Survey nonresponse research: appendices

Appendix C: Qualitative research with interviewers and stakeholders

Summary of key findings

• Findings are based on a focus group with Scottish Government survey leads and contractors and three focus groups with experienced survey interviewers working on the three main Scottish Government cross-sectional general public surveys.
• Key perceived reasons for (increasing) non-contact discussed by participants included:
  • Working from home: While increased home working makes it easier to find potential respondents at home, it also means they may be more stressed and less willing to participate.
  • Challenges accessing buildings: Access to flats was felt to be increasingly difficult, especially newer builds lacking trade buzzers. Doorbell cameras also enabled some potential respondents to avoid contact, even if they are at home.
  • Broken appointments: Interviewers reported an increase in broken appointments, hypothesising that respondents may not feel able to refuse on the doorstep, so agree to appointments without intending to keep them.
• Key perceived reasons for (increasing) refusals included:
  • Lack of time: Potential respondents frequently cite lack of time, though interviewers suspect this may often be a perceived rather than actual time constraint. Views on whether shortening the interviews would improve response rates varied, however.
  • Lack of interest/perceived value: Potential respondents often express disinterest or question the survey's value as a reason for not taking part.
  • Negative perceptions of institutions: Interviewers felt that negative views of the Scottish Government, NHS (for SHeS), or criminal justice system (for SCJS) were increasingly common reasons for refusal.
  • Decline in civic engagement: A perceived decline in civic engagement and social responsibility was felt to be a factor in declining response.
  • Fear of scams: Heightened fear of scams was felt to be making it more difficult to convince potential respondents of interviewers' legitimacy, despite ID cards.
  • Data protection concerns: Increased awareness of data protection and misuse was felt to have increased concerns regarding data security and usage. Reluctance to provide personal information for random respondent selection also poses a challenge for interviewers.
  • Survey fatigue: The prevalence of survey requests, particularly for feedback on goods and services, was felt to be leading to desensitisation and reduced willingness to participate.
  • Social isolation: There were mixed views on the impact of social isolation, with some suggesting it decreases willingness to participate while others believed it can be a motivating factor.
• In terms of groups that were more difficult to include than others, interviewers noted that assessing capacity, combined with a perceived increase in proxy refusals from family members, was increasing the difficulty of including some older people, particularly those with fluctuating dementia. Language barriers could also hinder interviewing those with limited English.
• In contrast, public sector workers (who may attach higher value to government survey data) and people in rural areas (who feel underrepresented) were felt to be more willing to participate.
• Stakeholders and interviewers discussed various factors that might help mitigate non-response, including:
  • Skilled interviewers: Experienced interviewers who can engage potential respondents and address their concerns were considered crucial.
  • Interviewer techniques: Gathering information about the household in advance, presenting a professional and friendly demeanour, emphasising the importance of their individual response, and tailoring descriptions of the survey to the respondent's circumstances were all felt to be useful techniques.
  • Advance materials: While doubts remain about how many people read them, advance letters were considered important as conversation starters and proof of legitimacy. However, some weaknesses in current materials were identified.
  • Incentives: Mixed views were expressed on the overall impact of incentives. Concerns were raised about potential negative consequences of discretionary or differential incentives.

Introduction and method

This Appendix presents findings from qualitative research with Scottish Government survey leads, survey contractors from Ipsos and ScotCen, and field interviewers from Ipsos and ScotCen. The research team conducted four online focus groups between 4 February and 20 February 2025, each lasting around 90 minutes.

Three of the focus groups were with field interviewers, one with interviewers working on each of the Scottish Government’s main general population surveys. Eight interviewers were recruited for each group, and 22 took part in total. Interviewers were recruited through the Ipsos and ScotCen field teams. They were asked to prioritise experienced interviewers with good response rates, as this group were considered most likely to have useful reflections on what does (and does not) work in persuading respondents and to be able to reflect on changes in response over time.

