Understanding survey nonresponse behaviours: evidence and practical solutions

3. Barriers and potential solutions: opportunity

The ‘opportunity’ element of COM-B focuses on external influences that shape people’s opportunity to take part in a behaviour. In the survey context, this might include survey setting and social dynamics, both of which can impact people’s opportunity to participate.

This research identified two main barriers relating to opportunity: time, and social norms about taking part.

Time

Summary

Lack of time is consistently given as a reason for not taking part when studies have asked nonresponders about this directly. While the literature emphasises that ‘time’ can be used to mask other underlying reasons, qualitative research with the general public highlights genuine concerns about time as a key barrier, particularly (though not only) for younger people. Survey interviewers felt that time issues were exacerbated by rising home working, blurring the lines between ‘work’ and ‘free’ time.

Existing literature consistently identifies lack of time and disinterest as primary reasons people give for not participating in surveys^[7]. For example, ‘too busy’ was the most commonly reported reason for noncompletion of the 2022 Scottish Census, despite the census being mandatory in the UK – a requirement that sets it apart from most social surveys.^[8]

Within the general public qualitative research, there were genuine concerns about the length of the survey and if participants could realistically spare the required time. Indeed, phrases such as “time consuming” and “how long is it going to take?” arose spontaneously across general public interviews when participants were asked what came to mind upon hearing the word ‘survey’.

While lack of time may sometimes be used as a polite excuse, concern about the actual time commitment involved in surveys is nonetheless an important barrier for many potential participants. The influence of time in the decision to take part is likely to vary considerably between subgroups. It might be hypothesised that for older, retired people, time is not necessarily a major issue, but for parents of young children or others with a lot of caring and work responsibilities, it may be critical. The SHS, which was used as a case study in general public groups, takes around 45 minutes on average to complete^[9]. Some general public participants, particularly (though not only) younger people, felt that 45 minutes was too long a time commitment. In addition to concerns about the stated length, participants were concerned that an interviewer-administered survey would actually take up more of their time in practice, with additional time needed for small talk and introductions:

“It's quite a long time. And then it won't just be 45 minutes because they'll be like coming into the house. It'll take time. It'll take more time than just that. It's just too long for me, I think.” (Public Group 2, aged 16-24, Rural areas)

Disabled people also highlighted challenges around the time commitment involved, particularly where there were already additional burdens on their time around managing their conditions.

Survey interviewers noted that the rise in home working since the Covid-19 pandemic has complicated how people respond to survey interviewers calling during the working day. While people may be more likely to be at home, they may be less likely to have time available when the interviewer calls, which survey interviewers felt might be leading to more refusals. They also noted the increased challenge of potential respondents working late into the evening, therefore becoming unavailable to be surveyed at times when they previously would have been free. The potential impact on refusals of blurring boundaries from home working was supported by comments from the general public:

“You could be working from home, but that's the last time you want anyone to come and visit you because you've got work to do.” (Public Group 4, aged 25+, Rural areas)

While existing research highlights lack of time as a common reason for nonparticipation, it is important to recognise that stated reasons may not always reveal the full picture. As Loosveldt and Joye (2016) note, the reasons people state for not taking part may not always reflect the underlying reason or their true motivations, with ‘lack of time’ potentially serving as an excuse or a more socially acceptable explanation when someone does not want to disclose their actual reasons. This was acknowledged by members of the general public interviewed for this research – as one participant said, “if I want to do something, I’ll find the time”.

Potential solutions

The most direct solution to ‘time’ as a barrier would be to shorten the length of surveys. This could be approached in various ways, including:

• Reducing the content but keeping the same design (that is, continuing to interview respondents face-to-face but reducing the content).
• Dividing the survey into a shorter ‘core’ face-to-face survey, followed by an option to continue to a ‘full’ interview, either face-to-face or by another mode. This design could be applied universally (such that all respondents complete the ‘core’ survey, and those who are willing to do so continue to a full interview). Alternatively, interviewers could be allowed to offer a much shorter ‘core’ interview only to those who are uncertain about taking part to avoid refusals.^[10]
• Offering the option of taking part by other modes (such as telephone, video or online) from the outset. Providing greater flexibility about both when and how people take part recognises that some people who are not able to make time for a face-to-face survey during interviewer working hours might nonetheless be willing and able to complete an online survey in their own time.

Whatever approach is taken, shortening the surveys and/or incorporating alternative modes would involve significant trade-offs in terms of the scope, completeness and, potentially, quality of data collected. There would also likely be significant costs, at least in the short term, associated with redesigning the surveys in this way.^[11]

Younger people interviewed for this research indicated that they might be more willing to take part in the SHS if the survey was 20 minutes rather than 45, or if there was an option to complete some sections later online. However, there were more mixed views on shortening the length among older respondents, with questions raised about whether the survey would still produce ‘meaningful’ data if it was much shorter.

