|Home | About | Journals | Submit | Contact Us | Français|
Overall satisfaction levels with social care are usually high but lower levels have been reported among black and minority ethnic (BME) service users in England. Reasons for this are poorly understood. This qualitative study therefore explored satisfaction with services among informal carer participants from five different ethnic groups. Fifty‐seven carers (black Caribbean, black African, Asian Indian, Asian Pakistani and white British) were recruited from voluntary sector organisations and a local hospital in England, and took part in semi‐structured interviews using cognitive interviewing and the critical incident technique. Interviews took place from summer 2013 to spring 2014. Thematic analysis of the interviews showed that participants often struggled to identify specific ‘incidents’, especially satisfactory ones. When describing satisfactory services, participants talked mostly about specific individuals and relationships. Unsatisfactory experiences centred on services overall. When rating services using cognitive interviewing, explicit comparisons with expectations or experiences with other services were common. Highest satisfaction ratings tended to be justified by positive personal characteristics among practitioners, trust and relationships. Lower level ratings were mostly explained by inconsistency in services, insufficient or poor care. Lowest level ratings were rare. Overall, few differences between ethnic groups were identified, although white British participants rated services higher overall giving more top ratings. White British participants also frequently took a more overall view of services, highlighting some concerns but still giving top ratings, while South Asian carers in particular focused on negative aspects of services. Together these methods provide insight into what participants mean by satisfactory and unsatisfactory services. Cognitive interviewing was more challenging for some BME participants, possibly a reflection of the meaningfulness of the concept of service satisfaction to them. Future research should include comparisons between BME and white participants’ understanding of the most positive parts of satisfaction scales and should focus on dissatisfied participants.
Users’ satisfaction with services, especially health services, has been assessed for decades and many tools are available to measure satisfaction. However, the concept is inconsistently and often poorly defined (Coulter 2005, Gill & White 2009, Bjertnaes et al. 2012) making it difficult to draw general conclusions about satisfaction and its relationships with service quality. It is generally accepted that user satisfaction is a reflection of experiences, expectations and individual preferences, which are themselves influenced by factors such as cultural norms (Coulter 2005). Satisfaction surveys are therefore regarded as an important outcome measure in health and social care (Larsson & Wilde‐Larsson 2010) and are seen as a means of measuring service quality from the users’ perspectives (Malley & Netten 2008, Malley & Fernández 2010). Although there is a considerable body of literature relating to satisfaction with healthcare, relatively little research investigates satisfaction with social care (Bauld et al. 2000, Greenwood et al. 2015) which may in part be a reflection of the conceptual challenges and the complex, multidimensional nature of social care (Malley & Fernández 2010).
Service satisfaction is measured with a variety of both standardised quantitative, often self‐completion methods and with qualitative methods. However, 20 years ago, Williams (1994) argued that most measures of satisfaction were devoid of clear definition and theoretical context, and fail to recognise the concept's complexity. This continues to be a concern and their poor psychometric properties are also highlighted (Crow et al. 2002, Gill & White 2009). The reliability and sensitivity of the tools have been questioned (e.g. Williams et al. 1998, Gill & White 2009) with scales often dependent on authors’ perception of what satisfaction means (Gilbert et al. 2004) rather than being based on patients’ experiences and values (Avis et al. 1995). Furthermore, exactly what users mean when they rate services as ‘satisfactory’ or ‘unsatisfactory’ is unclear and studies seldom attempt to determine what participants actually have in mind when describing services in these terms.
Putting issues with the concept and its measurement aside, there is evidence, especially in healthcare, that most people say they are satisfied with their care (e.g. Jenkinson et al. 2002, Bjertnaes et al. 2012). Although inconsistent, there is evidence that expressions of satisfaction vary with demographic groups. Older people frequently report greater satisfaction (Commission for Healthcare Audit and Inspection 2006), while findings for gender are more variable (Crow et al. 2002). Findings for black and minority ethnic (BME) groups are also inconsistent but several studies report that BME groups are less satisfied than majority ethnic group users (e.g. Thomas et al. 2009, Health and Social Care Information Centre 2013). Differing experiences of services may influence BME service users’ expressions of satisfaction but other factors including perceptions of what ‘satisfaction’ means and interpretations of survey questions may also account for variations in ratings.
