lando (Center for Food Safety and Applied Nutrition, U.S. Food and Drug Administration)
11:25 Session 2 panel discussion FSANS FDA’s Food Safety and Nutrition Survey Moving from Phone to Address-based Sampling
Amy Lando, MPP Martine Ferguson, MS Fanfan Wu, PhD Linda Verrill, PhD
U.S FDA, CFSAN, Consumer Studies Branch
Presentation at ESRC Festival of Social Science (FOSS20) November 9, 2020 FSANS Project Team
Amy Lando, MPP Linda Verrill, PhD Fanfan Wu, PhD Martine Ferguson, MS
79 Center for Food Safety and Applied Nutrition U.S. Food and Drug Administration
• U.S. FDA is a scientific regulatory agency responsible for the safety of the nation's domestically produced and imported foods, cosmetics, drugs, biologics, medical devices, and radiological products.
• The Center for Food Safety and Applied Nutrition, one of FDA’s six product-oriented centers, is a national leader in protecting and promoting public health by safeguarding more than $1.5 trillion worth of food, cosmetics and dietary supplements.
80 The Value of FDA FSANS
Understanding Evaluating Monitoring
Understanding what consumers Developing and evaluating Monitoring trends in consumer know, believe and do consumer education materials behavior, knowledge, and attitudes
81 Transition to FSANS
Food Safety Survey (FSS), since 1988 Food Safety and Nutrition Survey Health and Diet Survey (FSANS) (HDS), since 1982
RDD ABS RDD Phone Web & Paper Comparison Surveys
82 2019 FSANS
• Change from an interview assisted phone survey (RDD) to self-administered web and paper survey (ABS multimode) – Increase response rate – Increase population coverage – Increase number of questions • Conduct a smaller phone survey to ensure mode continuity (RDD landline and cell phone) • Combine food safety and nutrition topics
83 Transition to FSANS
Food Safety and RDD Consumer Nutrition Survey Phone Surveys • Met with food safety and nutrition experts (FSANS) • Reviewed food safety and nutrition surveys Health and Diet • Drafted new survey questions for “hot” topics ABS Web & Paper Survey (HDS) • Updated question wording and skip patterns • Developed contact materials • Conducted cognitive interviews and pilot tests RDD comparison Food Safety Survey (Landline and cell (FSS) phone)
84 FSANS Questionnaires
ABS web version ABS paper and pencil RDD interviewer screen
85 Nutrition Facts Food package Organic GE foods label claims • General use • Low/lower • Purchase behavior • Knowledge Shopping • Serving size • Healthy • Beliefs related to: • Beliefs related to: • %DV • Natural – Harmful germs – Harmful germs • Perceptions • Organic – Environment – Environment – Long-term health – Long-term health – Nutrition – Nutrition
Food handling Raw foods Recipes and cooking ability • Hand washing • Following cooking instructions for • Use of recipes Preparing • Thermometer use frozen vegetables • Cooking frequency and ability FSANS Topics • Cutting board cleaning • Grilling meat and chicken • Raw meat and chicken
Restaurant Diet knowledge & Food Food Risky foods practice safety allergies
• Frequency / types • Overall diet • General risks • Prevalence • Consumption of raw Eating • Menu labeling • Label related • Who’s at risk and undercooked • Health inspection • Dietary fiber • Knowledge foods – meat, fish, scores • Calories vegetables, eggs, raw • Consumer advisory • Salt & sodium milk, raw flour
Proper storage of refrigerated foods
Storing • Meat and chicken dishes
86 Methodology ABS Multimode RDD N=4,398 N=834 Gen. pop. U.S. adults 18+ Gen. pop. U.S. adults 18+ Fielded October 1-November 20, 2019 Fielded October 14-December 22, 2019 USPS CDSF of households 80% cell / 20% landline • Web push with • Oversamples AA/Hispanic paper NR follow-up, high density (LL only) up to 5 contacts • Up to 2 contacts for LL only
English/Spanish English/Spanish $2 advance token appreciation No token appreciation
Response Rate = 28.1% (AAPOR RR3) Response rate = 6.6% (AAPOR RR3) 87 Next Steps
Continue data analysis Presentation and reports Publications
14 Questions & Thank you!
Please contact
Amy Lando: [email protected]
89 Session 2. Innovative data collection: Online Survey methods (11:05-11:35)
11:05 Using a web-push methodology for FSAs flagship survey, Food and You Dr Patten Smith (Research Director, IPSOS Mori, and Visiting Professor at Surrey University)
11:15 US FDA Food Safety and Nutrition Survey - Moving from Phone to Address-based Sampling Amy M. Lando (Center for Food Safety and Applied Nutrition, U.S. Food and Drug Administration)
11:25 Session 2 panel discussion Session 3. Innovative data collection: Other methods (11:45-12:30) 11:45 Participatory approaches in developing new risk assessment and communication tools Domagoj Vrbos and Giorgia Zamariola (Risk Communication, European Food Safety Authority)
11:55 Qualitative Handwashing research: An ethnographic app study Helen Heard and Katharine Porter (Social Science, FSA)
12:05 Food and Citizen Science Dr Christian Reynolds (Senior Lecturer, Centre for Food Policy, City University, and visiting researcher, at the Department of Geography, University of Sheffield)
12:15 Session 3 Panel Discussion 9 November 2020
Participatory approaches in developing new risk assessment and communication tools
Engaging grassroot stakeholders through focus groups
Domagoj Vrbos & Giorgia Zamariola Communication Engagement and Cooperation Department Discussion I Bee health
9 3 Beekeepers focus groups I Countries
Results to be published in 2021
Southern landscape
Central landscape
Northern landscape
9 4 Focus groups I Research “packages”
Package 1 Results to be published in 2021 Regulatory system
Package 2 Data for bees Package 3 Digital beehives
Package 4 Research for bees
9 5 As we speak I Transition to online discussions
9 6 Subscribe to www.efsa.europa.eu/en/news/newsletters www.efsa.europa.eu/en/rss
Engage with careers www.efsa.europa.eu/en/engage/careers
Follow us on Twitter @efsa_eu @plants_efsa @animals_efsa @methods_efsa
9 7 Session 3. Innovative data collection: Other methods (11:45-12:30)
11:45 Participatory approaches in developing new risk assessment and communication tools Domagoj Vrbos and Giorgia Zamariola (Risk Communication, European Food Safety Authority)
11:55 Qualitative Handwashing research: An ethnographic app study Helen Heard and Katharine Porter (Social Science, FSA)
12:05 Food and Citizen Science Dr Christian Reynolds (Senior Lecturer, Centre for Food Policy, City University, and visiting researcher, at the Department of Geography, University of Sheffield)
12:15 Session 3 Panel Discussion Qualitative Handwashing research: An ethnographic app study
09/10/2020 Helen Heard and Katharine Porter Social Science Team Overview - Qualitative Handwashing app study
Project aims A project in two phases
To qualitatively investigate how or Phase 1: why people wash their hands the way A two-week study, asking they do and the factors that influence 12 participants to record their behaviour. their handwashing behaviours Our main research questions are: and answering questions in real-time 1. What do people understand constitutes a ‘good’ hand wash? 2. What factors influence decision- making about when and how to wash Phase 2: hands? Does this vary in different Interviews conducted in-house with scenarios? up to 20 participants to 3. What barriers and levers influence explore handwashing effective handwashing inside and behaviours (conducted via Microsoft outside of the home.? Teams) 4. How has COVID-19 affected consumer handwashing behaviour, Publication in 2021. Initial findings and how do consumers explain this?’ from Phase 1 available now Why an ethnographic app study?