The fourth focus group was conducted with Scottish Government survey leads and survey contractors from Ipsos and ScotCen for each of the three main general population surveys, supplemented by email correspondence with one contractor who was unable to make the date of the group.

Each focus group covered the following discussion topics:

• Perceptions of the main reasons for nonresponse.
• Views on what works in practice to mitigate nonresponse.
• Perceptions of behavioural interventions or experiments.

The first section of this appendix discusses participants’ overall views on reasons for nonresponse, how (if at all) reasons for nonresponse have changed over time, and the factors that have driven these trends. The second section explores participants’ views on a range of potential initiatives, both formal and informal, to mitigate nonresponse.

Reasons for non-contacts

Contacting potential respondents is the first hurdle for interviewers to overcome when attempting to complete a survey. On face-to-face surveys such as the SHS, SCJS and SHeS, this can be divided into two parts: (1) calling at the potential respondent’s property while they are inside and (2) getting the potential respondent to open the door.

Home working

One of the most noticeable recent changes in this regard has been the increased prevalence of working from home. Stakeholders’ reflections on the impact of this indicate that it may help reduce non-contacts but potentially increases refusals. Interviewers noted that the increase in home working helped them to catch potential respondents while they are in the house. However, potential respondents who are working from home when the interviewer calls were felt to be more likely to be visibly stressed, and less likely to have time available to take part when the interviewer calls.

Interviewers described the need to navigate this situation delicately, pausing to give potential respondents space and offering to return at a more suitable time to avoid the interaction ending in a refusal. One group felt that, although it is easier to catch potential respondents in the house and to make an appointment, the boundary between work and home has changed. They 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.

Survey leads and contractors hypothesised that a shift towards working from home may also have changed the demographics of who answers the door, potentially leading to more immediate refusals before the interviewer is able to reach the person who was going to be selected for interview. Finally, some interviewers noted that working from home could be given as an easy excuse to turn the interviewer away, regardless of whether people are actually working.

Blocks of flats

Interviewers, survey leads and contractors were aware that a high proportion of Scottish households, especially in cities, live in flats (as noted in appendix A, the Scottish Census indicates that almost a third of households in Scotland live in purpose-built flats). They also felt that getting into flats has become more difficult: interviewers noted that newer blocks of flats often do not have ‘trade’ buzzers which they would otherwise use to gain access to the stairwell. This has created additional barriers for interviewers, especially in urban areas. At the same time, it was suggested that it rarely prevents interviews from taking place if the interviewer is persistent. For example, interviewers said they would simply resort to pressing other buttons on the intercom and politely asking someone to let them in to the stair.

Doorbell cameras

There were mixed views about the impact of modern doorbell technology on the rate of non-contacts. Doorbell cameras allow potential respondents to see the interviewer on their doorstep and choose not to respond. One view among interviewers was that there had been “massive” increases in the number of people who pretend not to be in the property, sometimes even though interviewers have seen them through their window.

Some interviewers said they do speak to the people through ring doorbells. These participants had occasionally found doorbell technology to be useful in helping them to get in contact with potential respondents who were genuinely away from home. The technology allowed some to explain their presence and to schedule callback appointments. However, others said they tend to knock on the door instead of using doorbell cameras, to try and increase the chance of their having an opportunity to speak to the potential respondent.

Broken appointments

Callback appointments are only useful if they are kept. Interviewers described what they perceived to be an increase in broken appointments. One group felt that people have become more willing to break appointments – they might hide in the house or drive away as the interviewer approaches. Others thought this was because potential respondents do not feel sufficiently empowered to say ‘no’ to a persuasive interviewer on their doorstep, and therefore book appointments which they do not intend to keep.

Making contact with the household is only the first hurdle in gaining an interview. As discussed in appendix B, the main reason for the increase in nonresponse on the Scottish Government surveys is an increase in refusals, rather than non-contact. Interviewers, survey leads and contractors identified a range of reasons why people refuse to take part in surveys, once the interviewers have successfully made contact with them.