Participants suggested that providing upfront information about the number of questions and the expected maximum completion time, and a commitment that the interviewer would leave on reaching this time, might help alleviate some of their concerns about the time commitment involved. In addition, it was suggested the government could offer people an ‘exemption’ from future government survey research for a fixed period after taking part (this was compared with the exemption people receive after serving on juries). In fact, the Scottish Government’s approach to selecting the sample for the three surveys that are the focus of this research already ensures that the same addresses are not resampled in a 4-year period. This could be more explicitly highlighted at the outset, as an advantage of taking part. However, it would not be possible to guarantee that the same households would not be sampled for any other Scottish or UK Government survey. It would also be important to consider potential implications for other Scottish Government studies that currently use recontact data (that is, contact details for those who are happy to be contacted about future research after taking part) from the three surveys as sample frames.

Social norms about taking part

Summary

Stakeholders and survey interviewers felt that civic engagement had weakened over time, with the consequence that people are less willing to take part in activities for social benefit, including surveys. The evidence for this from the literature was uncertain. Although there is some evidence that other sorts of prosocial behaviour (e.g. formal volunteering and giving blood) have declined, there are many confounding factors that could explain this. There was clearer evidence from the general public that feeling surveys are not for ‘people like them’ might deter some from taking part, particularly those who are already less well represented or in minority positions.

Stakeholders and survey interviewers who took part in this research felt that people’s sense of ‘civic engagement’ in general may have weakened over time, and that this might be contributing to higher nonresponse. Falling civic engagement has been a long-standing theme in social science literature, particularly since Robert Putnam’s ‘Bowling Alone’ (1995), which hypothesised that the networks, relationships and norms (collectively termed ‘social capital’) that facilitate cooperation and mutual benefit among individuals and groups had been eroding over time. While focused on the USA, Putnam’s theories have gained a wide audience across liberal democracies.

Establishing whether civic engagement, or ‘prosocial’ behaviours, are indeed in decline, and whether this is a plausible explanation of the rise in nonresponse observed on the Scottish Government surveys since 2012, is challenging. Many of the available measures are themselves derived from surveys and may therefore themselves be impacted by changes in nonresponse over time. However, there is nonetheless some evidence that certain sorts of civic engagement may have been eroded in recent years. For example, SHS data suggests that formal volunteering has decreased in recent years (Scottish Government, 2024). Research by Volunteer Scotland (2025) has suggested that economic pressures linked to the cost-of-living crisis may explain this decline; that people are less able financially to commit time or resource to formal volunteering, rather than having become less ‘civic minded’. This is supported by the finding that rates of informal volunteering (where people provide unpaid help to others they are not related to) have remained broadly stable over time (Scottish Government, 2024).

The proportion of people in Scotland giving blood, which could be taken as another indication of ‘civic participation’, has also reportedly declined, particularly since the Covid-19 pandemic. However, again there are a number of confounding factors, including an ageing population and scaling back of mobile blood collection units post-2020.

Taken together, although there is some evidence of a decline in ‘prosocial’ behaviours, it is hard to be certain whether these are a reflection of a general reduced willingness to engage in activities ‘for the public good’, which might also translate into a reduced willingness to complete surveys, or whether there are specific demographic, economic or social factors impacting on these behaviours. While these factors may also impact on people’s ability to take part in surveys (via reduced time, for example, discussed above), they require different solutions to those that aim to increase people’s sense of civic engagement per se.

However, a sense of ‘civic engagement’ is not the only way in which social norms might impact on people’s participation in surveys. Whether or not people see Scottish Government surveys as something people ‘like them’ would typically do may also be an important influence on behaviour. There was a clear stereotype among members of the general public interviewed for this research that those who take part in such surveys tend to be older, with more time available, as well as being more informed and/or more “used to being listened to”. For some participants who did not feel they fit this description themselves, whether because of age, educational background, or minority position, this led them to feel government surveys were not things people “like them” would normally participate in.

“A lot of young people like from the ages from like 16 to kind of like my age, don’t take part in surveys. A lot more informed intelligent people take part in surveys, maybe that are younger … but a lot of them that I know, people that are in an older demographic for the ages, maybe like 30 and a wee bit above that, they’re the ones that are actively taking the surveys. It’s mostly those kinds of people that are taking them, not really younger people.” (Interview 1, under 25, lower qualifications)

Potential solutions

There are many creative ways in which ‘social norms’ around who a ‘typical respondent’ is could be challenged, such as creating video clips or using quotes from respondents who do not fit the above stereotypes talking about their reasons for taking part. While such activities are arguably unlikely to have an immediate impact on response, they could form part of a wider effort to drive a ‘refreshed citizen relationship’ with Scottish Government surveys (see discussion in chapter 2).