What has employee loyalty to do with “love” to clients? Testing approaches to work as mediators

Abstract

Purpose

Drawing on the organizational psychology literature and social resource theory, this research aimed to investigate how attitude toward the employer (i.e. loyalty) and attitude toward the client (i.e. approach to work: professional, market-oriented and person-centered) relate to the perceived importance of socio-emotional resources in providing care to older people.

Design/methodology/approach

Swedish frontline care staff members participated in an electronic survey using a cross-sectional design. Mediation analyses were conducted to examine proposed direct and indirect effects of loyalty on the perceived importance of socio-emotional resources in care through three different approaches to work in care settings.

Findings

In general, the results confirmed the hypotheses. Thus, the analyses showed a positive association between employee loyalty and the perceived value of socio-emotional resources in care, which was partially mediated by the person-centered and professional approaches to work. Moreover, the analyses showed that the person-centered approach was more strongly related to the perceived value of socio-emotional resources in care than the other two approaches, lending support to the superiority of the person-centered approach in this context.

Originality/value

The study highlights that there exist multiple approaches to work in care settings. Also, the insights about how loyalty toward the employer relates to approach to work in care settings and the perceived value of socio-emotional resources in care are novel and of crucial importance to practitioners and the outcomes of care.

Citation

Kazemi, A. and Elfstrand Corlin, T. (2022), "What has employee loyalty to do with “love” to clients? Testing approaches to work as mediators", Employee Relations, Vol. 44 No. 7, pp. 149-161. https://doi.org/10.1108/ER-05-2021-0212

As we grow old our health deteriorates and our social network narrows. Receiving instrumental and emotional support from family and friends as well as from care staff has been shown to be associated with increased psychological well-being among older people (Chao, 2012). Instrumental support can be described as aid or tangible support (Cohen and Wills, 1985), while emotional support involves communicating reassurance, empathy and affirmation (Bottorff et al., 1995). Put differently, care is composed of two interrelated dimensions of task (instrumental support) and relation (emotional support) (Kazemi and Kajonius, 2015). Although provision of emotional support usually is not part of formal job descriptions of staff in elderly care services, and as such could be conceived as extra-role behavior (Van Dyne et al., 1995), previous research clearly demonstrates the crucial importance of relational factors for positive outcomes of care (e.g. Chenoweth et al., 2009; Kajonius and Kazemi, 2016). In this research we focus on the relational dimension of care work by investigating what elderly care staff members believe are the socio-emotional resources that older persons value as part of their care, in this paper referred to as socio-emotional support in care. Two types of attitudes are examined in the context of the present study as predictors of socio-emotional support in care, i.e. attitudes toward the older person (approach to work in care settings) and the employer (loyalty).

There are different approaches to work in care settings. Elfstrand Corlin and Kazemi (2019) discuss three such approaches – professional, market-oriented and person-centered – each grounded in different types of logics of care. Acknowledging differences between these approaches is crucial as they may ultimately affect how the older person is approached and what aspects of care are prioritized. Further, as loyalty has been found to have a positive influence on extra-role behavior (Cetin et al., 2015; Chen and Francesco, 2003; Feather and Rauter, 2004), we use employee loyalty as a predictor of the importance ascribed to socio-emotional support in care.

Previous research has shown that socio-emotional support (i.e. an extra-role behavior) provided by care staff members is instrumental in improving the outcomes of care in elderly care services (e.g. Chao, 2012; Kazemi and Kajonius, 2021). Given the crucial importance of this type of support in care, we need to look beyond the current dominant focus in gerontological and older people nursing literature on quality improving measures (in terms of structural resources, e.g. budget and process-related factors focusing on the quality of interactions between the staff and the older person), and embed this research within the context of work in care settings drawing on the organizational psychology literature and insights from social resource theory (stating that social behavior can be conceived as exchange of resources between individuals or between groups, Törnblom and Kazemi, 2012). Both offer important knowledge, as they help managers in care work settings to recognize that different parts of the care delivery system are connected as the links in a chain.