• To minimise social desirability bias • Distances the presence of the researcher • More time to narrow down onto potentially embarrassing subjects
• To minimise recall bias • Interruptive questioning captures very recent behaviours
• To place knowledge and reported behaviour into a wider context • Allows participants to explain or account for their actions How did it work?
Examples of images shared by participants Examples of the App Questions: Some findings so far…
“A good hand-wash is one using hot water, plenty of soap making sure you clean every part including 1. What do people understand backs of hands, nails and in- constitutes a ‘good’ handwash? between the fingers.”
– 54, F, multi-person household, living with dependent children
“When I am quite confident that my dog has been washed, I’m comfortable 2. What factors influence decision- touching and cuddling my dog without me washing after. If the dog has just making about when and how to been for a crazy long walk … loads of wash hands? Does this vary in people have been touching her, I different scenarios? wouldn’t touch her until she was washed.”
– 25, M, multi-person household Some findings so far…
"Using the gloves and the gels, your hands get really sore - you're constantly 3. What barriers and levers influence handwashing." effective handwashing inside and – 36, F, living with dependent children, care outside of the home? industry
“COVID-19 has changed how I clean my 4. How has COVID-19 affected hands. It has made me so aware that I am consumer handwashing behaviour, touching things all the time. It has made and how do consumers explain this?’ me clean my hands over triple my norm.”
– 54, F, multi-person household, living with dependent children Reflections/Conclusions
• Allows for collection of detailed behaviours, including the thinking behind them • Structure of the methodology (warm-up interview, app diaries, wrap-up interviews) allows for detailed probing after behaviours were completed • No technical issues – as a result of using an established app through a contractor • Engagement was high, and well maintained. Incentives were offered (£125) for completion of project • Once app was in progress, the FSA team could not be involved in immediate response/probing, until wrap-up interview stage
Full detailed report on both phases of the project will be published in 2021, via the FSA website THANK-YOU
[email protected] [email protected] Session 3. Innovative data collection: Other methods (11:45-12:30)
11:45 Participatory approaches in developing new risk assessment and communication tools Domagoj Vrbos and Giorgia Zamariola (Risk Communication, European Food Safety Authority)
11:55 Qualitative Handwashing research: An ethnographic app study Helen Heard and Katharine Porter (Social Science, FSA)
12:05 Food and Citizen Science Dr Christian Reynolds (Senior Lecturer, Centre for Food Policy, City University, and visiting researcher, at the Department of Geography, University of Sheffield)
12:15 Session 3 Panel Discussion Food and Citizen Science
ESRC Festival of Social Science: FSA Event in partnership with the University of Sheffield Understanding food in a digital world: 9th Nov 2020
Dr Christian Reynolds Centre for Food Policy, City, University of London @sartorialfoodie What will we be talking about today?
• Citizen Science – What is it?
• Food and Citizen Science
• How I got into Citizen Science
• Where have I used Citizen Science
• Online experiments, Gamification and Living Labs Citizen Science – what is it?
• Research where public citizens participate as investigators in research projects alongside professional scientists. • Citizen science engages a diversity of publics (Sauermann et al., 2020) enabled by scientists who are advocates (Stilgoe, 2009) is a powerful collaborative approach. • This is not only about engaging widely, but also about bringing fresh perspectives, solutions and enable a faster and smoother uptake.
For more information please read: Robinson L.D., Cawthray, J.L., West, S.E., Bonn, A., & Ansine, J. (2018). Ten principles of citizen science. In S. Hecker, M. Haklay, A. Bowser, Z. Makuch, J. Vogel, & A. Bonn. Citizen Science: Innovation in Open Science, Society and Policy. London, UCL Press. 1–23. The ‘escalator’ model of science engagement
Engagement level Real world example
Source - Doing It Together Science – an EU citizen science project
Food and Citizen Science
Citizen science engagement with the food system and impact pathways
Oakden et al (forthcoming/under review) Citizen engagement in science and the food system is essential to enable a shift to sustainable healthy dietary practices. Frontiers Ecological monitoring
• School children scientists in a project to monitor soil health
• Mussel pathogen survey to monitor local toxins, collected samples and also carried out sample analysis, allowing survey sites number to rise from 60 to 108.