Older people make up the largest group of social care service users and are often in receipt of long‐term personal care. An increasingly important subgroup of older users are from BME communities whose differing needs were mentioned in the National Service Framework for Older People (DH 2001) in its ‘person‐centred standard’. Assessing their experiences and satisfaction with services is important but research has largely failed to address this adequately (Bauld et al. 2000, Bridges et al. 2010). Moriarty (2008) reviewed research investigating the health and social care experiences of BME people and concluded that they received poorer treatment and are under‐represented in health and social care services. Barriers to accessing services include lack of information (e.g. lower awareness of services), differing service expectations and language difficulties. Expectations may influence how older people from different ethnic groups identify and choose help (Chahal & Temple 2005), and differing cultural responses to disability may delay help‐seeking (Bowes & Wilkinson 2003). As a result, services are often not accessed until crisis is reached after point (Moriarty 2008).
In the UK, there are over six million unpaid carers, often family carers, who provide personal and instrumental care to adults with disability (Carers UK 2012). The White paper ‘Caring for our future: reforming care and support (DH 2012) describes these carers as ‘partners’ in care, acknowledging the essential role they play and carers’ rights to assessments have recently been enhanced under the Care Act 2014. Therefore, understanding carers’ perspectives and satisfaction with services is very important, in part because there is so little research evidence here (Greenwood et al. 2015). Furthermore, even services provided for care recipients may benefit carers (Pickard 2004).
In order to improve our understanding here, this study therefore investigated satisfaction with social care services from the perspectives of carers from diverse ethnic groups. Two approaches were adopted to provide different perspectives in understanding satisfaction with services.
The critical incident technique (Flanagan 1954) is a widely used qualitative approach (Butterfield et al. 2005, Bradbury‐Jones & Tranter 2008) that has been applied in a range of service contexts including service satisfaction (Gremler 2004). Data are usually derived from participant reports of ‘critical incidents’ in interviews or questionnaires (Butterfield et al. 2005).
The technique has numerous strengths including its flexibility and provision of rich data in users’ own words from their perspectives (Gremler 2004). It is particularly useful when assessing perceptions of people from different cultures as it encourages descriptions of participants’ experiences, rather than responses to interviewer‐initiated questions. However, criticisms of the method also include insufficient clarity in its application (Butterfield et al. 2005), its retrospective nature (Gremler 2004) and the possibility that participants may have difficulty providing detailed descriptions (Edvardsson & Roos 2001 cited in Butterfield). Exactly what constitutes a ‘critical incident’ has also been debated and authors have tended to be inclusive in their approach. For example, participants’ accounts are often neither clearly demarcated nor a single incident but rather are often an amalgam of similar incidents (Norman et al. 1992).
Cognitive interviewing, where participants ‘think aloud’ while answering survey questions, has been used in developing survey questions since the 1980s. It can be described as ‘… administering draft survey questions while collecting additional verbal information about the survey responses’ (Beatty & Willis 2007, p. 288). An advantage is the relatively little interviewer training required but conversely, verbalising cognitive processes can be difficult, especially for less articulate participants (Collins 2003). Cognitive interviewing was used here to help understand how participants from differing ethnic groups interpreted questions from commonly used satisfaction questionnaires and also what they regarded as satisfactory and unsatisfactory social care.
Using cognitive interviewing and the critical incident technique, this qualitative study explores service satisfaction among carers from five ethnic groups. The strengths and weaknesses of the two methods with diverse ethnic groups are also investigated.
To be included, participants had to self‐identify as Asian Indian, Asian Pakistani, black African, black Caribbean or white British, be over 45 years and either currently or recently (last 2 years) caring for stroke survivors living in the community. They or the stroke survivor had to have used social care services in the last 2 years.