Reasons for refusals (explicit)

Interviewers said that they often asked respondents why they did not want to take part; sometimes, this gives the interviewer one last chance to persuade respondents to complete. As the groups took part while interviewers were conducting fieldwork across the Scottish Government surveys, some had decided to ask respondents for their reasons specifically in preparation for the discussion. The reasons that respondents give overtly could usually be categorised as a lack of time, lack of interest and perceived value of surveys, and negative attitudes towards public institutions.

Lack of time

Survey leads and contractors discussed what they perceived to be an increasing difficulty in persuading people to dedicate time to surveys. Interviewers framed this as potential respondents’ perceptions of their lack of time, rather than an actual lack of time. Interviewers working on the SHeS felt that these feelings are due to modern life and potential respondents being “bombarded” with different demands on their attention, via notifications, emails, texts and letters.

Interviewers generally felt that while time constraints were frequently given as a reason, they may not reflect the true reason for potential respondents not taking part.

“People feel like they’re busier, even if they might not be.” (Group 2: SHS interviewers)

Perhaps surprisingly, few interviewers believed that the length of the survey had a significant impact on response rate. This is arguably in line with the findings on survey length Ormston et al (2024) – this research found no clear consensus among survey methodologists as to the ‘optimal’ interview length for different modes of administration. On the SCJS, interviewers said that victim forms are time consuming, but that these do not cause people to withdraw from survey. This was echoed by interviewers working on SHeS, which can be prolonged depending on the number of biological measurements and the number of household members taking part. Their view was that, once people start, they are likely to go through right to the end. However, there was a concern that respondents can feel the interviewer has ‘taken advantage’ of them if the interview lasts too long, which might gradually erode the public’s willingness to take part in surveys in future.

Lack of interest

One of the most common explicit reasons for refusals, when interviewers asked potential respondents directly, was that they are not interested in it. Survey leads/contractors discussed their perception that there is a lack of understanding of the purpose of surveys and of the importance of statistics. They felt that public communications around the surveys have struggled to convey this. As a result, respondents may think the survey is about their individual household rather than producing aggregate statistics.

However, field interviewers found that potential respondents often are interested in the surveys if they successfully tailor their responses and give examples of how the surveys have been used to change policy previously.

Interviewers for the SHS said that the most common question they field is, "will this make any difference?" Depending on the survey, interviewers felt they could not always easily show what difference it makes. This was felt to be easier for interviewers on the SHeS to explain, as they can provide examples such as changes to public health policy on smoking and alcohol. Meanwhile, SHS interviewers said that they used to mention the winter fuel payment as an example; however, this is now a topic that interviewers avoid (given the recent withdrawal of universal winter fuel payments). As this appendix will discuss later, however, interviewers have often found creative ways to think of something relevant to the participant.

Negative perceptions of institutions

Interviewers report that many of those with negative perceptions of the Scottish Government will give this as their first reaction. Interviewers’ felt that the political landscape has been more polarised in the past ten years, citing events such as the 2014 Independence referendum, the 2016 Brexit referendum and the enforcement of lockdown rules between 2020-2021 due to the coronavirus pandemic. Survey leads and contractors also felt that there could, for some potential respondents, be a potential negative impact of association between the Scottish Government logo branding and the current party of Scottish Government, though equally this could have a positive association for others.

Interviewers echoed this sentiment, stating that people can be put off when they hear 'Scottish Government'. Perceptions among field interviewers were that they see more people now with anti-establishment views or feeling disillusionment with politicians.

“Quite a lot of people are very vocal about not wanting to help the government. I don't know if that's just the areas I work in, but I'll get a lot of people say, ‘I know, I've read the letter, I'm not interested in helping this government’.” (Group 3: SCJS interviewers)

Some interviewers respond to this by emphasising that is the non-political (civil service) part of the Scottish Government, rather than politicians, that primarily design and use the survey.