• Local Environmental Observer (LEO) is a community situated network that started in Alaska. This group has tested for food pathogens, and monitored for advanced permafrost melting, to safeguard household and community food storage depots Urban Growing
• Citizens generated data on the suitability of growing conditions in different geographies for up to 10 lines of soya bean plants, aiming to reduce reliance on imported soya • MY Harvest initiative is investigating the geography of urban food growing, examining the scale of urban cultivation and yields in Leicester and across the UK. Artisanal/ Specialist Food Processing & Production
• Peer networks and by experts (e.g. brewers or bakers) • Commercial food substitute Soylent • ‘biohackers’ such as Real Vegan Cheese biohacking group (Wilbanks, 2017) or the Shojin Meat project. Artisanal/ Specialist Food Processing & Production Food Safety & Fraud
• Allergen testing, food spoilage etc. See https://theanalyticalscientist.com/fields- applications/citizen-science-and-food-safety • ‘Citizen Radioactivity Monitoring stations’ – Japan post Fukushima Community Based Public Health
• Onevoice or Photovoice methodology (using smartphones) • Examine enablers and barriers to health and food practice • Citizen scientists in New Jersey (USA) identified problems with a Healthy corner stores’ scheme and suggested strategies for stores’ to modifying logistics together, improving scheme implementation (Chrisinger et al., 2018). • Sharing recipes and meeting to cook healthier foods together (Rogers et al., 2018). These were just some of the aspects…
Citizen science engagement with the food system and impact pathways
Oakden et al (forthcoming/under review) Citizen engagement in science and the food system is essential to enable a shift to sustainable healthy dietary practices. Frontiers How I got into Citizen Science
2012-2014 2018-2020
Dr Philip Roetman from the University of South Australia demonstrates the Koala Counting app Online experiments
Zooniverse - You Know Food? (2019) Online experiments Zooniverse - You Know Food? https://doi.org/10.3389/fsufs.2020.00120
N=~516, 8484 valid image classifications. 10 Foods types 3 portion sizes, with and without weights. Results: Citizens are unable to accurately estimate carbon footprint and energy content, with the majority of citizens overestimating values. Portion size impacts perceptions, with estimations increasing alongside size. Weight information influences perception, but the direction varies by factor. Input method significantly affects citizen estimations. Citizen feedback confirms the lack of knowledge surrounding carbon footprint values. Online experiments Food Safety: Controls
Chicken is rated as a higher risk food than pasta and apple.
https://doi.org/10.35241/emeraldopenres.13711.1 Women rate chicken as higher risk than men or other, however there are no gender difference for pasta or apple.
Significant diff. vs control
Low Risk High Risk
0 1 2 3 4 5 6 7 8 9 10 Online experiments Chicken: Safety
Vegans, vegetarians, pescatarians, those with another dietary preference https://doi.org/10.35241/emeraldopenres.13711.1 rate risk as significantly higher than omnivores.
Low Risk High Risk
0 1 2 3 4 5 6 7 8 9 10 Gamification
http://climatefoodchallenge.online/
https://drive.google.com/file/d/1FxitzwFFnsloH2kc3LnqlAYfHQmoYyib/view Living Labs - University of Sheffield
Actions • Low impact logo on menu • Milk guides • Surveys and Interviews with students • Student ambassadors • “Climate Strike” menu changes Take Away thoughts
• Technology is enabling Citizen Science research in new and exciting ways. • There are many ways that Citizen Science can help engage, and educate, advance and evaluate existing types of (food) research. • Let’s get to it! Thanks to my (many) collaborators…
Citizen Science; climate change, cooking and food habits: Carla Adriano Martins; Marcelo Vega; Ian Vázquez Rowe; Gustavo Cediel ; Ximena Schmidt; Angelina Frankowski ; Sarah Bridle ; Carolyn Auma ; Jacqueline Silva ; Gemma Bridge ; Libby Oakden; Hibbah Osei-Kwasi ; Alana Kluczkovski ; Robert Akparibo; Tahir Bockarie; Daniel Mensah; Maria Laura Louzada; Changqiong Wang ; Luca Panzone ; Astrid Kause ; Charles Ffoulkes; Coleman Krawczyk ; Grant Miller; Stephen Serjeant; Fernanda Rauber; Renata Levy, and the catering team and students at University of Sheffield. … and funders. This work was funded by STFC Food Network+ pilot funding (ST/P003079/1), and STFC 21st Century challenge funding (ST/T001410/1) “Piloting Zooniverse for food, health and sustainability citizen science”. Christian Reynolds was supported from the HEFCE Catalyst- funded N8 AgriFood Resilience Programme and matched funding from the N8 group of Universities. Additional funding was provided by Research England via the project “Food based citizen science in UK as a policy tool”. Thanks to Sheffield Sustainable Food Futures group and the Institute for Sustainable Food, University of Sheffield for providing funding to extend the Living Labs project. Comments? Questions?
Dr Christian Reynolds Centre for Food Policy, City, University of London @sartorialfoodie [email protected] Financial Support: This work was funded by STFC Food Network+ pilot funding (ST/P003079/1), and STFC 21st Century challenge funding (ST/T001410/1) “Piloting Zooniverse for food, health and sustainability citizen science”. Christian Reynolds was supported from the HEFCE Catalyst-funded N8 AgriFood Resilience Programme and matched funding from the N8 group of Universities. Additional funding was provided by Research England via the project “Food based citizen science in UK as a policy tool”. Thanks to Sheffield Sustainable Food Futures group and the Institute for Sustainable Food, University of Sheffield for providing funding to extend the Living Labs project. Place-based citizen science
• “Health and Local communities” project in Bornholm, Denmark was community led public health intervention, in a defined geography that aimed to reduce childhood overweight and obesity (Bloch et al., 2014; Toft et al., 2018) • Sustainable food places? Food Literacy & Culture
• Canadian healthy eating project (Growing roots): Immigrant communities, to gain familiarity with cooking healthy Canadian meals and to explore positive nutritional elements in their indigenous food cultures. (Henderson and Slater, 2019) Session 3. Innovative data collection: Other methods (11:45-12:30)
11:45 Participatory approaches in developing new risk assessment and communication tools Domagoj Vrbos and Giorgia Zamariola (Risk Communication, European Food Safety Authority)
11:55 Qualitative Handwashing research: An ethnographic app study Helen Heard and Katharine Porter (Social Science, FSA)
12:05 Food and Citizen Science Dr Christian Reynolds (Senior Lecturer, Centre for Food Policy, City University, and visiting researcher, at the Department of Geography, University of Sheffield)
12:15 Session 3 Panel Discussion After lunch, at 1310, please rejoin the event using Audience link 3. Session 4. The potential of social media analysis (13:10-13:55)
13:10 Using social data to surface food trends and understand audiences Anna Rudkevych (Senior Research Manager, Pulsar)
13:20 FSA Covid-19 Social Media Listening Beatrix Siemering (Strategic Surveillance, FSA)
13:30 Understanding plant-based food images beyond the wellness hype Alexandra Boutopoulou (PhD candidate, University of Sheffield)
13:40 Session 4 Panel Discussion Pulsar x FSA Using social data to surface food trends and understand audiences
Anna Rudkevych | Senior Research Manager 138 How this project came to be
The FSA team engaged Pulsar to conduct a social listening food trends study.