Recruitment took place from summer 2013 to spring 2014 and focussed on voluntary sector organisations and the local NHS stroke unit. After piloting, the final questionnaire contained three main sections. First, open‐ended questions were used to encourage participants to talk generally about their experiences with social care services. Using the critical incident technique, participants were then asked to describe examples of very satisfactory or unsatisfactory social care. Third, participants completed selected items from the CSQ‐8 (Larsen et al. 1979) while ‘thinking aloud’. Background demographic information including age, religion, place of birth and time in the UK was also recorded. Interviews were undertaken by two researchers (JH and TE) with experience in both in‐depth and semi‐structured interviewing. Interviews were audio‐recorded with consent, transcribed and entered into Nvivo 10.
All data were subjected to thematic analysis (Braun & Clarke 2006). Analysis of cognitive interviewing data focussed on what participants reported made services either satisfactory or unsatisfactory and how they explained their responses to closed questions. The aim was to summarise and describe the incidents using both inductive and deductive approaches (Gremler 2004). Analysis started with independent open coding and identification of themes. All transcripts were analysed independently at least twice by two or more researchers, and the process was contiguous with data collection.
Ethics approval was gained from the National Social Care Research Ethics Committee (Reference number: 12/IEC08/0003).
Fifty‐seven carers were interviewed, usually in their homes. Table 1 shows their demographic characteristics. Approximately two‐thirds were female. Most were aged over 50 years (from 45 to 91 years). Two‐thirds were supporting spouses and the remainder were mostly caring for parents. Caring duration ranged from less than 1 year to over 30 years.
In each section, overall findings are described followed by comparison between ethnic groups. Carers’ quotes are provided and all participants have been given pseudonyms.
Participants were asked to describe one very satisfactory and one very unsatisfactory experience.
Generally, rather than describing distinct incidents, participants talked very positively about specific individuals, usually care workers but occasionally social workers. Aspects of behaviour or care they regarded as very good were highlighted. Descriptions included reliable, proactive individuals with an especially beneficial impact on carers or stroke survivors who performed their job well, sometimes going above and beyond expectations. Care workers’ manner when interacting with stroke survivors was emphasised. For example, if they were particularly kind, understanding or treated the stroke survivor ‘as a person’:
She used to bend down, she used to hold Mum's hands and say ‘How's my Nell today?’ Do you think you can walk or shall I get a wheelchair? Or ‘Are you a bit wobbly today? (Dorothy, white British)
Where participants mentioned specific incidents, these generally related to individuals on isolated occasions:
… he was having a fit and the carers (care workers) really supported me, I was on my own … both carers stood by me even though it was a 15‐minute call, they were here for an hour…. (Kalyn, Asian Indian)
Good relationships were emphasised, for example where care workers joked with carers and stroke survivors or treated them like family members.
Sometimes, rather than focussing on specific individuals, participants referred to professional groups, again often care workers. Similar positive behaviours were mentioned and included professionalism, reliability and good communication, for example using care tasks as opportunities for social interaction:
…they have got to have a sense of humour, haven't they? Because you know it is pretty depressing being washed and so on … They have a chat with her and dress her and wash her … talk to her about female type things. (Valerie, black Caribbean)
Other descriptions related mostly to overall services or service providers, for example day centres or re‐ablement services providing exceptional support for stroke survivors or providing information or equipment. Services, such as day centres, that stimulated stroke survivors and were enjoyed and looked forward to were described:
The day centre … really tried to involve my Dad in things … he really enjoyed getting up and going outside the house … they gave him back a bit of himself…. (Vivienne, black Caribbean)
However, some participants did not answer the question as intended saying services were ‘fine’ or they received too little care to describe anything specific. A few said they could not identify anything particularly satisfactory or that the service did not fit their needs. Others focussed on healthcare.
Clear differences between ethnic groups were rare. However, only white British and one black Caribbean participant described specific incidents, and BME participants (Asian Indian, black African and black Caribbean) tended to highlight individual professionals.