Similarly, interviewers reported that negative perceptions of the NHS (for SHeS) or criminal justice system (for SCJS) are regularly brought up by potential respondents who are unwilling to take part in the interview. In many cases, interviewers said that prolonged waiting lists and the difficulty of scheduling medical appointments had contributed to reactions to SHeS. Interviewers on SHeS talked about the effort which can be required to explain to potential respondents that they are not working for the NHS.

These frustrations sometimes manifested in respondents describing the surveys as a “waste of time and money”. People sometimes do not see the surveys as worthwhile as they feel that money could be better spent elsewhere, or that the government already know the problems that need to be addressed. This was brought up by interviewers for the SCJS and SHeS.

"[They say that] there's not enough police, so [they ask] how is there funding for this?" (Group 3: SCJS interviewers)

Interviewers working on the SCJS said that some people they come across express negative views on the criminal justice system, and do not want to take part in something that helps it. This was observed to have been a particular problem among young men living alone; not only are they more difficult to make contact with to begin with, but SCJS interviewers said that some young men living alone were more likely to have been in contact with the police and may even feel targeted by the survey, or concerned that they might say something which incriminates themselves. Interviewers described how they had tried to reassure potential respondents that felt this way:

“Usually when I get that, I'll say, “Well, do you know that people between 16 and 24 are more likely to be stopped by the police? And we're looking to see if people have had experience of that.” […] Sometimes you'll maybe change their mind, or they'll make an appointment, but you go back and they don't answer the doors.” (Group 3: SCJS interviewers)

At the other end of the spectrum, interviewers said that other potential respondents feel they do not know anything about the criminal justice system and therefore do not think the SCJS is relevant to them. The perceived relevance of the SCJS to the wider public was also raised by survey leads and contractors, primarily in relation to people who have not experienced any crime.

Reasons for refusals (implicit)

In addition to the reasons above, interviewers, survey leads and contractors discussed a variety of other factors they thought contributed to survey nonresponse but which people might not overtly mention to an interviewer.

Civic engagement and civic responsibility

Survey leads, contractors and field interviewers all felt that there has been a general decline in civic engagement and the sense of civic responsibility that would intrinsically compel people to participate in government initiatives.

“I think there's just not the same sense of social responsibility that there maybe used to be. A lot of people will go, ‘no, I just don't want to do it’. They won't really give you an answer. And if you try and probe and say how it helps their community, they'll just kind of poo poo it.” (Group 3: SCJS interviewers)

Although social responsibility was identified as a motivating factor for potential respondents who are more civically minded, interviewers described it as “an uphill struggle” to persuade potential respondents who do not think in this way:

“People who do have a good sense of social responsibility will say things like, ‘oh, I think these things are important. You know, you don't change things unless you take part in these things’, you know. And so I think people either have a good sense of social responsibility or they don't” (Group 3: SCJS interviewers)

Fear of scams

Field interviewers from all three groups noted that some people do not take part in surveys because of fear of scams. There was uncertainty about whether this was due to an increase in the prevalence of scams, or an increase in people’s awareness of scamming techniques, but the fact remained that interviewers reported finding it increasingly difficult to reassure potential respondents that they are legitimate. Although they carry ID cards, interviewers find these are not always sufficient to reassure a potential respondent. There was a belief that this could pose a particular challenge for younger male interviewers (compared with older female interviewers), who might more easily arouse suspicion from a stranger.

Survey leads and contractors thought that the expansion of ‘no cold calling’ schemes has made it more difficult to get completes. This was confirmed in discussions with interviewers. For example, an interviewer working in the Scottish Borders reported that signs had been placed on lampposts warning potential respondents not to let strangers into their property. Another interviewer working on SHeS spoke about occasions on which potential respondents had phoned their local general practice to ask about whether the survey is genuine. Reception staff, unaware that the survey was taking place in their local area, had confirmed that it had not come from the surgery and was likely to be a scam. Interviewers said that, on some occasions, these warnings had spread through social media to an even broader audience.