We have devised a multi-step framework to go from large-scale trend discovery to trend deep-dives.We used social media data to conduct the study: Twitter data for trend discovery and volumetric analysis, and Forum, Twitter and Reviews data for trend deep-dives.
We started with broad keyword syntax to capture the wider food conversation, proceeding to develop bespoke keyword search syntax and keyword exclusions for every trend uncovered to ensure we’re capturing relevant results. For this study, we combined three unique approaches to data analysis to arrive at the big picture Food image recognition AI
Pulsar is the only platform that lets you choose which algorithms to deploy to Audience Segmentation & Social Listening analyse a specific dataset. Our 2 image integrated recognition modules specifically dedicated to food enriched our trends identification Different communities within an audience and insight deep dive processes. talk about the same topic very differently. There’s a lot more value in the data than generic listening can deliver. To unlock the full potential of the data, we segmented key audiences of interest using the food trends community lens to glean deeper insights from the food category conversation.
Proprietary trend virality framework
Pulsar’s research team have developed a framework for effectively identifying and examining trends. This helped us structure the broad and complex food trends dataset, providing insights into the shape and evolution of individual food trends as as well as across categories.
140 How we’ve approached this: analysis steps
We collected an initial broad pool of circa 4 millions tweets mentioning keywords such as cooking, eating, ingredients in March- August 2019. We proceeded to use a number of techniques, including keyword visualisation and filtering, top Trend discovery content, and top image tags, to identify prominent topics that could be considered to be trends.
Trend validation We used Pulsar TRENDS to compare the number of tweets about each trend in the August 2018- 2019 vs 2019-2020. We illuminated some of the trends, and showed ~200 noteworthy trends in our report.
Selection and Virality Modelling Following discussions with the FSA team, we agreed on ~30 trends to be explored further. We interrogated the shape of discussion for these trends using Pulsar’s virality metrics.
We collected Twitter, Forum and Review data for these qualified trends going back a whole year. Deep dives We then deep dived into how each of these trends is being discussed and what are the key driving forces behind the conversation..
Audience Mapping We have profiled audiences of people discussing Allergies and Hypersensitivities and a combined audience of people making up Food standards post-Brexit macro-trend discussion
Pattern identification We have included analysis of discussion patterns throughout the study Introduction to trend classification: categories, higher level vs specific trend and theme classifications
For this research, we have considered a large number of topics which relate to food consumption, with conversations varying in size and nature. While each topic area is referred to as a trend, there are varying levels of positivity expressed towards an issue at hand, with some trends seeing a high level of mentions due to people’s negative sentiment towards a phenomenon, e.g. chlorinated chicken.
In order to put this into context, we’ve employed the following ways of classifying trends:
Category topics Higher level Theme classification vs specific trends We have divided trends into the following The growth rate of some of the trends present has clearly been categories: driven by external factors. Analysis in this report is guided, aside 1. Food systems from discussion volume, by growth To contextualise this, we have introduced categories of: 2. Ingredients and process rate and trends’ belonging to a wider Brexit- topics directly influenced by existing or speculated outcomes 3. Health macro-trend category. As such, you of Brexit 4. Intolerances may see trends of a larger size COVID-19- topics which either emerged, or gained significant 5. Diets, Habits and Routines alongside emerging phenomena. prominence brought forth as result of COVID-19 6. Food Safety Public Concern- issues that are recognised to affect large population 7. Sustainability We have divided all trends into Higher- groups and are debated among public level and Specific topics, to signify the Everyday Life- ”business as usual” topics relating to individual These categories were created by the Pulsar team trends that combine larger topic areas, concerns, routines and choices as a result of the food conversation interpretations. and specific medium-sized and macro A number of trends may belong in multiple trends. The distinction is not always clear-cut. Some of the trends discussed categories, depending on the lens applied. In such existed as topics long prior to Brexit or COVID-19, but have gained cases, our team used our discretion to assign a traction as a result of the changes triggered by both factors. Thus, trend to a category. we have applied discretion in this area. Introduction to trend classification: macro-trends and how to read them
In order to help further contextualise key changes which have taken place within each category, we’ve grouped identified Macro trend visualisation examples trends into a smaller number of macro-trends.
We have visualised these using bubbles of different size, reflective of macro-trends’ overall discussion volume on Twitter in 2020 (a sum of mentions contributed by trends that make up Stable or Growing macro- Declining macro- the corresponding macro-trend). inconclusive macro- trend trend trend We have also signified macro-trends’ growth in 2020 using green-coloured bubbles, and decline in 2020 using red- coloured bubbles, both in comparison to 2019 discussion volume.
We used a grey bubble where a macro-trend have remained stable. We have also assigned grey to macro-trends where one definitive direction of travel could not be identified, as individual trends making up that macro-trend showed conflicting direction of travel. Growing individual trend Some trends fell within the Other category, as opposed to being Declining individual trend assigned to macro-trends. These are presented individually, with white and red arrows showing upward or downward trend. The shape of food discussions on social media in 2020 We considered annual growth to understand how the landscape has changed Example: ingredients and process category overview
Ingredients and process category reflects a non-exhaustive list of specific ingredients, dishes and cooking processes which surfaced during the exploration of the broad food, cooking and ingredients dataset.