Descriptions of unsatisfactory experiences contrasted with satisfactory descriptions. First, when describing unsatisfactory incidents, participants tended to focus on unsatisfactory services overall, rather than on individual people. Second, it appeared easier for many participants to give examples and talk in detail about unsatisfactory experiences.
Services generally were criticised for poor communication and taking too long to start:
Not getting back to you. And then when you do phone up and query, you leave a message and nobody phones you back. (Rosa, black Caribbean)
Insufficient support provision that carers thought affected stroke survivors’ dignity was highlighted:
One of the things was she's got these pads, like a nappy pad type thing and they said: ‘Well if she uses it, don't worry, we will clean her up’… We couldn't do that to Mum. (George, white British)
Care workers were criticised for frequently being late, unreliable, rushing or taking insufficient care:
Sometimes they are late … they come on the bus and that is a big problem about timings. If they call at 6:30 they'll end up here at 7:30 or 7. … you and I can go to bed whenever we want, he can't, he need somebody to take him. That gets really annoying and he gets achy. (Upma, Asian Indian)
Where an individual's behaviour was highlighted, it related to poor care:
… I think she was a bit sergeant majorly… and we said… That lady is a bit rough with her … this lady was a bit brutal and I was saying… Look, hang on, she is paralysed but she does have feelings…. (Robert, white British)
However, individuals’ behaviour was commonly attributed to insufficient time allocations or lack of training:
… but I think if people were trained properly in the first place and there were systems in place and everybody was following those systems, we wouldn't be bombarded with all this extra work we are struggling with now. (Samiya, Asian Indian)
Two participants described incidents where stroke survivors had fallen, apparently due to poor care. In one case, it was attributed to care worker inexperience, in another it related to agency policy:
… she fell over, smacked her head on the ground because nobody had thought to actually assist her. And then you find out that they are told that they are not allowed to for health and safety issues… Now that to me is nonsense. If you are working in an environment where you are caring for people, surely making sure they don't fall over is part of the…. (David, white British)
Again, some participants did not answer the question as intended and did not describe any unsatisfactory incidents, often saying the service was very good, and they had no complaints.
South Asian and white British participants gave the most detailed responses here. Asian participants appeared most critical about overall services stressing insufficient support, poor training and not addressing their needs:
…when the social (worker) came to discuss it, I did say to her that obviously bathing was a big issue for us, because it is part of our religion for sort of being clean … but she came back and said that she would have to be at the day centre all day and I said that it probably would not suit her because she doesn't speak English. (Radhika, Asian Indian)
Compared to other groups, white British participants were again more likely to describe specific incidents, rather than general concerns:
I asked to see the social services lady there at the time to ask her about taking benefits. And I can still see her there and she looked at me and she said… Don't bother because you won't get much. (Joan, white British)
Here, participants were asked to think out loud and to explain their answers to structured survey questions. All the questions had four possible responses (e.g. excellent, good, fair and poor). For simplicity, participants’ explanations and reasoning are grouped into first (highest, e.g. ‘excellent’), second, third or fourth level (lowest, e.g. ‘poor’) responses.
Participants’ ratings were concentrated in the top two levels, thus providing more information about more positive ratings. Expectations of the standard of care appeared low:
… she was perfectly safe with her because she was not going to be yelled at or shoved around. Not that I ever heard anybody yelling at her, in fact they – the other lot didn't actually talk to her. (Dorothy, white British)
The positive impact of good care on both stroke survivors and carers was emphasised (especially where carers felt relieved of caring responsibilities). Overall, individuals’ (e.g. care workers) behaviour and positive characteristics were associated with explanations for higher service ratings. Lower ratings related more to overall services.
