Dissertation

Welfare of adult caregivers in domiciliary and

residential services

J02024

Psychology (conversion) MSc part-time

PS7112 Research dissertation

2014/2015

University of Chester

Supervising tutor: ……………….

Word count: 12,610

Abstract

Domiciliary and residential caregiver well-being was examined in response to reports of

improper organisational support practices. Caregiver susceptibility to anxiety and

depressive symptomology were examined. Additionally, to what extent internal locus of

control, affective and cognitive empathy, perceived organisational support and self-

esteem could predict anxiety and depression within these populations was explored. A

between-subjects cross sectional design was used to compare self-report

questionnaire response data from UK domiciliary (n=51), residential (n=85) and a non-

caregiver control groups (n=67). Hospital anxiety and depression scale scores showed

that caregivers were more susceptible than non-caregivers to anxiety and depressive

symptomology. Tukey’s post hoc analysis confirmed one statistical difference between

high domiciliary and low control depression scores (p=.01). Correlation analysis

indicated one negative association between perceived organisational support and

depression. Hierarchical multiple regression analysis established low levels of internal

locus of control (p<.001) and increased affective empathy (p<.02) to be statistically

significant predictors of anxiety. Similarly, low levels of internal locus of control (p<.001)

and domiciliary occupation type (p<.001) were found to be statistically significant

predictors of depression. However, findings must be interpreted with caution due to the

small size and distribution range of the questionnaires. Furthermore, anxiety and

depression scores were below clinical range. These, findings do not fully substantiate

an association between the impact on caregiver well-being and the perception of

negative organisational practices. A questionnaire designed from focus group

outcomes of caregiver organisational support perceptions may be a more accurate

method to access the impact poor organisation support has on caregiver well-being.

Introduction

The consequence of outsourcing publically funded social care to the

independent sector is the competition between financially orientated residential (res)

and domiciliary (dom) providers who ultimately drive out the good providers (Koehler,

2014). As such, the consequence of poor care will be paid for by the elderly, the

vulnerable, and by the caregivers they depend on (CQC, 2014; Koehler, 2014). When

considering the projected increase in need of care due to greater life expectancy, better

health care, and rise in population (National Audit Office, 2014) employers have an

operational and moral responsibility to safeguard their staffs’ psychological well-being.

Previous res and dom caregiver health and well-being research focused the

requirement for staff retention and financial and practical support form organisations to

maintain the health and social care needs for those they care for (Castle & Engberg,

2007; Chester, Hughes & Challis, 2014; Colton & Roberts, 2007; Denton et al., 2002;

Eustis, Kane & Fischer, 1993; Hunter, Ward & Camp, 2012; Wibberly, 2013). However,

the challenge to implement change appears to remain difficult in light of recent think

tank (Franklin, 2014; Koehler, 2014), National audit office (National Audit Office, 2014),

The National Care Forum (The National Care Forum, 2015), employer-led workforce

development bodies (Skills for Care, 2015) care quality inspectorates (CQC, 2013,

2014; CSSIW, 2014) and increasing media reports (for example, Daly, 2008; Farooq,

2009; Triggle, 2014, 2015). Perhaps by redirecting the focus of this research to

ascertaining imbalances in caregiver psychological well-being, poor service providers

may re-access their stance on negative operational practices.

Definitions

Characteristically, a res caregiver will provide a number or person centred

social, personal and domestic care duties to a variety of vulnerable individuals (for

example, learning disabilities, elderly, dementia and mental health) in a single place of

work generally following a regular shift pattern. A dom caregiver undertakes similar

duties to a variety of care receivers’ with similar needs over the course of a day but

within their own homes to maintain independence and quality of life. A dom caregiver

will usually work irregular hours, irregular shift patterns and is required to provide their

own transport between calls. Demographically, females account for 83% of the dom

care workforce. The average age of a dom caregiver is forty two. Furthermore, 37% are

understood to work part-time with 23% on a zero-hours contract (Skills for Care, 2015).

These figures are comparable to prior research (Zeytinoglu, Denton, Davies &

Plenderleith, 2009) as is the disproportion in gender and age (Crown, Ahlburg, &

McAdam, 1995), indicating consistency in demographics over time that are also

comparable to res caregiver figures (CQC, 2014; Skills for Care, 2015). Recent data

exposed the comparable rates in staff turnover between these two groups (CQC, 2014;

Koehler, 2014; National Audit Office, 2014; The National Care Forum, 2015; Skills For

Care, 2015). Of the three hundred thousand social care workers leaving their role each

year, this data indicates that adult dom caregivers account for 30.6% of leavers

compared to 24.2% adult res caregivers (Skills For Care, 2015). It is estimated that

whilst 42% continue to work in this sector 58% leave altogether. (Skills For Care,

2015). Such similarities between the two groups make them ideal for comparative

analysis. Likewise, the addition of a non-caregiver group (control group) as used in

previous well-being research (Butterworth, Pymont, Rodgers, Windsor & Antsey, 2010),

provides a reliable baseline with which to compare the two groups.