Data protection concerns

Related to concerns about scams, focus group participants cited a lack of trust related to data privacy and usage as a growing reason for nonresponse. Survey leads and contractors felt that there is greater public awareness of data protection, including the value to companies and governments of their personal data, and of high-profile cases of data misuse. They hypothesised that this may have contributed to an erosion in potential respondents’ propensity to trust that their personal data collected for surveys will be stored securely and used appropriately.

Interviewers also spoke about trust related to data usage, as this sometimes affected their process for randomly selecting a respondent to take part from the household. For example, SCJS interviewers said that respondents have been reluctant to provide any personal information before they have consented to take part. However, the procedure for selecting a random adult requires each adult’s name (or at least their initials) and some interviewers had faced resistance against this from potential respondents. Interviewers reported that they had less freedom to adapt their procedures to account for this resistance, because the random selection process is central to each survey’s sampling design. Previously, interviewers had been permitted to use a 'Kish grid'. This method uses a pre-assigned table of random numbers to identify the person to be interviewed, and meant respondents did not need to give any personal details prior to finding out who had been chosen to take part. Interviewers advocated for bringing these tools back into use to overcome barriers around the selection process.

Survey fatigue

Field interviewers and survey leads expressed their perception that, when potential respondents claim a lack of interest, this might reflect survey fatigue. Participants noted an increasing prevalence of survey requests, especially those asking for feedback on goods and services. As a result, it was believed that potential respondents have become desensitised to being asked to take part in surveys, which may previously have been a relatively novel and unusual occurrence.

"People get quizzed about everything...there's maybe something around that." (Group 1: Survey leads and contractors)

“You get asked to do a survey now when you’ve just bought a cup of coffee.” (Group 2: SHS interviewers)

Participants said it was important to make every effort to differentiate the SHS, SCJS and SHeS from these more common, and often repetitive, types of surveys (one group of interviewers referred to these as ‘coffee cup surveys’). However, a related challenge, identified by survey leads / contractors, was a perception that people can be surprised to be asked to be interviewed at length in their own home, and that it can be difficult to explain why this kind of data collection is needed. It was suggested that many people now assume either that everything can be done online rather than face-to-face, or that organisations can access information (e.g. health records) to answer the kinds of questions these surveys cover without needing to collect further data.

Social isolation

Finally, there were mixed views on the extent to which social isolation has been a factor driving the change in response rates over time. A member of the survey leads and contractors group hypothesised that the pandemic may have contributed to people becoming less ‘pro social’ and therefore decreased their willingness to participate. However, some interviewers felt that isolation and loneliness had been a motivating factor for taking part for some respondents.

Variation by sociodemographic factors

Survey leads and contractors noted two specific issues which they felt may be making it more difficult to get some older people to participate in surveys, especially those who may have fluctuating dementia (an issue which is likely to increase with Scotland’s ageing population).

1. It could be difficult for an interviewer to assess whether older respondents have the capacity to take part

2. A perceived increase in the number of complaints from family members concerned about their elderly relatives participating in surveys, resulting in proxy refusals.

The latter point was mirrored in discussions among field interviewers working on the SHS, who noted that family members occasionally assume that their parent cannot take part, even if they might be capable of participating.

Beyond this, survey leads and contractors expressed uncertainty about whether there are subgroup-specific drivers for nonresponse, noting the difficulty of obtaining definitive evidence on drivers of nonresponse for specific subgroups.

Field interviewers noted that language had presented a major barrier to interviewing those with limited English in the past. They felt this had been extremely difficult to overcome, even when they tried to use telephone interpreters. However, while this might limit interviews with those with limited English specifically, there was also a view among interviewers that, provided they had reasonable English comprehension, migrants to Scotland were often more willing to take part than potential respondents who had been born in the UK. Interviewers working on SHeS specifically noted that people from countries sometimes described as ‘Eastern European’ (including Poland) had been more likely to take part. They were not sure of the reason for this, but hypothesised that they were perhaps more enthusiastic about being part of society in Scotland having chosen to move here.