In conclusion, the present research poses the question of whether a positive relationship between the employer and the employee (as reflected in employee loyalty, i.e. a positive attitude toward the employer) has a positive bearing on staff's attitude toward their work (i.e. how care staff members approach their work and clientele) and the perceived importance of socio-emotional support in care (i.e. relational aspect of care). In other words, besides shedding new light on the association between employee loyalty and socio-emotional support in care, another unique feature of the present study is that alternative approaches to care work than the widely researched person-centered care are investigated. In addition to contributing to theory and research, the novel cross-fertilization between gerontological research in older people nursing, organizational psychology and social resource theory is expected to help aged care service managers gain a more nuanced understanding of how to improve the quality of care.

Theory and previous research

Loyalty

According to the Cambridge dictionary (2022), loyalty refers to “feelings of support or duty towards someone or something” (italics added). Loyalty has also been defined as a feeling of dedicated attachment, which may manifest itself as a sense of obligation to remain in a relationship in good as well as in bad times (Rundle-Thiele, 2005), a facet reminding us of the notion of commitment, lexically defined as the state or quality of being dedicated to a cause, activity or an organization (Lexico, online dictionary powered by Oxford, 2022). There seems to be a conceptual confusion in the literature discussing “this feeling of dedicated attachment” (or faithfulness, support and duty) in that what we refer to as loyalty here, most frequently, has been studied under the term commitment. Apparently, there is some overlap between the two concepts. Thus, in addition to referring to studies explicitly involving investigations of loyalty, we also draw on studies of organizational commitment in developing the rationale for our research.

Commitment appears to be a more comprehensive concept than loyalty, at least in some conceptualizations appearing in the organizational psychology literature. Thus, in a sense, loyalty could be seen as a facet of commitment, as reflected in multidimensional conceptualizations of commitment. For instance, early on, Porter et al. (1974) conceptualized organizational commitment as containing three factors: “(1) a strong belief in and acceptance of the organization's goals and values; (2) a willingness to exert considerable effort on behalf of the organization; (3) a definite desire to maintain organizational membership” (p. 604). In a similar way, Meyer and Allen (1991) conceived organizational commitment as a psychological state, reflecting the employee's relationship with the organization, which has implications for the decision to stay with or to leave the organization. Commitment was conceptualized by Meyer and Allen (1991) to be composed of three components – affective commitment (reflecting a desire to stay with an organization because of positive feelings and emotional attachment to the workplace), continuance commitment (reflecting a need to stay with an organization because of too high costs associated with leaving) and normative commitment (reflecting an obligation to stay with an organization because of an internalized loyalty). In contrast to these multidimensional views of commitment, McCaul et al. (1995) found support for commitment being a global attitude toward the organization or more specifically as “an affective and evaluative reaction toward the organization” (p. 81). Following this view, we conceive employee loyalty in the same way, that is, as an overall attitude toward the employer, reflecting the employee's relationship with the organization.

Socio-emotional resources

According to social resource theory (Törnblom and Kazemi, 2012), also called resource theory of social exchange (Foa, 1971), everything that can be exchanged between individuals or groups of individuals can be conceived as resources. All types of social behavior fall within one or more of six resource classes – love, status, information, money, goods and services. Love, status and services, which are the focus of the present study, are more particularistic in their nature, meaning that the value of the resource depends on the persons involved in the exchange. For example, you value a hug more when it comes from someone you love than from a stranger. Moreover, it takes longer time to exchange more particularistic resources than more universalistic resources (money, goods and information).

In the present study, we merged the particularistic resources of love, status and services into what we refer to as socio-emotional resources. When care staff members provide socio-emotional resources to a client, we refer to this as socio-emotional support in care – reflecting the provision of approval, esteem, affiliation and emotional support to the older person. More specifically, it reflects to what extent staff members feel that the older persons appreciate situations in which staff members “go the extra mile” , that is, make the older persons feel important and competent, show that they enjoy their company, and sometimes do something “extra” for them. Although seeing each older person as a unique individual and treating him or her with respect is part of good care, providing socio-emotional resources is often not formally required of the care staff and thus can be seen as extra-role behavior, reflecting voluntary behavior which goes beyond formal job requirements/contractual tasks and aims to benefit the organization (e.g. other employees or clients) (Van Dyne et al., 1995). As mentioned earlier, there are different sources of emotional support, elderly care staff members, however, occupy a unique position in this regard (Chao, 2012). It is thus relevant to investigate how attitudes toward the employer in terms of loyalty and attitude toward the older person in terms of approach to work relate to the perceived role of socio-emotional support in care from the perspective of the care staff.

