Survey on knowledge and behaviours towards smoked fish: background and methodology

In November 2022, the Food Standards Agency (FSA) and Food Standards Scotland (FSS) commissioned Ipsos UK to conduct an online survey to explore UK consumers’ knowledge of and behaviours towards smoked fish consumption, particularly among people vulnerable to a higher risk of food poisoning caused by the bacteria Listeria. The key research questions were: 

• do consumers know that there is a difference between ready-to-eat (RTE) hot-smoked fish and RTE cold-smoked fish?
• how are RTE hot-smoked fish and RTE cold-smoked fish currently consumed? 
• would consumers be receptive to cooking advice for RTE hot-smoked fish and RTE cold-smoked fish?

In line with the research questions, the survey covered three broad topics: awareness of RTE smoked fish, consumption of RTE smoked fish, and vulnerable consumers’ response to risk communications and advice.

Fieldwork ran from 4th to 8th November 2022 using the Ipsos UK i: Omnibus service, an online platform that provides a nationally representative online sample with participants from Ipsos panels. Ipsos UK surveyed a representative sample of 2,478 respondents aged 16 to 75 living in the UK, including 564 respondents living in Scotland, 85 in Wales, 58 in Northern Ireland and 1771 in England. Due to the high smoked fish market share in Scotland, a boost sample in Scotland was included. This allowed for comparisons between consumers in Scotland and the rest of the UK. 

Of the total sample, 2,390 consented to the collection of their personal health-related data. Therefore, the results presented in this report are based only on this group of respondents. 

A key element of this survey was to explore the views and opinions of consumers who are at higher risk of food poisoning caused by the bacteria Listeria when eating RTE smoked fish products (referred to as ‘vulnerable consumers’ in this report). This includes people with a weakened immune system, people aged over 65 years and those who are pregnant. People who prepare food for these vulnerable consumers were also included in the survey. Of note, this survey only captured the views and opinions of people up to age 75.  

Of those who consented to the collection of their personal health-related data, 605 were consumers in groups considered vulnerable to Listeria: 301 were aged 65-75 (13% of the sample), 17 were pregnant (1% of the sample) and 331 self-reported as having a weakened immune system (14% of the sample). 

Both the overall sample and the Scotland sample were weighted to be representative of the UK adult population aged 16-75 on key demographics: age, gender, education and social grade. All the differences which have been commented on within this report are statistically significant unless otherwise stated. This report only comments on key findings from the survey. Further information, including more detailed demographic analysis, is published alongside this report in full data tabulations.

A note on the tables used within this report 

Within this report, tables have been used to present some of the findings. Within these tables statistically significant differences are indicated using letters. Each column has been assigned a letter. If a letter is shown below a percentage in the table, this indicates that this figure is significantly different compared to the figure in the corresponding column. Comparisons are only made between variables within the same row. ‘EWNI’ refers to England, Wales and Northern Ireland. 

Note on interpreting the data

Results presented in this report should be interpreted with care. All surveys are subject to a range of potential sources of error including sample imbalances which are not easily identified and corrected through weighting and errors in respondents’ interpretation of survey questions and response options. The data reported within this report relies on respondents’ self-reported behaviours. Errors could occur due to imperfect recollection, or respondents’ tendency to overreport behaviours which are perceived as being desirable and underreport undesirable behaviours. As with all survey research, correlation does not imply causation therefore causality has not been commented on in this report.