First‐level explanations focussed on staff personal characteristics (either specific individuals or more generally), how staff interacted with stroke survivors, and trust and relationships formed. Descriptions highlighted trained, caring, kind, punctual, flexible staff who understood stroke disability, spoke directly to stroke survivors and could be trusted to care for them. Participants emphasised individualised, personalised care which valued stroke survivors as individuals. Sometimes this was only possible because of seeing the same care workers regularly. Relationships with care workers who could share jokes with carers and stroke survivors were highlighted:
I put any old clothes on her and they (the care workers) say: ‘Oh you can't possibly put that on, it's got a stain on it!’ So we have a joke about it. (Robert, white British)
Second‐level explanations included similar topics as in first‐level ratings but included caveats about less consistent behaviour or services. For example, where participants said care workers were sometimes unreliable, were mostly polite, sometimes rushed or had to be told what to do:
It's difficult to know when they're coming … if you're expecting them at half past eight and they don't come until ten … if they don't come until ten, which is very rare, I say ‘Shall I get Linda dressed?’ or, you know ‘Shall I have breakfast?’ … But that's another minor thing really. (Michael, white British)
Unreliability and lack of continuity affected both carers and stroke survivors negatively. Participants selecting second‐level ratings often justified this by saying they wanted more or improved support from services, more flexibility and more individualised care for stroke survivors:
Little things that, there's seven days in a week, they'll do it six and a half days and the other half day they wouldn't do it … no particular reason … human, human nature. …. And if you say to them ‘Oh, you didn't do so‐and‐so’ they'll come and do it, but I wouldn't bother with that. I'll just leave it and do it myself (Paulina, black Caribbean)
Third‐level explanations echoed second‐level ones but also included poor communication or wanting more services. Again the variable nature of provision was highlighted and participants were more likely to mention care workers ‘doing the minimum’, being poorly trained or not doing the job properly:
…whenever the regular carer (care worker) was on leave or sick and they sent another one. I found that most of them didn't seem to have any training. (Usha, Asian Indian)
Some participants described services as ‘better than nothing’:
… so you know that's why I appreciate it … I am slightly satisfied that at least I get this as well, rather than get nothing at all. (Sabih, Asian Indian)
Fourth‐level ratings were rare. Descriptions included poorly trained staff lacking compassion and who were not trusted but the focus here was almost entirely on service providers, for example poor communication and insufficient services:
Poor because lack of communication. No follow‐up when I've left messages. The advice is – someone listen to what I'm saying. All the problems I'm facing… ‘Stick him in a home.… Everything is Stick him in a home. (Samiya, Asian Indian)
Clear differences were rare but South Asian participants appeared to diverge from the other ethnic groups by being more likely to mention appreciating care workers from their own communities, who spoke their language and understood their culture:
Because they are speaking my language and er, kind of heart to helping. It's work but their helping is more in their nature, they have more humanity. (Chetna, Asian Indian)
South Asian participants were also more likely to say that good care workers treated stroke survivors as members of their family:
She treated her like a mother because I think she had a bond with her, that love with her. (Sadar, Asian Indian)
The only other identifiable difference was that white British participants spoke more about services meeting or exceeding expectations and care workers ‘going the extra mile’.
This section concerns how participants explained their ratings.
First level: Responses here mostly came from Asian Indian and white British participants and often involved comparisons with expectations and services actually received. Many emphasised services overall:
So initially I didn't expect the services to be that good but I'm happy now. (Dilit, Asian Indian)
I suppose now I have to say I'm ‘very satisfied’. And overall, most of the time I'm satisfied. There are bits where I wasn't, you know, particularly satisfied, but… In the overall scheme of things now, having had 10 or 12 years of it…. (David, white British)
For this carer, the care workers had done more than asked:
Because whatever I wanted, they do it so that's how… It definitely is, yes. (Kalyn, Asian Indian)
High ratings were associated with believing it was unrealistic to expect more to be done and having trust in services allowing carers to feel comfortable leaving stroke survivors.