As we are investigating the relationship between employee loyalty and the perceived importance of a behavior that can be seen as extra-role behavior, we draw on research studies showing a positive association between loyalty and extra-role behavior (Cetin et al., 2015; Chen and Francesco, 2003; Feather and Rauter, 2004). Parallels can also be drawn to service-profit chain theory, originating from the service industry, which posits that good management practices (called internal service quality) generate employee satisfaction and loyalty which in turn is expected to lead to improved external service quality (Heskett et al., 1994). In the context of care, we believe that this association translates into loyal staff being more perceptive of the older persons' need for socio-emotional support, which may add value to the care experience. Thus, we propose that:

H1.

There is a positive association between employee loyalty and the perceived importance of socio-emotional support in care.

Approaches to care and to older persons

Depending on how work is organized, there are different ways to approach work, each with its own rules and role expectations to guide staff in how they should behave in various situations, in particular, how they should view their clientele and accordingly behave toward them. These approaches to work imply different expectations of staff in terms of work ideal, decision-making and the dominant view of service users. Previous research has identified three distinct but related approaches to care work in elderly care services – the professional, the market-oriented and the person-centered approach (Elfstrand Corlin and Kazemi, 2019). We propose that these approaches may be conceived as attitudes toward the older person. Thus, different approaches may be argued to reflect different conceptions of the client (e.g. the older person) and how care could best be provided. An organization's dominant approach to work may shape staff's attitudes which in turn will affect their care behavior, but it also may be the case that the organization's dominant approach to work can influence staff's care behavior which in turn may affect their attitudes (cf. Ajzen and Albarracín, 2007). In the context of the present study, we propose staff's attitude toward the employer (i.e. loyalty) to be associated with their attitude toward the older person (i.e. approach to work), which in turn may influence the extent to which they believe socio-emotional support is valued by older persons in the process of care. Hence, we hypothesize that:

H2a.

Approach to work mediates the association between loyalty and perceived importance of socio-emotional support in care.

Professional approach

This approach prioritizes competence and ethics in providing good care. Professional competence is often based on formal education but may also refer to performing job tasks safely and skillfully (Egidius, 2011). Following this approach, care is more focused on instrumental care (tangible aid) as opposed to emotional care. Staff with professional skills may be granted extensive autonomy and is thus trusted to make their own assessments in different situations. To uphold ethical standards is central in care provision. According to this approach, older persons are viewed as clients who are entitled to the best professional treatment by the staff.

Market-oriented approach

According to the market-oriented approach, which is grounded in the principles of marketization, older persons are viewed as customers who can change provider if they are unhappy with the care they receive. Staff's main priority is to acknowledge the older persons' preferences and to satisfy their needs, mainly because of loyalty to the employer and the organization, and primarily driven by improving the reputation of the company and ultimately its profitability. The manager has the freedom, within defined legal boundaries, to manage the care unit in the way he/she feels maximizes its market share.

Person-centered approach

With a person-centered approach the older person is seen as a unique individual with unique needs and preferences, placed at the center of attention for care. Staff members know each person's life story as well as likes and dislikes. Older persons are viewed as individual service users and are supported in making decisions about their care. Thus, there is a dialogue between the staff and the older person who actively gets involved in the planning and implementation of care services. Moreover, in involving the older person in the process of care, staff members focus on the older persons' strengths rather than their weaknesses (Slater, 2006).

We expect the three approaches to work to show different patterns of association with the perceived value of socio-emotional support. The argument is that the three approaches to work seem to have differing degrees of emphasis on two distinct, albeit interrelated, aspects of care – task and relation (Kazemi and Kajonius, 2015). A task focus involves “assisting behaviours necessary to perform the expected work in a particular situation” (p. 144), and a relation focus reflects “voluntary acts of care considered as not necessary for performing the work falling outside the scope of the caregivers' formal responsibilities. These services are provided as they contribute to the overall comfort of the older person, to do ‘that little bit extra’” (p. 144). Both the professional and the market-oriented approaches tend primarily to have a task focus, that is, care services being mainly focused on the instrumental aspects. The person-centered approach, on the other hand, is a more relational approach mainly focusing on the interpersonal aspects of care (i.e. the extent to which the care staff members believe socio-emotional support is valued by older persons). Along this line of reasoning, we hypothesize that:

H2b.