Review of the literature

A systematic review of the literature suggests that anxiety and depression

appear to be the most common consequences of caring for a vulnerable adult (Blanco,

Rohde, Vázquez & Torres, 2014; del-Pino‐Casado, Pérez‐Cruz, & Frías‐Osuna, 2014;

Otero, Smit, Cuijpers, Torres, Blanco & Vázquez, 2014; Rodakowski, Skidmore,

Rogers, & Schulz, 2012). Both reported to have greater impact caregivers than non-

caregivers (Butterworth et al., 2010) with a focus on the research based around

caregivers of the frail and elderly (Aggar, Ronaldson, & Cameron, 2011; Phillips,

Gallagher, Hunt, Der, & Carroll, 2009; Silva, Teixeira, Teixeira & Freitas, 2013),

dementia caregivers (Edwards, Zarit, Stephens & Townsend, 2002; Hatch, DeHart,

Norton, 2014; Molyneux, McCarthy, McEniff, Cryan & Conroy, 2008) learning disability

(LD) caregivers (Gray-Stanley, Muramatsu, Heller, Hughes, Johnson, & Ramirez-

Valles, 2010; Li, Shaffer and Bagger, 2015; Minnes, Woodford, Passey, 2007; Mutkins,

Brown & Thorsteinsson, 2011; Taggart, Truesdale-Kennedy, Ryan & McConkey, 2012)

and some research based on mental health (MH) caregivers (Cummings & Kropf,

2015; Taylor & Barling, 2004). Such consequences are also recognised cross culturally

(Rezende, Coimbra, Costallat & Coimbra, 2010; Wang, Shyu, Chen & Yang, 2011).

This body of research suggests that regardless of the caregiver’s health or disability

status, it is the role of having to care for vulnerable adults that cause anxiety and

depressive symptomology. As such, research indicates that there is great relatability

between anxiety and depression. Notably, that one will usually predict the other (Aggar

et al., 2011; Phillips et al., 2009).

However most of this research focusses on paid MH caregivers and unpaid

family caregivers. The impact on family caregivers’ psychological well-being can be

interpreted by their own individual means to find a balance between life, caregiving and

work (Wang et al., 2011). Contrarily, Edwards et al. (2010) found there to be no

difference in role overload, worry, strain, and depression between employed and non-

employed caregivers. Suggesting that employed and family caregivers can be

susceptible to similar psychological imbalances regardless of the caregiver’s personal

or professional relationship.

Psychological variables

Perceived organisational support (POS). POS relates to the value of the

employee employer relationship. Here, the employee recognises their worth in relation

to the organisations operational goal, enabling the caregiver to proficiently complete

their job roles and to effectively manage demanding circumstances (George, Reed,

Ballard, Colin, & Fielding, 1993). Similarly, organisational support theory stipulates that

employees must develop a belief that the organisation values their contribution and

well-being before the organisation will recognise and subsequently reward them

(Eisenberger, Huntington, Hutchison, & Sowa, 1986). A positive reciprocal relationship

between employee and employer is believed to satisfy the socioemotional need of the

employee in terms of esteem, affiliation and approval (Rhoades & Eisenberg, 2002). An

absence of this in residential services has led to reductions in job satisfaction, positive

mood, affective commitment and a performance increasing staff turnover (Castle &

Engberg, 2007; Colton & Roberts, 2007; Rhoades & Eisenberg, 2002). Similarly,

research has found that low organisational support and depressive symptoms to be

associated with burnout in LD services (Mutkins et al., 2011). Not just related to private

companies, compared to charity home care staff, home care NHS staff believe that

organisations had unrealistic expectations and perceived that their views are not

listened to (Blumenthal, Lavender & Hewson, 1998).

Quality assurance research indicates that there are a number of fundamental

pre-requisites required from caregivers by their organisations in order to deliver

appropriate care. For example: appropriate training, adequate pay, good

communication, contact with peers and supervisors, supporting documentation and

information, supervision, and clear accountability and authority (Eustis et al., 1993).

Results from a meta-analysis of 70 studies concerning POS found that fair treatment,

favourable job conditions, supervisor support and rewards were significantly related to

POS (Rhoades & Eisenberg, 2002). However, personality and demographic

characteristics were weakly associated. Equally, subcategories of organisational

rewards found that job stressors and job security had a strong relationship with POS.

As such employees with a high POS commonly suffer fewer strain symptoms including

burnout, anxiety headaches and fatigue as well as job satisfaction (Rhoades &

Eisenberg, 2002).

It would appear beneficial to the welfare of the caregiver to adhere to these pre-

requisites. However a vast number of reports do not reflect this. Concurrent, figures

indicate a disproportionally high use of zero-hour contracts amongst res and dom

caregivers (Skills for Care, 2015). Whilst they offer employers more flexibility to

respond to the fluctuation in care demand (CSSIW, 2014) they simultaneously impact

on job insecurity destabilising caregivers’ lives and incomes (Koehler, 2014). Job

insecurity has affected job competition in MH services leading to burnout and work

related stress exacerbated by poor management and inadequate resources (Taylor &

Barling, 2004). Contrarily research has indicated that zero-hour contracts suit some

caregiver’s lifestyles (Eustis & Fischer, 1991). Growing reports reflect the inequality in

pay particularly amongst dom caregivers met attributable to a lack of fuel and uniform

allowance (Koehler, 2014; National Audit Office, 2014; Skills for Care, 2015) known to

impact well-being (Denton et al., 2014). Furthermore, non-specified sick pay and

payment fort training within social care contract documentation (Chester et al., 2014)

could further impact on financial burden and sense of feeling devalued by the

employer.