Category macro-trends Other trends (standalone)
Cooking methods
Cooking during (134,761 overall mentions) Haggis (417,523) COViD-19 Faux meat (18,430)
(207,623 overall mentions) Seasonal food (15,194) Recently Rapeseed oil (3,871) popularised food Wellness focus additions (10,497 mentions) (109,224 overall mentions)
Click to add footnotes 145 Year-on-year Trend (ranked by volume) Brief description Volume 2019 Volume 2020 Topic level Theme Growth Cooking during COVID-19
Sourdough All mentions 55498 99414 79% Specific Everyday life Substitutes All mentions mentions alongside food and recipe terms 19402 29995 55% Higher-level COVID-19 Canned food All mentions 54389 78214 44% Specific Everyday life Cooking methods
Slow cooking All mentions mentions alongside food and recipe terms 49208 61253 24% Specific Everyday life Stir fry All mentions 22470 26316 17% Specific Everyday life Air fry All mentions 12491 13424 7% Specific Everyday life Pizza oven All mentions 9337 10100 8% Specific Everyday life Marinade All mentions mentions alongside food and recipe terms 4376 6372 46% Specific Everyday life Instant pot All mentions 8004 6114 -24% Specific Everyday life Griddle All mentions mentions alongside food and recipe terms 3535 4265 21% Specific Everyday life Charcoal All mentions mentions alongside food and recipe terms 3895 3958 2% Specific Everyday life Campfire food All mentions 4817 2959 -39% Specific Everyday life Wellness focus
Matcha All mentions 30673 27939 -9% Specific Everyday life Coconut oil All mentions 31190 27580 -12% Specific Everyday life Superfoods All mentions 34615 23610 -32% Specific Everyday life Kombucha All mentions 18569 15501 -17% Specific Everyday life Clean Eating All mentions 19194 11858 -38% Specific Everyday life Cauliflower as steak/pizza Mentions of cauliflower as substitute for steak and pizza base 3286 2736 -17% Specific Everyday life Recently popularised solutions
Nut butter (not peanut) All mentions 8149 7447 -9% Specific Everyday life Alternative fruit syrups All mentions 2532 2532 0% Specific Everyday life CBD and cooking All mentions 779 518 -34% Specific Everyday life Other
Haggis All mentions 339243 417523 23% Specific Everyday life
Source: Twitter, 28 August 2018/19, 2019/2020, volumetric preview data by Pulsar TRENDS We then went beyond growth rate, applying virality metrics to deep-dive into the shape of key conversation
Metrics glossary Growth Rate - % change in total volume of mentions (from pervious period to current period); a key metric used here to understand the difference between pre- and post- COVID-19 lockdown discussions
Lifespan, Velocity and Volatility Peak Velocity - peak amount of volume per hour (max daily value from the period divided by 24) used to ascertain the speed of the metrics slide were all taken into discussion spreading. Discussion with high initial saturation rate the account in combination to plot often have a shorter lifespan trends on Streaky vs Steady scale. Volatility - standard deviation of daily rate of change in volume used to ascertain the stability of the conversation, with high volatility rate meaning the level of discussion fluctuates often
Lifespan - Consecutive days above median volume during the time period measured used to gauge the level of a trend’s sustainability over time
147 Example: the trajectory of high-volume social media trends in Marc- August
HEATING UP
Food & going out worries Food standards
Chlorinated chicken
Farm work STREAKY STEADY
Conscious consuming/ Plant-based green diets
COOLING DOWN Source: Twitter, 28 August 2019/2020, March -August 2020, volumetric preview data by Pulsar TRENDS 148 Overall shape of trending conversations: COVID and ongoing Brexit conversations contribute to the high level of volatility seen across most discussions
Overall volume of tweets referencing COVID and Brexit in the last year (UK)
Most trends assessed appear streaky regardless of the overall discussion volume, seeing a high level of volatility, as Brexit and COVID-19 have a major influence on reshaping all areas of people’s lives. COVID-19 The majority of trends seeing high recent growth are largely driven by Brexit and Coronavirus, suggesting public concerns ~44million Brexit overshadowing private choices. ~18 million A degree of volatility prevails for Everyday Living topics, such as chickpea flour, faux meat and skipping meals, emphasizing the ripple effects of these two major drivers.
149 From broad discussion to audience communities: Allergies & Hypersensitivities case study Intolerances and hyper-sensitivities category overview
Intolerances and hyper-sensitivities contains all mentions of allergies and intolerances related to food in general, as well as specific ingredients. Allergy- triggered behaviours and lifestyle impact (e.g. avoiding restaurants and takeaways) are also included.
Category macro-trends
Allergies & Intolerances –general (66,673 mentions)
Allergy to specific ingredients Allergies & eating (85,425 overall mentions) prepared food (6,438 mentions) Intolerances and hyper-sensitivities : overall mentions are down, however mentions of reactions to spice and chocolate is on the rise
Allergies and Intolerances conversation is generally declining, including lower levels of fish and dairy allergy discussions. However, discussion around spice and specifically cinnamon intolerance is on the rise.