Person-centered approach to work is the strongest mediator of the three approaches, accounting for the association between loyalty and perceived importance of socio-emotional support in care.

Summarizing the proposed connections

As previously stated, we propose that there is a positive association between employee loyalty and the importance ascribed to socioemotional resources in care. We believe that care staff members who have a feeling of loyalty toward their employer carry over this feeling toward the older persons, i.e. loyal staff members recognize to a larger extent the importance of socio-emotional support in care as added value.

Research has shown that loyalty can be seen as a global attitude toward the organization (McCaul et al., 1995). Further, approach to work can be seen as an attitude toward the older person and care work. Both represent attitudes from the perspective of the care staff, i.e. how staff members relate to the employer and their work. We propose that approach to work can account for the association between employee loyalty and socio-emotional support in care. Also, as the person-centered approach is relation-focused, we expect this approach to be the strongest mediator.

Material and methods

Participants and procedure

Swedish municipalities were invited to participate in a web survey for frontline elderly care staff. The municipalities were selected on the basis of their overall rank (low, medium and high) in an annual national survey in which older persons using elderly care services were asked about how they perceived the care they had received (Swedish National Board of Health and Welfare, 2014). An invitation with instructions and a link to the web survey were sent to staff members via care unit managers in the participating municipalities. Hence, a response rate could not be established. Care staff members from 12 municipalities representing all three ranking groups participated and, in total, 1,342 valid responses were registered. 93% of the participants were female, 90% were employed by the municipalities and 9% were employed by private care providers. 75% worked in nursing homes and 23% worked in home care. 8% of the participants stated compulsory education as highest educational attainment, 62% stated high school and 29% stated post-secondary education. 80% of the respondents were born in Sweden, 10% in other European countries and 10% in countries outside Europe. Mean age was 46 years, and the largest age group (27%) was 40–49 years. The Swedish Research Council's (2002) research ethics principles were followed. That is, participants were informed about the purpose of the study. They were also informed that participation was voluntary and anonymous, and that the collected data would be held confidential. Consent to participate was implied by submitting responses to the electronic survey.

Measures

Approach to work

We used a nine-item measure of approach to work constructed by Elfstrand Corlin and Kazemi (2019), distinguishing between three distinct and theoretically meaningful factors reflecting unique approaches to work. The professional approach reflected viewing the older persons as clients and prioritizing knowledge and experience in providing care. The scale included two items, e.g. “As a member of the elderly care staff, I provide good care to the older persons based on my knowledge and experience”. The market-oriented approach reflected care with a focus on acknowledging the older persons' preferences as the older persons primarily are viewed as customers. The scale included two items, e.g. “My work with the older persons is largely exercised with the aim to improve the unit's reputation among the older persons and their relatives in the municipality”. The person-centered approach reflected a focus on older persons as unique individuals who receive assistance to participate in the process of care. The scale included five items, e.g. “My work with the older persons is largely exercised based on the older persons' points of view”. All items were measured on a five-point response scale ranging from 1 (strongly disagree) to 5 (strongly agree).

Loyalty

A one-item measure assessed employee loyalty toward the employer, “How loyal are you toward your employer?”, using a five-point response scale ranging from 1 (not loyal at all) to 5 (very loyal). This item was a revised version of the item used by Coughlan (2005) (i.e. “I feel a sense of loyalty to my co-workers”) in his three-item measure of attitudinal loyalty.

We chose to use a single-item scale to measure employee loyalty as the principal intention was to assess the overall perception of loyalty, and not the various facets of loyalty. All the questions, including the loyalty item, were piloted prior to data collection, and the loyalty item prompted the overall feeling of attachment toward the employer that we aimed to assess. It has been suggested that single-item scales are appropriate to use for measuring a global attitude (cf. Scarpello and Campbell, 1983). Single-item scales have also shown to have a higher face validity than multiple-item scales (Nagy, 2002). In addition, recent studies reveal that single-item scales may perform nearly as well as multiple-item scales in other measurement respects as well (e.g. content validity, test-retest reliability, no comprehension concerns) (Bergkvist and Rossiter, 2007; Matthews et al., 2022). In a recent large-scale multi-study paper, Matthews et al. (2022) reported that single- and multi-item measures of commitment correlated substantially, suggesting that our single-item measure may be a reasonable reflection of the underlying construct of loyalty.