Likewise, training and work related qualifications within dom and res services

are low (Franklin, 2014; Colton & Roberts, 2007; Skills for Care, 2015). Whilst training

has led to caregiver positivity, simultaneously impacting on turnover rates seen in a

residential study (Hunter et al,, 2012), its absence will undoubtedly risk both caregivers

and care receiver’s well-being when encountering challenging behaviours (Denton et

al., 2002; Rose, Mills, Silva, Thompson, 2013). This is notable in cases of dementia

(Hatch et al., 2014; Molyneux et al., 2008), LD (Minnes, et al., 2007; Li, Shaffer &

Bagger, 2015) and MH (Taylor & Barling, 2004). By reinvesting back in staff training

caregivers will be more equipped to deal with challenging behaviour (Denton et al.,

2002) and reinforce POS (Shore & Shore, 1995). Similarly, working in dirty and unsafe

homes due to insufficient times to clean has been linked to the devaluation of caregiver

work and risking well-being (Wibberly, 2013), believed attributable to organisations

purposely cutting down times between clients to increase profit (Daly, 2008; Farooq,

2009; National Audit Office, 2014). Subsequent time pressures to complete roles are

shown to lead to stress (Michelson & Tepperman, 2003; Taylor & Barling, 2004) and

are associated with depression and an inability to control one’s life events (Gray-

Stanley et al., 2010). However it is important to understand that this is not the status

quo and there are numerous accounts of good service provision in the UK (CQC, 2013,

2014).

A review of the literature has highlighted the requirement for fair treatment and

retention of quality caregivers through policy change (Chester et al., 2014; Denton et

al., 2002; Eustis et al., 1993; Franklin, 2014; Koehler, 2014; Skills for Care, 2015). As

such, social, legal and financial intervention support strategies (Hatch et al., 2014; Silva

et al., 2013) have helped to reduce emotional stress (Otero et al., 2014), physical

fatigue (Phillips et al., 2009) and caregiving resentment (Aggar et al., 2011) amongst

informal caregivers.

Locus of control. Locus of control (LOC) revolves around an individual’s

internal and external belief in which they can control the outcomes of life events

(Rotter, 1966). Commonly, an internal LOC relates to the belief that a person can

influence their own outcomes and are mostly in control of what happens to them. An

external LOC is the belief that an individual is not in control of live events subsequently

perceiving negative consequences as a result of luck or chance that is out of their

control (Gray-Stanley et al., 2010; Rotter, 1966). Internality of control has been related

to coping strategies that manage stress (Hay & Diehl, 2010) and increased self-esteem

(Anderson, 1998) which in turn relates to psychological adjustment (Elfström & Kreuter,

2006). Those who subscribe to external LOC endorse higher levels of stress, illness

and lower levels of self-efficacy (Roddenberry & Renk, 2010). As such external LOC is

associated with depression (Presson & Benassi, 1996) and anxiety (Molinari & Khanna,

1981).

In relation to the caregiving context, a study by Chan (2000) found those with

an internal LOC to have greater life satisfaction, less caregiver burden, low levels of

depression and greater role satisfaction due to greater coping strategies. Similarly,

higher internal LOC scores lead to a better quality of life and fewer depressive

symptoms (Gibson et al, 2013). It is said that those overoptimistic individuals purposely

attribute negative life events to external outcomes to as a method of protection to

maintain control (Seligman, 1991), suggesting that there is some interchangeability

between the two concepts (Lefcourt, 1991). As such it is plausible to assume that low

levels of internal control can impact on psychological well-being dependent on the

situation. Current negative caring cultures deter autonomy of care, career progression

and maintain a status quo of low status (Koehler, 2014) controlled by a lack of pay,

poor management and a lack of training (Eustis et al., 1993; Koehler, 2014). This

cyclical nature known to impact on job satisfaction, stress, self-esteem and resentment

(Colton & Roberts, 2007; Denton et al., 2002; Taylor & Barling, 2004; Michelson &

Tepperman, 2003) will undoubtedly impact on the perception that one believes they are

in control of their lives. However, the perception of some caregivers’ situation may not

be related to fate or chance but to an acknowledgement that they are not being

supported appropriately. This reflects the work by Lefcourt, Martin, & Saleh (1984) who

believed that those with an internal LOC compared to those with an external LOC

benefit from social interactions and support, suggesting that this is a possible coping

method to relieve stress. As such internal LOC has been associated with greater social

support that was significantly correlated with better psychological health and lower

levels of perceived stress in the caregiving context (Pelletier, Alfano & Fink, 1994). In

this instance it is plausible to assume that low levels of internal control can relate to low

levels of psychological well-being.