Second level: This was the most frequently selected level. Participants tended to make explicit comparisons with top ratings, ‘excellence’ or ‘perfection’. In general, this level was chosen because the service was perceived as ‘good’ overall but not top level. Some aspects were seen as insufficient, not what was expected, and did not suit the carer and stroke survivors:
Because, excellent means no mistake, no fault. But good, you can be good, at times you can be fairly good. So I can say ‘good’. But not poor. (Abike, black African)
It was not uncommon for participants to say that services could never be perfect either because this was simply impossible or external constraints, such as the economic climate, restrict what services can afford to provide:
Well I just think that there's nothing that can be excellent, because they're working with, you know, strict budget, so they can only give so much. (Esther, black African)
Third level: Participants explicitly compared this level with those above and below. Overall services given this rating were seen as ‘better than nothing’, often insufficient, only good in parts or doing ‘most’ of what was required. Therefore, services were given this, rather than the lowest ratings:
As I said, something is better than nothing. (Abdul, Asian Pakistani)
Fourth level: Responses here were rare and usually related to when stroke survivors were unhappy with care or services were not trusted:
…the service should be all about making sure that you leave that person happy. But if you're not going to leave that person happy because you're clock watching or all you're concerned about is getting your paperwork done and going, I don't call that ‘service supply’. (Omar, Asian Indian)
Comparisons across ethnic groups were difficult because, apart from Asian Indian participants, few BME participants responded at the top level. At this first level, the Asian Indian and white British participants were superficially similar although the Asian group were more likely to mention being unable to find fault, while the white British participants mentioned small irritants but still gave top ratings. In this sense, the white British participants appeared to take a more overall perspective and to treat this category as wider than Asian Indian participants. At the second level, several BME participants specifically stated that they would virtually never give top‐level ratings either because external constraints made providing an ‘excellent’ service impossible or because assigning top‐level ratings might lead to complacency. Two BME carers compared services positively with their home country but, despite offering similar explanations, they gave very different ratings. This participant gave a third‐level rating:
I know how things are worked in India, Pakistan where they don't get these things and therefore you know I mean they have to pay for it, if you're wanting help and so on and still wouldn't get the same kind of a service as here … so therefore that's why I sort of appreciate the help I get here. (Omar, Asian Indian)
This exploratory study provides important insights into perceptions of satisfactory and unsatisfactory social care services. Expectations across all ethnic groups appeared low. Reliable, caring staff who understand carers’ and stroke survivors’ needs and who provide the agreed support might be assumed to be standard care but for many this was seen as ‘very good’ or exceptional care.
Findings from both cognitive interviewing and the critical incident technique showed that participants rating services as satisfactory more frequently appeared to be evaluating individuals, often care workers. In contrast, those rating services as unsatisfactory were more focussed on service providers. This suggests that satisfaction scales may be used to evaluate different aspects of services depending on the point in the scale selected. Scales relying on overall satisfaction ratings may therefore fail to capture this distinction. In order to understand participants’ experiences fully, separate specific questions may need to be asked about these two, often very different, aspects of experiences with services.
Our findings do not allow us to know if any facet of social care overrides any other aspects when evaluating services, but the dominance of discussions about relationships and interactions with individual practitioners suggests that these relationships with specific people were extremely important (Malley & Fernández 2010, Lewis & West 2014). Some participants highlighted poor care from social care staff but then qualified this by saying that it was probably not their fault and was probably their employers’ responsibility or due to financial constraints. In line with some other research with older people, many participants appeared to prefer to accept poor care rather than to complain (Rabiee & Glendinning 2014). This may suggest that participants value these relationships, wanting them to work and may try to shift the responsibility away to relatively distant, anonymous organisations. Blaming ‘an agency’ or the economic climate is perhaps easier than blaming individuals.
In general, it appeared easier for participants to talk about unsatisfactory, than satisfactory, social care whether as part of the critical incident technique or cognitive interviewing. This bias is important as other research has reported the apparent discrepancy between quantitative satisfaction surveys and findings from qualitative studies. Quantitative surveys usually report high satisfaction levels, while qualitative research suggests users often identify unsatisfactory aspects of services when given the opportunity with open‐ended questions (Pickard 2004). Perhaps unsatisfactory experiences are easier to describe because they appear clearer than satisfactory experiences. Certainly, participants here frequently struggled to highlight what made satisfactory services. This may be related both to their lack of clear expectations and also to how grateful many are for any support they receive, even if it is imperfect. Only when care is clearly poor (e.g. allowing someone to fall) or goes against accepted norms (such as rudeness, poor punctuality or loss dignity) do participants feel able to articulate this to researchers. However, regardless of the reasons, if services are to be improved, qualitative or mixed methods may be helpful in identifying problems or solutions.