Socio-emotional resources

This scale reflected the extent to which the staff perceived socio-emotional resources to be important to the older persons, i.e. how important it was that staff offered supportive care to the older person. The overall question was “What do you think the older persons who you meet appreciate the most of what you do for them?” The scale included five statements, e.g. “That you show consideration in different ways” and “That you make the older person feel important”. The items were generated for the purpose of the present research following the definitions of resource classes in social resource theory provided by Foa (1971) and Törnblom and Kazemi (2012). The items were measured on a five-point response scale ranging from 1 (not important at all) to 5 (very important). We conducted a principal component analysis with Varimax rotation which rendered, as expected, a simple structure with the five statements showing positive and high loadings (ranging from 0.70 to 0.85) on a single underlying component (i.e. socio-emotional resources/socio-emotional support in care), providing preliminary evidence for the validity of the newly constructed measure. This one-factor solution, using the Kaiser's criterion of eigenvalues greater than 1, accounted for 61% of the variance.

Statistical analysis

The associations between the study variables were examined using Pearson's correlation analysis. Gignac and Szodorai (2016) suggested that r = 0.10, r = 0.20 and r = 0.30 should be seen as relatively small, typical and relatively large associations respectively in individual differences research. An online calculator (http://quantpsy.org/corrtest/corrtest2.htm) was used to test the differences between correlations. We used Cronbach's alpha to calculate the scales' reliability–professional (0.55), market-oriented (0.67), person-centered (0.81) and socio-emotional resources (0.83). Mediation analyses were performed to examine the connections between loyalty (predictor), socio-emotional resources (outcome variable) and approaches to work (mediators). We conducted all the analyses in SPSS v. 27 and used PROCESS macro v. 3.5 for SPSS, model 4 with 5,000 bootstrap samples (Hayes, 2018) for the mediation analysis. All coefficients reported are unstandardized (b) except in Figure 1, where also standardized coefficients are presented. Indirect effects were denoted ab. Significance level cut-off points were set to p < 0.01 in correlational analyses, and p < 0.001 in the mediation analysis. Significant indirect effects were inferred when zero was not included in the 95% confidence interval. The confidence intervals of small but significant indirect effects included zero when rounded to two decimal places.

Results

Pearson's correlation analysis showed different magnitudes of associations between socio-emotional resources and different approaches to work. The strongest association was found between socio-emotional resources and the person-centered approach (r = 0.36), which was significantly stronger than the corresponding correlation coefficients for the professional (r = 0.27, z = 3.08, p < 0.01) and the market-oriented (r = 0.20, z = 5.37, p < 0.001) approach to work. Thus, the more staff members had a person-centered approach toward work, the more they believed that a socio-emotional focus in care was appreciated by the older persons. Moreover, the difference between the correlation coefficients for the professional and the market-oriented approach was also significant (z = 2.28, p < 0.05) (see Table 1).

Next, we conducted a mediation analysis (Figure 1). The model featured loyalty as the independent variable, socio-emotional resources as the dependent variable and approaches to work (i.e. professional, market-oriented and person-centered) as parallel mediators. The analysis showed a positive association between loyalty and socio-emotional resources, confirming hypothesis 1. The three approaches to work together revealed a significant indirect effect (ab = 0.05, CI [0.04, 0.07]), however there was still a significant direct effect (b = 0.11, p < 0.001), reflecting a partial mediation thereby providing partial support for Hypothesis 2a. This indicates that approaches to work accounted for a significant part of the association between loyalty and socio-emotional resources, but that there are additional factors accounting for this relationship.

Both the correlation and the mediation analyses showed that the person-centered approach was more strongly related to the perceived value of socio-emotional resources than the other two approaches. The largest specific indirect effect was observed for the person-centered approach (ab = 0.04, CI [0.02, 0.05]), followed by the professional approach (ab = 0.01, CI [0.01, 0.03]), while the indirect effect through the market-oriented approach was non-significant (ab = 0.00, CI [−0.01, 0.01]) (see Table 2). The differences between the indirect effects from the three approaches to work were tested by pairwise contrasts showing a significant difference between professional and person-centered (ab diff = −0.02, CI [−0.04, −0.01]), and between market-oriented and person-centered (ab diff = −0.03, CI [−0.05, −0.02]) approaches to work, altogether confirming Hypothesis 2b.