Self-esteem. Self-esteem is a universal assessment of self-acceptance, self-

satisfaction, self-respect, and self-worth (Bowling, 2005). Stability in self-esteem is

related to reduced mood swings (Campbell, Chew, & Scratchley, 1991) and happiness

(Galambos, Fang, Krahn, Johnson, & Lachman, 2015). Also, high and low levels of

self-esteem has found to impact significantly on success and failures in life domains

(Orth, Robins & Widman, 2012). Similarly, the absence of self-esteem is related to a

number of psychological symptoms including depression (Riketta, 2004; Sowislo, Orth

& Meier, 2014; Zunzunegui, Llácer, & Béland, 2002) and anxiety (Lee & Hankin, 2009;

Riketta, 2004). A meta-analysis of the available literature suggests that the effect of low

self-esteem on depression in robust across time, samples and characteristics (Sowislo

& Orth, 2013) seemingly related to the exposure of stressful events (Orth, Robins, &

Meier, 2009) and daily hassles (Kernis et al, 1998).

Characteristically, greater self-esteem levels have been attributed to women

with a higher educational status than those who have not, who had embraced religion

in childhood and who have a greater sense of LOC in the caregiving context (Moen,

Robinson & Dempster-McClain, 1995). Furthermore, Moen et al. (1995) also found that

those women who had life satisfaction and self-esteem prior to obtaining a caregiving

role managed to retain self-esteem and well-being. High self-esteem is important in the

provision of care because it is essential to the reciprocity of the caregiver and caregiver

relationship (Kramer, 1997; Lopez, Lopez-Arrieta & Crespo, 2005; Neufeld & Harrison,

1995). Maintenance of self-esteem is particularly important when dealing with difficult

and stressful situations related to the complex need of passive or aggressive care

receivers (Colton & Roberts, 2007; Molyneux, et al., 2008; Rose et al., 2013). In order

to maintain self-esteem organisations must provide holistic support to their caregivers

as suggest by a recent think tank report: self-esteem is important to the caregiver

workforce suggesting that in order to provide compassionate care; greater training, free

influenza vaccinations, key-worker status, working carer tax credits, appropriate pay,

licence to practise for all homecare workers and the promotion of career pathways are

required (Koehler, 2014). Currently, res and dom research suggest that there is a lack

of such operational and social organisational support (Colton & Roberts, 2007; Denton

et al., 2002; Mutkins et al., 2011). As such, research suggests that those whose self-

esteem is dependent on workplace performance are likely to feel ostracised in the face

of poor performance (Lance Ferris, Lian, Brown & Morrison, 2015). Equally, self-

esteem has been shown to moderate workplace conflict and role strain (Dijkstra,

Beersman & Cornelissen, 2012), thus having a direct impact on the ability to provide

compassionate care for the caregivers. Research has shown that by emphasising

mutual respect (Lyons, Sayer, Archbold, Hornbrook, & Stewart, 2007) and

readdressing the imbalances in caregiver and care receiver relationship can strengthen

caregiver resolve and resilience leading to greater levels of satisfaction, reward and

happiness that ultimately reinforces and strengthens the bond between caregiver and

care receiver (Brouwer, van Exel, van den Berg, van den Bos, & Koopmanschap, 2005;

Chappell & Dujela, 2008; Cohen, Colantonio, & Vernich, 2002; Kramer, 1997).

However, without readdressing negative organisational practices this will only

exacerbate caregiver MH deficits (Denton et al., 2002; Eustis et al., 1993).

Empathy. A modern day conception of empathy is based around two

components: affective empathy and cognitive empathy (Davis, 1983; Jolliffe &

Farrington, 2006) measured by the basic empathy scales (BES) (Jolliffe & Farrington,

2006). As such, empathy is recognised as an automotive response as well as an

interpersonal attentive cognitive reaction to the emotions and mental states of other

individuals (Carré, Stefaniak, D’Ambrosio, Bensalah & Besche-Richard, 2013).

Recognising that the scale was initially designed for adolescents, through its

development and validation Jolliffe and Farrington (2006) found that females scored

higher in both empathy components through particularly higher in affective empathy.

Similarly, females were more likely to respond prosocially to another individual’s

distress compared to males. Whilst it is recognised that a residential or dom caregiver

role is not financially rewarding the choice of a caring career pathway may instead be

prosocially incentivised noting the large disproportion of male caregivers to female

caregivers (Skills for Care, 2015). As such, in a caring context empathy is essential to

producing the right socioemotional therapeutic response (for example, comforting the

individual when they are upset or proving sensitive personal care not leaving it for

someone else to do) interpreting the care receivers’’ emotional and contextual state.

Well supported caregivers experience great satisfaction and reward from their

caregiving roles (Hatch et al, 2014; Silva et al., 2013) many individuals that become

report enjoyment, companionship and others find a great sense of reward, fulfilment

and enjoyment out of being able to provide a quality of life (Cohen et al., 2002). The

reciprocal nature of this relationship is based on self-awareness, non-judgmental

approach requiring a mutual regard (Davis, 1990).