Trend (ranked by annual Year-on-year Brief description Volume 2019 Volume 2020 Topic level Theme growth level) Growth Allergy to specific ingredients
Allergy - dairy Mentions of allergies as well as “can’t have” alongside dairy 47902 42746 -11% Specific Everyday life Allergy - nuts Mentions of allergies allergies as well as “can’t have” alongside nuts 25677 25939 1% Specific Everyday life Allergy - chocolate Mentions of allergies allergies as well as “can’t have” alongside chocolate 3590 5691 59% Specific Everyday life Allergy - seafood Mentions of allergies allergies as well as “can’t have” alongside seafood 4344 3237 -25% Specific Everyday life Allergy - fish Mentions of allergies allergies as well as “can’t have” alongside fish 2730 2450 -10% Specific Everyday life Allergy - spicy Mentions of allergies allergies as well as “can’t have” alongside spices 1329 1895 43% Specific Everyday life Allergy - oranges Mentions of allergies allergies as well as “can’t have” alongside oranges 1836 1721 -6% Specific Everyday life Allergy - ingredients Mentions of being allergic to specific “ingredients” in a dish 2622 1514 -42% Specific Everyday life Allergy - cinnamon Mentions of allergies allergies as well as “can’t have” alongside cinnamon 196 232 18% Specific Everyday life Allergies & Intolerances (general) Allergies, allergens, intolerances, All mentions of allergies, intolerances and reactions to food 76965 66673 Higher-level Everyday life allergic (general) -13% Allergies & eating prepared food
Allergies & Takeaway Mentions of allergies and takeaway 873 747 -14% Specific Public concern Allergies & Eating out Mentions of allergies and cafes, restaurants, eating out 6744 5691 -16% Specific Public concern
Source: Twitter, 28 August 2018/19, 2019/2020, volumetric preview data by Pulsar TRENDS Allergies and intolerances Deep Dive
Allergies and intolerances discussion is multi-faceted and involves consumers, food business owners and employees. Peaks in Key food allergy topics conversation are driven by news stories about people who suffered from improper ingredient labelling, contributing to labelling becoming the most prominent discussion topic, alongside awareness. Within the awareness conversation, there is a Labelling 2521 concern that not all food businesses recognise the seriousness of improper labelling. On the other hand, we see occasional posts from hospitality workers suggesting customers sometimes exaggerate their allergy concerns, making hospitality Awareness 2341 worker’ jobs more difficult. Affordability 1112 Affordability emerges as another prominent theme, as awareness is raised about the difficulty of sticking to the proper diet Training for allergy sufferers on lower incomes. Access is another issue discussed, with posts urging people not to buy special diet 999 foods for no reason. Access 958 With allergy awareness raising, there is some underlying worry from people undiagnosed as having allergies about potential FSA 248 harms of their diets, and commercial food allergy tests are sometimes mentioned. The latter are sometimes deemed controversial.
3500 A widely-shared tweet urging Discussion of Viral tweet about difficulties people to not panic-buy free 3000 Owen Carey’s of living with a garlic allergy from products simply because death following they are the only ones left 2500 him unknowingly available, leaving those who eating buttermilk Another post about really needs them with no King’s college research A story about difficulties of 2000 in chicken garlic allergy options suggesting early exposure Joke about lactose having a lactose allergy and 1500 in infants can prevent nut intolerant people being coeliac whilst on allergies later in life indulging in diary benefits sees some shares 1000
500
0 Aug-20 Sep-20 Oct-20 Nov-20 Dec-20 Jan-21 Feb-21 Mar-21 Apr-21 May-21 Jun-21 Jul-21
Source: Twitter, 28 August 2019/2020, for volumetric data; forums and reviews incorporated into deep-dive analysis 153 Food Allergies Audience comprises of varied and distinct segments We re-examined the allergies conversation through the communities lens
Define Segment Cluster Discover the audience the audience the conversation unique engagement patterns
Attitudes Language Topics Behaviours Occasions Mindsets Example: vegans and animal lovers discuss allergy risks relating them back to their cause, while mummy bloggers focus on meeting family needs
Older, more female Similar age, compared to more full audience female compared to full audience
Mummy Bloggers & Nutritionists – 7% Vegans & Animal Lovers – 4% Audience comprised of bloggers and writers interested in vs People following a vegan lifestyle, many of whom for nutrition, healthy lifestyles and parenting. They seek and reasons related to animal rights. They highlight the risk of give advice/recipes on catering to family members with allergy to dairy products to discourage consumption. Some specific food intolerances (e.g. nuts) and dietary discuss own allergies to plant-based replacements (e.g. requirements (e.g. vegan). soya) and tweaks to vegan recipes to avoid reactions to gluten. Online Affinities (People & Organizations): Online Affinities (People & Organizations):
Jo Middleton Victoria Sully Sarah Anguish Honest Mum Jaime Oliver Peter Egan John Oberg Vegan Olive Veganella Philip Morris
Your Healthy The Vegan Plant Based Brit Mums MumsNet Sainsbury’s Netmums Herbiⓥore Veganuary Vegan Future Living Society News Thank you! Session 4. The potential of social media analysis (13:10-13:55)
13:10 Using social data to surface food trends and understand audiences Anna Rudkevych (Senior Research Manager, Pulsar)
13:20 FSA Covid-19 Social Media Listening Beatrix Siemering (Strategic Surveillance, FSA)
13:30 Understanding plant-based food images beyond the wellness hype Alexandra Boutopoulou (PhD candidate, University of Sheffield)
13:40 Session 4 Panel Discussion FSA Covid-19 Social Media Listening
09/11/2020 Why listen to social media?
Just been to local Chinese take Waiting for the storm to hit is away. They are really suffering as painful. Knowing what is they have less people ordering. coming feels unreal. People actually think they are at Watching the general public risk from a chop suey...Support hoard food, ignore advice your local Chinese take away!!! and continue to spread this (27.02.2020) virus, knowing that in just weeks we will be seeing mass fatalities, its infuriating, depressing and, frankly, enraging. (19.03.2020)
That period in lockdown when there was no food in the supermarkets & people were fighting for tissue doesn't even seem real (28.06.2020) But…
Twitter ≠ Real World What to listen for?
“Deep “Corona” Clean” “Virus” “Spread “Covid” by” “Infected”
“Food” “Lockdown” “New “Quarantine” Normal”
“Pandemic” “Self “Rona” Isolation”
“2020” ? Topics
Restaurant: Mentions of restaurants, pubs and cafes peaked in response to lockdown announcement Food insecurity: The key themes are food banks and free school meals Delivery: A lot of discussion on social media is about food delivery availability with some mentioning of most vulnerable at-risk people. Supply: People on social media voice concern or anger over stockpiling/ hoarding, with some concerns over food shortages and availability/ safety of workers Local: There is an increased focus on local communities and buying local, which might reflect changes in shopping behaviour. Trends Unsafe behaviours Example: Are people worried about catching covid from food and are they using unsafe methods to clean food? What have we learned?