Discussion

Drawing on the organizational psychology literature and social resource theory, we used care staff's attitude toward their employer (as reflected in employee loyalty) and their approach to work (attitude toward the clientele) to understand the perceived importance of the relational aspect in care work. Our results revealed that care staff's perceptions of the employer and the older persons shape their perceptions of the extent to which they believe that socio-emotional support is valued by the clients.

Loyalty is an important element of service-profit chain theory, according to which loyal employees are assumed to provide better service to customers (Heskett et al., 1994). In support of this reasoning, our results showed that care staff members who have a positive attitude toward their employer tend to have more positive perceptions of the value that older persons place on socio-emotional resources, i.e. the more loyal the staff members are the more they acknowledge the importance of socio-emotional support in care. Apart from loyalty, we examined different approaches to work, or attitudes toward work and clientele. We found that approaches to work explained a significant portion of the relationship between loyalty and perceived value of socio-emotional support in care, with the person-centered approach, having a more pronounced relational focus, as the strongest mediator. The mediation was partial implying that the mediators did not fully explain the relationship, and that further search for additional explanatory mechanisms is needed.

The three approaches to work – professional, market-oriented and person-centered – correlated differently with socio-emotional resources, which supports the presence and co-existence of distinct approaches to work in care settings. Not surprisingly, and reasonably expected, the person-centered approach, according to which the respect for and the dignity of the older person as a unique individual with unique needs is in focus, was more strongly correlated with the perceived importance of socio-emotional support in care compared to the professional and market-oriented approaches. That is, when you know and acknowledge someone as a unique individual, and when care is given in collaboration between the care provider/staff and the older person, you tend to be more attentive to the older person's socio-emotional needs.

Notably, loyalty predicted the endorsement of the market-oriented approach which in turn did not predict the perceived importance of socio-emotional support. Possibly, the older persons' needs being fulfilled in a market-oriented approach are more practical in nature than interpersonal, or put differently, task-oriented rather than relation-oriented. If we assume that socio-emotional aspects of care have a positive influence on older persons' satisfaction with care, and that market-oriented approach to work is most weakly correlated with use of socio-emotional resources, it presumably suggests that the market-oriented approach according to which the older person is viewed as a customer or consumer of a priced service may not be instrumental in increasing satisfaction with care.

Practical implications

The present results have managerial as well societal implications. For managers, the results suggest that it is enlightening to see the different parts of the care delivery system as the links in a chain, that is, as connected. Specifically, managers who invest in a loyal relationship with their staff may influence how they approach work and their clientele. Loyal employers tend to breed loyal employees, and loyal employees tend to be more attentive to their client's needs (i.e. focus on the relational aspect of care). That is, loyal employees tend to be more person-centered. In turn, the care experience is enhanced when care staff members provide the kind of care that the older persons appreciate. Considering socio-emotional support as an essential element of care quality, this study provides insight into how managers may improve satisfaction with care. This is a pressing issue for the welfare societies as they need to provide quality care for a growing aging population that increasingly places higher demands on the care they receive.

Limitations and future research

Employee loyalty was measured with a single item. Whether single items are appropriate to use or not has been the subject of many empirical studies. Some studies show that multi-item measures outperform single-item measures (e.g. Sarstedt and Wilczynski, 2009), whereas others show that single-item measures are correlated with multi-item measures and may perform nearly as well as multi-item measures (e.g. Bergkvist and Rossiter, 2007). Abbreviated scales and single-item measures have shown to perform satisfactorily in assessing various psychological constructs, e.g. personality (Thalmayer et al., 2011), self-efficacy (Hoeppner et al., 2011) and job satisfaction. For example, a meta-analytic study by Wanous et al. (1997) found the mean corrected correlation between single-item and multi-item measures of job satisfaction to be 0.67, and the authors estimated reliability for the single-item measure to be close to 0.70, which is deemed to be an acceptable level of reliability. In support of this, Nagy (2002) noted that face validity is another advantage of using single-item measures of job satisfaction. In a recent study, Matthews et al. (2022) investigated to which extent single-item measures reliably and validly could assess a large number of constructs in organizational research. Relevant to the context of the present research, Matthews et al. (2022) reported that single- and multi-item measures constructed to assess commitment correlated substantially, and one main conclusion was that single-item measures do not necessarily imply weak research design (i.e. non-valid measures). However, all in all, to our knowledge there are no studies of whether the single-item measure of employee loyalty used in the present study perform as well as multi-item measures. Thus, this issue should be addressed in future research investigations as single-item measures have several advantages such as lowering the costs of data collection, averting mental fatigue and increasing response rates (e.g. Bergkvist and Rossiter, 2007; Kwon and Trail, 2005).