Austin, Globe, Leier and Bryne (2009) compared a reduction in empathy to a

term called ‘compassion fatigue’ where by one’s ability to empathise with another’s

psychological distress is hampered by redirecting the caregivers focus on themselves

attributed to psychological issues including burnout, stress and emotional contagion.

Similarly, barriers to relive another’s emotional distress (for example, less time with the

care receiver) may induce a corresponding response. For example, it may impact on

caregivers’ emotional state including anxious feelings or tension at their inability to

perform their job. However, this may be interpreted as a sympathetic not empathic

response if the caregiver cannot transpose themselves into the care receiver’s position

(Davis, 1990). As such, research appeals for caregivers to participate in perspective

taking to influence one’s interpretation of what another person is experiencing

(Lobchuk, 2006).

Factors related to the maintenance of empathy include, patient recovery,

professional support, personal support, professional resources and compassion

satisfaction (Austin et al., 2009). As discussed res and dom are equally susceptible to

burnout (Blumenthal et al., 1998), stress (Gray-Stanley et al., 2010) and other

psychological symptomology due to negative organisational practices (Mutkins et al.,

2011) and challenging nature of the work (Rose et al., 2013; Wibberly, 2013). Whilst

most empathy research is based around nurses and their care receivers (Austin et al.,

2009; Lobchuk, 2006), it is plausible to assume that res and dom caregivers

experience reduction in empathy levels due to similar roles and responsibilities.

However, a review of the research has not found any substantial accountancy of

empathy to anxiety or depression. Given the impact POS, LOC, self-esteem and

empathy can have on caregivers and subsequently their care receivers. It is important

to explore their relationship with anxiety and depression amongst res and dom

caregivers due to the comparability in roles and responsibilities between, NHS staff,

MH workers and informal caregivers (Eustis et al., 1993).

Aims

Due to the increase in negative reports surrounding the poor organisational

support within res and dom services (for example, CQC, 2014; Franklin, 2014; Farooq,

2009; Koehler, 2014) that fail to meet the quality assurance standards proposed by

Eustis et al. (1993), known to impact on caregivers’ psychological well-being (for

example, Denton et al., 2002; Gray-Stanley et al., 2010), this research had three aims.

Firstly, to add to the small body of psychological well-being literature surrounding res

and dom caregivers (for example, Bartoldus, Gillery & Sturges, 1989; Colton &

Roberts, 2007; Denton et al., 2002; Hunter et al., 2012; Wibberley, 2013). The second

aim was to ascertain associations between caregivers’ perception of organisational

support and the possible impact on psychological well-being. Thirdly, that by redirecting

caregiver well-being research towards a psychological focus this study will

optimistically inspire future research to raise public, caring organisation and policy

makers awareness’ about the possible need to change caring organisations standards.

Hypotheses

Based on previous research, it was hypothesised that domiciliary caregivers will

experience greater levels of anxiety and depression than residential caregivers and the

control group due to slightly raised levels of turnover rates (Skills for Care, 2015),

negative impacts of irregular hours and the abundance of zero-hour contracts,

decreased time to care for care receives due to additional responsibilities, lower levels

of pay due to additional responsibilities like keeping the car on the road (Denton et al.,

2002; Franklin, 2014; Koehler, 2014; National Audit office, 2014; Wibberley, 2013) and

less benefits and pay than those working in an institutional setting (Crown et al., 1995).

Secondly it was hypothesised that low levels of internal LOC, low levels of

affective empathy, low levels of cognitive empathy, low levels of self-esteem and low

levels of POS will predict variance in anxiety and depression symptomology. All

variables were entered independently into the hierarchical multiple regression, in this

order, to ascertain their individual contribution over and above the other. However,

affective and cognitive empathy were entered together in the same step as research

has yet to define their individual contribution to anxiety and depression. POS was

entered after all of the variables to see if it had any affect above and beyond the other

psychological variables because of its assumed relatability to them. Occupation was

entered last into the regression to see if it had any overall affect above and beyond all

other variables.

Method

Participants

Caregiver data was collected from res and dom caregivers working in the North

West and the South of the UK currently caring for elderly adults, adults dementia,

adults with LD and adults with MH disabilities. Some care receivers were comorbid and

others also had a physical disability. Control group data was not regionalised but were

contacted through Facebook (FB), a social networking site, and university research

site. Total number of participants was 203 comprised of 51 dom caregivers, 85 res

caregivers and 67 non-caregivers. Ethnicity was not accounted for. Participation was

voluntary and participants were prior warned in the information sheet that if they

suffered from a pre-existing or present psychological condition, including depression

and anxiety, that taking part was at their own discretion (see Appendix 1, 2 and 3). Age

was restricted to eighteen and above. This was controlled for by sending a private

invitation to take part in the electronic copy of the questionnaire and was a stipulation

written in to the paper questionnaire.

Domiciliary caregiver demographic information. Of the 51 caregivers 5

(9.8%) were male, 43 (84.3%) were female and 3 (5.9%) were not disclosed.