• Key themes – What are people talking about • Insight into themes – How are people talking about these topics • Insight into specific questions, e.g. food hygiene • Reaction to specific events, e.g. changes in restrictions Thanks!
Beatrix Siemering [email protected] [email protected] Session 4. The potential of social media analysis (13:10-13:55)
13:10 Using social data to surface food trends and understand audiences Anna Rudkevych (Senior Research Manager, Pulsar)
13:20 FSA Covid-19 Social Media Listening Beatrix Siemering (Strategic Surveillance, FSA)
13:30 Understanding plant-based food images beyond the wellness hype Alexandra Boutopoulou (PhD candidate, University of Sheffield)
13:40 Session 4 Panel Discussion Understanding plant-based food on Instagram: The case of Deliciously Ella
Alexandra Boutopoulou PhD Candidate, Information School, University of Sheffield
Twitter: @alexgrwords About me u PhD research at the Information School, University of Sheffield, working on big data, visual social media and food cultures
Short bio
A Media and Advertising professional, with 20 years working experience, specializing in media planning & buying, as well as digital and social media strategy
-MSc Data Science – University of Sheffield -MA Communication – University of Westminster -BA English Language and Literature – University of Athens My research
u Research drawing upon the increasingly popular practices of a set of highly active online actors self-defined as ‘wellness’ bloggers. u These actors, many of whom have popular blogs and social media profiles, generate a wealth of pictures of their shopping, cooking preparations and outcomes and post them online to communicate with their audiences. Project Aims
u To chart the ways in which (visual) social media have shaped recent social phenomena of healthy eating, wellness discourses and lifestyle practices in the UK u To provide evidence, information and data-based insights that could be further used by FSA to understand and eventually influence consumer behaviours in relation to healthy eating u To achieve its aim, the thesis will focus on the analysis of the visual social media content generated by a particular actor, the influential food blogger Deliciously Ella on Instagram. Original Approach u The originality of the approach consists in the focus on distinctively visual analysis, which has not been conducted before on this topic (to the best of my knowledge) u Furthermore, due to Instagram’s API limitations (and especially compared to the more popular Twitter’s API), Instagram visual data analysis is still scarce Theoretical Framework
➢Self presentation and Authenticity (Goffman, 1959; Marwick, 2010; Jackson, 2013; Johnston and Bauman, 2015)
➢ Distinction and Luxury (Bourdieu, 1979; Friedman, Savage, Hanquinet and Miles, 2015)
➢ Enrichment (Boltanski and Esquerre, 2017) Methods and Data Collection
➢ A combination of qualitative and computational methods, in order to analyse a large dataset of Instagram images * Compositional Interpretation – Visual Methodologies, Gilian Rose, 2012
➢ Manual data collection of 4.000 Instagram posts from the first image ever uploaded on Deliciously Ella’s Instagram account, in 17 January 2013, until 24 October 2018 when she partially shut down her B&M business in London
➢ Although more laborious and time consuming, manual data collection was selected, in order to a) ensure all relevant post and image information at one go and b) eliminate the cost of an Instagram analytics third party service Today’s presentation u Contributing towards the understanding of the wider social phenomena of healthy eating and wellness lifestyle practices, as particularly mediated through Instagram A brief selection of insights taken from a pilot study of 100 Instagram posts of u Seeking a productive panel discussion Deliciously Ella aiming through the juxtaposition with previous presentations: Listening to the audience at: VS listening to the person who talks to the audience Who is Deliciously Ella
➢ Ella Mills is a plant-based food advocate, an entrepreneur, a cook-book writer and…an Instagram influencer (2 million followers in October 2020)
➢ She gained publicity in 2013, after creating the blog ‘Deliciously Ella’ and sharing her personal story about fighting and managing PoTS (Postural Tachycardia Syndrome) through a plant-based diet that practiced after all medical treatments had allegedly failed her
➢ She was an advocate of clean eating right from the start of her blogging career, but denounced it in the early days of 2017, after a media backlash that put in question its effects on health and well-being Throughout the next few slides… u Five examples selected from a pilot study analysing 100 visual Instagram posts from 11 June to 4 August 2017 u These Instagram posts reflect a part of Deliciously Ella’s online and offline activities during a period of almost two months time in summer 2017 u They involve the promotion of new product launches and work events, the promotion of new menus and dishes served at Deliciously Ella’s London delis, as well as some significant personal life moments like the wedding of Deliciously Ella’s brother in Florence, Italy. Questions asked
What do the posts What kind of food What type of show? is depicted? meal is shown?
Who is in the Where is it all picture? happening? What do the posts show?
Food 53%
People 13% Food and People (in various settings) 14%
Other 20% (Nature, Animals, Inspirational Quotes, and so on) What kind of food is depicted? Top 10 foods
Ella’s Avocado Raspberries Granola
Sweet Carrots Chia Pudding Potatoes
Coconut Ella’s Yoghurt Bircher Musli
Ella’s Energy Blueberries Balls What kind of meal is shown?
Breakfast 29% Lunch 19% Snack 13%
Other 14% Dessert 10% Smoothie 9% Salad 6% Who is in the picture?
Ella and Matt Ella, Matt and Ella (21%) (38%) Austin (8%)
Other 33% (Ella with friends, relatives, colleagues and so on) Where is this all happening?
UK (London mainly) 75% Italy 25% What is this all about? u It is certainly not only about food u It is also about a projection of a certain kind of appearance, which includes personal life details
Authenticity (Goffman)
u This is not just marketing of products or the promotion of plant-based food, but a staging of personal experiences and real-life moments in ways such as to make the actor appear ‘authentic’ or ‘real’ to the intended audience What is this all about?
u A comparison with social distributions of food tastes among classes seems to possibly explain why there is focus on certain foods:
Distinction (Bourdieu) (a) Johnston and Baumann, 2007 (b) Fuster, 2013 (c) Costa, Zepeda and Sirieix, 2014 (d) Atkinson and Deeming, 2015 (e) Pechey and Monsivais, 2016 (f) Palma, Ness, Anderson, 2017
u Avocado, chia seeds, granola, coconut, poppy seeds, turmeric and so on What is this all about? u Mingling food with people/objects can be based on the idea of valuing and marketing ‘artisanal’ forms of food production
u Food here is not just a product but the nexus “Authenticity- and-cosmopolitan (person)food”
Enrichment (Boltanski/Esquierre)
u There is a sense in which the person markets herself (and her family) along with her foods to create extra economic value. What she sells is not just a product, but a whole narrative (‘authenticity’) that accompanies the (prosaic, in many respects) product and ‘transforms’ it into an object desired by others.