The low reliabilities of the professional and the market-oriented approach scales need also to be addressed. A possible reason might be the few number of items included in each of those two scales. Low internal consistency attenuates (lowers) the association between two variables, while in multivariable models it may result in uncertain estimates. Thus, future research should further develop and refine the professional and the market-oriented scales.

Cross-sectional studies are primarily used to describe characteristics that exist in a group and to make inferences about possible relationships or to collect preliminary data to conduct further research studies. The temporal sequence between the independent and dependent variables cannot be established in cross-sectional studies as they are assessed at the same point in time (e.g. Field, 2013). Hence, we cannot draw any causal inferences in the present study given the cross-sectional nature of the data (e.g. Stone-Romero and Rosopa, 2008). The proposed and tested causal directions in the mediation model are therefore theoretical in nature. That is, employee loyalty is presumed to precede approach to work, which in turn is presumed to precede socio-emotional support.

Another potential limitation is that 93% of the research participants were females and this could potentially constrain the generality of the obtained results. However, this unequal gender distribution is nationally representative in that female care staff working in elderly care services in Sweden outnumber male staff by nine to one (Swedish National Board of Health and Welfare, 2019).

As the survey was administered to the care staff by the unit managers, we cannot determine the response rate. Moreover, we do not have any information about whether the staff filled out the survey during working hours using computers in the workplace and, if so, how much time they had been given to answer the survey.

Conclusions

The present study contributes to research on loyalty, job role perceptions and extra-role behavior (cf. socio-emotional support in care) in the workplace (Cetin et al., 2015; Chen and Francesco, 2003; Feather and Rauter, 2004). Using elderly care as work setting, the present study provides insight into how care staff's attitudes relate to the perceived value of socio-emotional support in care, i.e. focusing on the older person and making them feel important in the process of care, and that the person-centered approach with its pronounced relational focus stands out in relation to the alternative approaches to work in care settings.

We have a reason to believe that the present study also provided a novel and fruitful shift in research focus in the context of quality of care studies. We described multiple approaches to work in care settings, where focus previously had solely been on the person-centered care. Acknowledging differences between different approaches to work in the context of care is crucial as they affect how the client is approached and what aspects of care are prioritized, ultimately affecting the outcomes of care.

Figures

Figure 1

Mediation analyses with approaches to work as mediators of the association between loyalty and socio-emotional resources

Table 1

Means, standard deviations and correlations for study variables

	M	SD	1	2	3	4	5
1. Professional approach	4.56	0.54	0.55
2. Market-oriented approach	3.79	0.89	0.34	0.67
3. Person-centered approach	4.08	0.64	0.36	0.35	0.81
4. Loyalty	4.32	0.83	0.23	0.27	0.28	*
5. Socio-emotional resources	4.69	0.43	0.27	0.20	0.36	0.30	0.83

Note(s): M = Mean, SD = Standard deviation. Cronbach's alphas are shown in the diagonal. All correlations are significant at p < 0.01 level. ^*Single-item measure

Table 2

Indirect effects and pairwise contrasts

Indirect effects	Effect	BootSE	BootLLCI	BootULCI
Total	0.0530	0.0079	0.0388	0.0698
Professional	0.0147	0.0049	0.0062	0.0253
Market-oriented	0.0021	0.0048	−0.0072	0.0118
Person-centered	0.0363	0.0063	0.0246	0.0495
(Contrast 1)	0.0125	0.0076	−0.0017	0.0281
(Contrast 2)	−0.0216	0.0081	−0.0379	−0.0060
(Contrast 3)	−0.0342	0.0087	−0.0526	−0.0179

Note(s): Non-standardized coefficients. Contrast 1 = Professional vs Market-oriented; Contrast 2 = Professional vs Person-centered; Contrast 3 = Market-oriented vs Person-centered. BootLLCI = Bootstrapped lower level confidence interval; BootULCI = Bootstrapped upper level confidence interval