Caregivers’ ages ranged from 19 to 77. The minimum and maximum length of time in

current caring position was 1 month and 39 years, respectively. The minimum and

maximum length of time working within the caring profession was 4 months and 39

years, respectively. The number of dom caregivers providing care for adults with LD

was 22, elderly was 34, dementia was 31, physical disabilities were 5 and MH was 1.

Physical disabilities and MH were comorbid with another social care requirement;

likewise participants caring form more than one care requirement was 32. The number

of dom caregivers employed full time were 25 (49%). With 17 (33%) working part time,

8 (15.7%) did not disclose. The number of dom caregivers working regular hours were

24 (47.1) with 16 (31.4%) working irregular hours, 11 (21.6%) did not disclose. The

number of domiciliary caregivers currently caring for a relative or close friend was 5

(9.8%), 37 (72.5%) did not and 9 (17.6%) did not disclose. Whilst 20 (39.2%) dom

caregivers had previously cared for a relative or close friend, 24 (47.1%) had not, 7

(13.7%) participants did not disclose. Of the 51 domiciliary caregivers 17 reported

being single, 27 reported being married/co-habiting, 5 were divorced/separated and 1

reported being a widow/er (see supplementary CD).

Residential caregiver demographic information. Of the 85 res caregivers 16

(18.8%) were male, 65 (76.5%) were female and 4 (4.7%) did not disclose. Ages

ranged from 19 to 71 and the minimum and maximum length of time in current caring

position was 1 month and 22 years, respectively. The minimum and maximum length

of time working within the caring profession was 3 months and 30 years, respectively

with a median average of 10 years. The number of participants caring for learning

disabilities was 68, elderly was 27, dementia was 27, physical disabilities were 10 and

mental health was 1. Physical disabilities and mental health were comorbid with

another social care requirement; likewise participants caring form more than one care

requirement was 36. The number of res participants in full-time employment was 63

(74.1). With 19 (22.4%) working part-time, 3 (3.5%) did not disclose. The number of res

caregivers working regular hours was 54 (63.5%). With 25 (29.4) working irregular

hours, 6 (7.1%) did not disclose. The number of res caregivers currently caring for a

relative or close friend was 10 (11.8%). Whilst 74 (87.1%) were not, 1 (1.2%) did not

disclose. Those previously cared for a relative or close friend was 31 (36.5%). Whilst

53 (62.4%) had not, 1 (1.2%) did not disclose. Of the 85 res caregivers 29 reported

being single, 48 reported being married/co-habiting, 6 were divorced/separated and 2

reported being a widow/er (see supplementary CD).

Control group demographic information. Of the 67 control participants 33

(49.3%) were male and 34 (50.7%) were female. Age ranged from 19 to 61.

Occupational type was not accounted for but 64 participants were students, 13 were in

employment, 12 of which were also students. Of the 67 control participants 24 were

single, 40 were married/co-habiting and 3 were divorced/separated (see supplementary

CD).

Ethical considerations

Prior to data collection this project was reviewed and passed by the University

of Chester, Psychology Department Research Ethics Committee, which conforms to

the ethical principles laid down by the British Psychological Society (see Appendix 4).

Additional amendment forms were sent for ethical approval relating to amendments to

the initial ethical approval for form (see Appendix 5) further amendments to study

including organisational consent to participate in research (see Appendix 6) and further

organisational consent to participate in the research (see Appendix 7, 8 and 9).

Measures

The hospital anxiety and depression scale (HADS). The HADS (Zigmond &

Snaith, 1983) is one of the most commonly used self-report questionnaires for the

study of anxiety and depression (DVs). It is a 4-point type Likert (0-3) type scale

consisting of 14 question divided evenly between the two factors (anxiety and

depression). HADS scoring assigns higher scores to indicate greater levels of anxiety

and depression. Scoring between 0 and 7 is considered normal, scoring between 8 and

10 is borderline abnormal and 11 and 21 is abnormal. An example of an anxiety

question was related to the whether an individual felt they had to be on the move due

to restlessness. Similarly, an example depression question was related to whether an

individual felt they had no need to maintain their appearance (Zigmond & Snaith,

1983). Prior research has demonstrated it to be robust in screening separate

dimensions of depression and anxiety in the general population, during general

practice and among psychiatric patients (Bjelland, Dahl, Haug & Neckelmann, 2002). It

has also construct and re-test validity with high internal subscale consistency and has

found to be reliable cross culturally (Reda, 2011). A high level of internal consistency of

the anxiety subscale α=.75 (see Appendix 10), depression subscale α=.76 (see

Appendix 11) and combined HADS α=.82 was duplicated using Cronbach’s alpha (see

Appendix 12).

The perceived organisational support (POS) scale. The eight item POS

scale was developed to be more user friendly (Rhoades, Eisenberger & Armeli, 2001)

from the original 40 items (Eisenberger, Huntington, Hutchinson & Sowa, 1986) so as

not to disinterested participants. As such it was designed to quantify POS in the work

place. The questionnaire is scored on a seven point Likert scale from 0 – 6 (strongly

disagree- strongly agree) with negative questions requiring reversing (2, 3, 5 and 7).