The importance of language differences in relation to service delivery (Manthorpe & Moriarty 2009) cannot be ignored but our research also suggests that when assessing service satisfaction, the issues may go beyond translation. The researchers here had the impression that sometimes the concept of satisfaction, as opposed to language difficulties, was difficult for these BME participants most of whom had not been born in the UK. With time, as more BME users and carers are born and educated in the UK, issues of language whether in relation to service access and provision or in assessing service satisfaction, are likely to change.
During the interviews, spontaneous references to religious or culturally specific requirements or expectations were rare. Only South Asian carers highlighted these, suggesting that a more directed approach than the critical incident technique may be more useful when working with some BME groups. Similarly, if services are to be improved, opportunities should be given to explain the reasoning behind survey responses which cognitive interviewing allows. The inclusion of specific prompts about cultural and religious requirements may also be instructive.
Collins and O'Cathain (2003) suggested that services rated as satisfactory are merely ‘acceptable’ and could be improved. We would endorse this, but our findings also suggest that there may be ethnic differences in how participants use the top level ratings in survey questions. South Asian participants sometimes explicitly stated that very few services would ever receive very top rating a either because they could never be perfect or because giving top ratings would not facilitate service improvement. In contrast, white British participants often gave services top ratings, simultaneously highlighting issues with services. Future research should therefore compare, in depth, how BME participants and white British participants use satisfaction scales. Too few of our participants gave the lowest ratings to detect ethnic differences but this too deserves further exploration. Overall, however, research should focus on participants describing themselves as dissatisfied in order to improve services (Pickard 2004).
There are a number of limitations to this exploratory study. The ethnic groups included were broad and the diversity within these groups needs recognition. The fact that the South Asian participants here stood out as wanting care workers from their own communities supports the importance of always bearing the heterogeneity of these groups in mind. Further research with larger sample sizes would be better placed to comment on, for example, the possible influence of intersectionality of participant characteristics or the impact of specific factors such as religion.
It proved very difficult to recruit Asian Pakistani and black African participants, especially men. The resultant small numbers undermined ethnic comparisons. It also proved harder than anticipated to recruit participants who did not speak English. Given the issues around language and communication as barriers to accessing services, this was disappointing.
A further limitation is that only carers of stroke survivors were included in the research. However, although a relatively specific group, they are unlikely to be substantially different from other service users in relation to service satisfaction. In terms of identifying differences between ethnic groups, the fact that they were all carers of stroke survivors may have reduced the impact of ethnic group membership.
The findings here are useful for service providers developing and evaluating their services and offer insight into what those receiving services mean when they rate services as satisfactory and unsatisfactory. They suggest that reliance purely on quantitative satisfaction surveys does not offer sufficient insight into the experiences of social care services for carers from diverse ethnic groups. The positive experiences with social care services reported by many participants need highlighting but more could be done to improve services. Organisations wanting to develop and evaluate services for all social care users should employ more sensitive methods than have mostly been used previously. More open‐ended approaches and focussing on service users who specifically articulate satisfactory or unsatisfactory aspects of care may be required. BME service users appear similar to majority users in terms of their experiences but further understanding of how different ethnic groups understand the concept of service satisfaction is needed.
The views expressed in here are those of the authors and not necessarily those of the NIHR School for Social Care Research or the Department of Health/NIHR.
We thank the carers who took the time to share their experiences; the advisory group and carer advisors: Paul Chambers and Janette Gledhill; the voluntary organisations who helped with recruitment including Carers’ Support Merton, NESCAF, Afiya Trust; local support groups affiliated with the Stroke Association and Different Strokes and to Professor Jill Manthorpe for her comments on this paper. The preparation of this article was made possible by a grant from the National Institute for Health Research School for Social Care Research (NIHRSSCR).