Other sub-groups interviewers felt were more likely to be willing to take part included:

• Public sector workers, who tended to be more interested in the findings of these surveys and were felt to more clearly understand their value.
• People living in rural areas, where potential respondents often felt that their communities were underrepresented. Interviewers said they used this as a selling point to persuade potential respondents to take part.

Mitigating nonresponse

Survey leads and contractors emphasised the crucial role of skilled interviewers, particularly the importance of experienced, skilled interviewers who can effectively engage with potential respondents and address their concerns.

"The role of the interviewer is absolutely crucial in... turning people around." (Group 1: Survey leads and contractors)

Field interviewers were asked to share their tips and tricks during the focus groups, and any other less formal measures they had found to be successful in increasing response rates.

Interviewer skills to avoid non-contacts

“Your interview starts before you get to the door.” (Group 2: SHS interviewers)

Interviewers said they tried to gather as much information as they could as they approached each property, to identify clues about who the potential respondent might be. They would use this information to assess whether it is an appropriate time to call, before they knock. For example, if there is evidence to suggest that an older person lives in the property, interviewers would come back the next morning rather than knock during the evening (when they felt older people might be more reticent about answering the door to a stranger). Similarly, if they could see children’s toys and smell cooking when the door is opened, interviewers said they would quickly acknowledge that they have called at an inconvenient time and would aim to make a call-back appointment rather than start explaining the survey.

Contacting neighbours to ask whether the property is occupied, explain their reason for trying to get in touch, and ask if they could suggest a good time to catch their potential respondent, was also mentioned. There was acknowledgement that this was more effective in rural areas and less useful in blocks of flats:

“Some people in flats keep themselves to themselves and genuinely don’t know their neighbours.” (Group 2: SHS interviewers)

Interviewer skills when interacting with potential respondents

An overarching theme across the different groups of field interviewers was the need to present themselves in a professional way, and maintain a friendly, confident demeanour. First impressions were considered extremely important, so interviewers would ensure they were well-organised, neatly dressed, and able to display a high degree of competence on the doorstep. Specific verbal elements interviewers mentioned using when trying to gain a positive response included:

• Finding a reason to give the potential respondent a compliment during the early part of their interaction, to build immediate rapport and instil a feeling of reciprocity.
• Saying something which shows the potential respondent that their individual response matters. Across all three surveys, interviewers would say they have come to the potential respondent’s address specifically, explain that they are not just knocking every door, and that they are not able to replace the respondent with someone else.
• Tailoring their descriptions of the survey to the circumstances of the potential respondent and their local area. Interviewers working on the SCJS said they would tap into a person's sense of responsibility for their neighbourhood by mentioning issues known to be causing problems locally. On the SHS, interviewers said they would ask if potential respondents have had any issues with their local services: “Whatever they say, it doesn’t matter, but tell them that this is the reason why they should take part” (SHS).

Overall, interviewers felt their own skills and approach in building rapport was central to gaining a good response rate:

"I don’t think anyone does an interview for the Scottish Government, they do it for you, they’re talking to you" (Group 2: SHS interviewers)

Advance materials

Survey leads and contractors considered advance letters and other materials to be important as a prompt or prop for interviewers to use on the doorstep. While there were doubts about how many people actually read them in advance, interviewers strongly warned against taking them away, as they were an important conversation starter.

Where potential respondents were concerned about scams, interviewers said that the letter provided legitimacy and often acted as their key to get through the door (again, even where they had not previously read it). This was particularly useful where they could sense that they were close to being refused:

“Show them the address in print. Show them the letter. Pause so they can tell you if they have changed their mind. Give them a copy of the letter, and back off.” (Group 3: SCJS interviewers)

On some studies, interviewers said they could personalise the advance letters. For example, interviewers said they would introduce themselves, sign their name on the envelope and give the day that they planned to visit (if known). They felt that seeing something handwritten meant respondents were more likely to register it.