Session 4. The potential of social media analysis (13:10-13:55)
13:10 Using social data to surface food trends and understand audiences Anna Rudkevych (Senior Research Manager, Pulsar)
13:20 FSA Covid-19 Social Media Listening Beatrix Siemering (Strategic Surveillance, FSA)
13:30 Understanding plant-based food images beyond the wellness hype Alexandra Boutopoulou (PhD candidate, University of Sheffield)
13:40 Session 4 Panel Discussion Session 5. Public engagement with the National Food Strategy (13:55-14:25)
13:55 Public engagement with the National Food Strategy Professor Peter Jackson (Co-Director of the University of Sheffield, Institute for Sustainable Food).
14:10 Session 5 Q&A. The National Food Strategy (public dialogue and digital engagement)
Professor Peter Jackson Co-Director, Institute for Sustainable Food
ESRC Festival of Social Science 9 November 2020 National Food Strategy
• National Food Strategy, led by Henry Dimbleby on behalf of the Department for Food, Environment and Rural Affairs (Defra) • My role to chair the Oversight Group, responsible for delivering the public dialogue part of the NFS – speaking today in a personal capacity • Relevant to ‘understanding food in a digital world’ because the COVID-19 pandemic meant that the dialogues were forced to move from face-to-face to online delivery between phases 1 and 2. National Food Strategy
• The Strategy will be published in two parts • Part 1 focuses on the needs of ‘our most disadvantaged children’ (extension of free school meals) and on ‘sovereignty, standards and scrutiny’ in the context of post-Brexit trade deals • Part 2 will focus on the food system: the economics and power dynamics that shape it and the benefits and harms it brings (health, climate change, biodiversity, pollution, AMR, zoonotic diseases and sustainable resource use). The public dialogues
• Delivered by Hopkins van Mil, with input from Sciencewise • Preliminary work included the collection of a series of ‘vox pop’ videos and interviews with a wide range of experts • Phase 1: face-to-face meetings with diverse groups of c.40 people in Grimsby, Kendal, Bristol, Norwich and Lewisham • Carefully structured to maximise participation. Phase 1 findings
• Importance of cost (affordability, access) • Links between diet and health • Questions of ‘choice’ (variety) • Challenges in understanding the complexity of the food system (and range of actors and institutions that shape what we buy and eat) • Challenges in recognising trade-offs (rather than simple interventions that only address one issue). Moving online
• Phase 2 involved the same participants in the same places but delivered online • HvM worked hard to maintain engagement between Phases 1 and 2 asking participants to keep a food journal, share their thoughts and maintain contact • Only one or two people dropped out • Aim to ensure that ‘quality of dialogue remains as high on Zoom as it would in person’ (breakout groups of 6 ppts per facilitator); allowing more time to think/reflect • All had access to laptop, tablet or mobile phone – but online engagement is more than access to technology. Epistemologies of engagement
• The nature of evidence gained from different engagement methods • No less valid that face-to-face, but hard to make direct comparisons between findings from the two phases – need to acknowledge and address the differences • Issues of voice, confidence, positionality; visual cues, turn-taking, body language and eye contact. Challenges and opportunities
• Potential exclusion of those unfamiliar with digital technologies? • Potential empowerment of those who are reluctant to ‘speak out’ in face-to-face settings? • Group dynamic likely to be different from phase 1: e.g. discussion of cooking on a budget and alleged decline of cooking skills (deficit model); very limited discussion of climate crisis, until prompted, or the links between food and environment (except in context of food waste). Impact of COVID-19 • Not just a methodological shift from F-2-F to online • Context also changed with the pandemic, leading to changes in people’s dietary attitudes and behaviours • C-19 highlighted the vulnerability of the food system (just-in-time delivery, temporary shortages and ‘panic buying’); increased food poverty; changes in hospitality sector (bars and restaurants); more use of online delivery etc. • Paradoxical trends: consuming more convenience food but also more cooking from scratch: need evidence of these trends – not moralization. On-going issues
• Narrow focus on specific issues (such as meat reduction) or wider debate about the environmental and health impacts of different dietary choices? • How much to prompt the dialogue (sufficient information to enable participants to engage effectively but not to steer the conversation)? • How ensure an appropriate balance in stimulus materials (e.g. from the meat industry), avoiding support for a particular view (‘to restore our health and environment’)? On-going issues
• Avoiding jargon: references to the ‘social contract’ replaced with an emphasis on ‘roles, responsibilities and expectations’ • Avoiding loaded terms and false dichotomies (e.g. about ‘normal’ vs ‘novel’ foods) • Stretched timetable: long gap between phases (Phase 2 delayed from April-May to Sept-Oct); no time for Citizen Assembly or National Summit? • Ensuring confidentiality while allowing future access to the data (archival challenges). Conclusion
• Public engagement at the science-policy interface is increasingly popular – see current work by Defra’s Social Science Expert Group (led by Susan Owens) • Some of the issues I’ve raised are relevant to any form of public engagement – some more specific to online/digital engagement • ‘Managing expectations’: will participants’ views be reflected in the final report? will recommendations be enacted (via White Paper and/or Spending Review)? Contacts:
Professor Peter Jackson Institute for Sustainable Food University of Sheffield Sheffield S10 2TN [email protected] Session 5. Public engagement with the National Food Strategy (13:55-14:25)
13:55 Public engagement with the National Food Strategy Professor Peter Jackson (Co-Director of the University of Sheffield, Institute for Sustainable Food).
14:10 Session 5 Q&A. Closing Remarks:
Professor Rick Mumford Head of FSA Science, Evidence & Research Directorate