Rated on a scale from lowest to highest (more negative to a more positive response)

with an average of score of 24, higher scores are attributed to greater levels of POS.

“The organization would ignore a complaint from me” (Eisenberger et al., 1986, p. 502),

is an example question from the questionnaire. Whilst “the SPOS measures with high

reliability” (Rhoades & Eisenberger, 2002, p. 699), to diversify between subject groups

the seventeen highest factor loading items from the original 40 appear to be the most

prevalent in POS studies (for example, Cleveland & Shore, 1992; Randall & O’Driscoll,

1997). As such the use of the 8 question version appears not to be problematic and is

still reflective of the well-being of the organisations employees (Rhoades &

Eisenberger, 2002). Cronbach’s alpha scores were calculated at α=.90 (see Appendix

13) indicating a high level of internal consistency.

The locus of control (LOC) scale. LOC is a scale designed to measure the

“the extent to which one regards one’s life chances as being under one’s control in

contrast to being fatalistically ruled” (Pearlin & Scholer, 1978, p. 5). “There is little I can

do change many of the important things in my life” (Pearlin & Scholer, 1978, p. 20) is

an example of a question from this scale. The self-report questionnaire consists of 7

items scored on a 4 point Likert scale from 1-4 (strongly agree – strongly disagree).

Positive questions (6 and 7) are reverse scored. Rated on a scale from lowest to

highest (more negative to a more positive response) with an average scored between

17 and 18 higher scores are related to a greater LOC. Developed for working adults

between the age of 18 and 65, researchers have used it in other populations including

adolescents and with mental health difficulties as well as being translated into

numerous languages (Brady, 2003). Cronbach’s alpha scores were calculated at α=.82

(see Appendix 14) indicating a high level of internal consistency.

The basic empathy scale (BES). The BES is a two factor 20 item scale

developed by Jolliffe and Farrington (2006) to measure empathy from two subscale

factors (affective empathy and cognitive empathy). “I get caught up in other people’s

feelings easily” (Jolliffe & Farrington, 2006, p. 593) is an example of an affective

question, ‘‘It is hard for me to understand when my friends are sad’’ (Jolliffe &

Farrington, 2006, p. 593) is an example of a cognitive question. Both affective and

cognitive states are essential to emotional functioning requiring the recognition of other

people’s feelings as well as their own. The affective empathy subscale is comprised of

11 items scored on a 5 point Likert scale from 1-5 (strongly disagree-strongly agree)

with an average score of 33. The cognitive empathy subscale is comprised of 9 items

scored on a 5 point Likert scale from 1-5 (strongly disagree-strongly agree) with an

average of 23. Below average scores are attributed to having less affective/cognitive

empathy and above average scores are attributed to having high levels of

affective/cognitive empathy. Included were 8 reverse scored items (1, 6, 7, 8, 13, 18,

19 and 20). Whilst the BES had been specifically designed for adolescents and has

now been adapted for adults (BES-A) showing some promising reliability (Carré et al.,

2013), the original BES was still chosen for this research. This decision was related to

the vast amount of positive attention it has received including satisfactory internal

consistency and moderate retest reliability also it has shown to be valid cross culturally

(D’Ambrosio et al., 2009; Geng et al., 2012; Sánchez-Pérez, Fuentes, Jolliffe &

González-Salinas, 2014).

Consent of BES

Consent was obtained from the authors prior to its implementation (see

Appendix 15). A moderate level of internal consistency of the affective empathy

subscale α=.72 (see Appendix 16), cognitive empathy subscale α=.74 (Appendix 17)

and combined BES α=.77 was found using Cronbach’s alpha (Appendix 18).

The Rosenberg self-esteem scale (RSES). The RSES (Rosenberg, 1965) is

one of the most widely used measures of self-esteem, popular for its succinctness (10

items) and its uncomplicated language “I take a positive attitude towards myself”

(Rosenberg, 1965, cited in Harvey & Keashly, 2003, p. 810). The RES is scored on a 4

point Likert scale 1-4 (strongly disagree-strongly agree) with an average of 25. Higher

scores correspond to greater levels of self-esteem with 5 items require reverse scoring.

Reliability and validity for the scale has been assumed among a number of groups

including nursing assistants (McMullen & Resnik, 2013). It is also seen to be stable

over time (Marsh, Scalas & Nagengast, 2010) and valid cross culturally (Meurer, Luft,

Benedetti & Mazo, 2012). Cronbach’s alpha scores were calculated at α=.87 (appendix

19) indicating a high level of internal consistency.

Materials

Three questionnaires were designed: a paper copy for res and dom caregivers

(see Appendix 20), a digital copy for the FB (see Appendix 21) control group and a

second digital copy for the SONA control group (see Appendix 22). A mixture of paper

and digital copies were used to access a wider breadth of participants. The outlay of

the questionnaire information changed from the proposed design for ease of use and to

cut down on pages so as not to make it appear too daunting for the participants. All

three questionnaires contained an information page (see Appendix 1, 2 and 3),

demographics page (see Appendix 23, 24 and 25), the POS questionnaire (see

Appendix 26), LOC questionnaire (see Appendix 27), BES questionnaire (see

Appendix 28), RSES questionnaire (see Appendix 29), HADS questionnaire (see

Appendix 30) and a debrief page (see Appendix 31, 32 and 33). Participation was

voluntary. The paper copy contained a separate consent form requiring a signature and

date (see Appendix 34). Consent from the SONA group was obtained via clicking a box

at the bottom of the information page (see Appendix 3) whereas consent from the FB

group was assumed post questionnaire submission (see Appendix 2).