“Even if they hadn't read the rest of the letter, they remember your name” (Group 4: SHeS interviewers).

Beyond their use as a visual prompt, interviewers said that the usefulness of advance materials is dependent on whether they are succinct, easy to understand, and followed up with a visit in a timely manner. Despite a number of changes to materials over the years, interviewers felt that there are still some weaknesses:

• Interviewers on SHeS fed back that the letter makes a lot of potential respondents think they are answering a survey about the NHS itself.

"We have to spend a lot of time telling people it's not about that." (Group 4: SHeS interviewers)

• Interviewers on the SHS fed back that the portal section of the advance letter should be removed, as it signposts potential respondents towards an easy way of opting out of the survey before the interviewer even calls.

“I understand why they brought it in during Covid, but it takes up space on the letter and people mainly use it for office refusals.” (Group 2: SHS interviewers)

At the same time, survey leads and contractors felt a lot of effort had been put into improving materials across the three surveys and while there was always room for improvement, they were not sure that this would make a big difference to response. They also noted the inevitable balance between length and how many key messages can be included.

Incentives

Survey leads felt there could be scope to explore different options with regard to incentives across the surveys. Two main options were discussed:

• Discretionary incentives, which would involve giving interviewers a small number of incentives to use at their discretion to encourage participation.
• Differential incentives, which would involve offering higher incentives to specific groups, such as those in deprived areas. It was noted that the SHeS experimented with this during the pandemic, so there would be a precedent for doing so again.

There were mixed views on incentives among field interviewers. Interviewers working on the SHeS believed that incentives were most effective at encouraging participation among households with children. For some children, particularly for teenagers, the incentive was felt to be the reason that they take part.

"It works well with child boost points as well because you hear the kids saying, when am I getting my money?" (Group 4: SHeS interviewers)

However, others thought that, for most respondents, the incentives are a bonus and are not the reason why people do the survey in the first place. One view was that not offering incentives helped to differentiate Scottish Government surveys from others, where private companies do have to pay people to share their views on products:

“In a way it adds to the legitimacy, that this is important research, this is government stuff, we’re conscious of taxpayers money.” (Group 2: SHS interviewers)

Interviewers were wary of discretionary or differential incentives. They suggested these could backfire and have a negative overall impact on response if respondents who are not offered an incentive hear about those who did. This was felt to be a particular risk in rural areas, where people are more likely to talk to their neighbours. There was also concern about the implication that it penalises respondents who are more willing or engaged.

The previous use of differential incentives on the SHeS was recalled by just one interviewer. They described an experiment where incentives had been doubled in the more deprived areas that they were working in. However, they did not notice a change in response rate, and they said it was "awkward, or even horrible" if people from areas receiving differential levels of incentive knew each other.

Other suggestions

Interviewer recruitment was highlighted as an underlying issue to increasing response rates, by survey leads, contractors and interviewers themselves. Survey leads and contractors doubted whether you can easily ‘teach’ being a good interviewer. In general, there was consensus that it is more about getting the right people in the first place. With this in mind, the group suggested more targeted interviewer recruitment: focusing recruitment efforts on individuals with the characteristics and experience associated with successful interviewers.

"[We should be] focusing more on the people that are more likely to stay and to be successful." (Group1: Survey leads and contractors)

Field interviewers mirrored this in their discussions. They described how it took them years to get good at their role and felt the skillset that they have built up cannot be taught quickly. They described how they need to be prepared with responses on the doorstep, listen and be responsive to "know which piece of your armoury to use to … persuade them." (Group 4: SHeS interviewers)

Participants felt that newer interviewers, who do not possess such an extensive ‘armoury’ of techniques, can find it difficult to meet the targets for response rates that they are aiming for, and that this in turn can discourage them from staying in their role.