Whilst there were subtle differences between all three questionnaires’

demographic pages because of the different target population samples, both the FB

and SONA questionnaires were not exclusive to non-caregivers (see Appendix 24 and

25. Both digital demographics pages contained a question relating to caregiver

employment status (see Appendix 24 and 25). This option was made available to

obtain as many caregiver responses as possible. There were also subtle content

differences within the information sheet content due to the participant type. For

example, the SONA group’s information page was written for the perspective of

psychology student thus the language and presentation of content would appeal to

their academic nature (see Appendix 3). Also, unlike the other groups an incentive of

two credits (an accumulation of 10 credits enables the student to use the SONA system

for their own research purposes) was awarded for participation indicated in the

information sheet (see Appendix 3). However all information sheets contained details of

the research, its intent, anonymity, instruction, ethical consent, dissemination and

contact details (see Appendix 1, 2 and 3).

There were also subtle differences between groups in the debrief page. This

concerned the practicality of returning the questionnaires. Whereas the paper copies

were to be handed back to an assigned staff liaison or manager charged with the role

of distribution and collection (see Appendix 31), the digital copies contained a

submission button which automatically uploaded the content on to a formic system

(software that collates digital information and stores it on a central database) (see

Appendix 32 and 33).

Procedure

Accesses to caregiver participants were sought through a series of phone calls

and emails directly to the organisational manager. Prior to the distribution of

questionnaires an email of consent was required from the organisational managers

(see Appendix 35). In some cases, face to face meetings were setup. Printed

questionnaires (see Appendix 20) were handed directly to the organisational manager

or to an arranged staff liaison to distribute amongst their care staff. All completed,

withdrawn or incomplete questionnaires were handed back to the liaison/organisational

manager and locked away safely until collected by the researcher. Raw data was

extrapolated form the paper questionnaires by a process of digital scanning. This data

was stored in a central database accessible only to University psychology technicians

or on request by the researcher. Simultaneously, control group participation was

sought through a private message via the researcher’s private FB account. The

message contained a hyperlink to the questionnaire (see Appendix 21). SONA

participants were not directly contacted. Details of the research were posted on the

University’s SONA page, only accessible to University SONA users (psychology

students and staff) (see Appendix 22). Questionnaire participation was voluntary.

Raw data from both digital questionnaires was transferred to the same

database along with the paper copy data after submission. Data collection lasted

approximately two and a half months. Consent forms were immediately separated from

the questionnaires, if not already done so by the participants, and stored separately in

a lockable filing cabinet. Once data had been extrapolated from the paper copy

questionnaires they were locked away in the same cabinet as the consent forms. There

were no identifiable markers to match corresponding consent forms and

questionnaires. Raw paper, SONA and FB questionnaire data was transferred into an

SPSS spreadsheet and emailed to the researcher after data collection had concluded

(see supplementary CD). This was then stored on an encrypted and password

protected USB stick. Due to technical issues following the shutting down of the formic

system with two weeks still remaining of the data collection period, the researcher input

data manually from questionnaires returned within this time period. Missing data points

in the paper copy SPSS spreadsheet was input manually from the original

questionnaires then data from all three spreadsheets were input into one SPSS spread

sheet (see supplementary CD).

Reverse scoring of appropriate questions were undertaken as was the division

of HADS and BES into their two subscales. Means scores were calculated and input

into a new SPSS data sheet with the demographic data (see supplementary CD). Data

analysis followed. Following research submission, a condensed version of the findings

was shared with the consented organisations that took part. This was within the interest

of communication and data sharing. The findings will enable the organisations to reflect

on their own practice. All organisations were made aware that this is a summation data

from all consented organisations that individually cannot be made identifiable.

Methodological issues

Initial analysis of the two factor empathy scale found the mean scores to be

abnormally low. This was attributed to the direction of the Likert scale which was

reversed on the questionnaire (See Appendix 28). All empathy scores were reversed in

the SPSS data sheet and means were recalculated. For replication of study Likert

scoring direction of the BES will need to be revised. Whilst non caregiver respondents

had the opportunity to fill out the questionnaires as a dom or res caregiver, the section

indicating what type of individual do they care for (for example, Elderly and dementia)

was not available (See Appendix 24 and 25). Furthermore, because the question

relating to what type of individual do you care for was a multiple choice option. A

number of respondents had ticked more than one option due to comorbidity of

disabilities. The scanning system that inputs the data into the formic system that

collates onto SPSS could not distinguish between options (See Appendix 23). A better

method would have been to number the options or to have had a comorbid option to

prevent data being input by hand.

Design and analysis

A between-subjects, cross sectional design was used to compare two subject

groups, who were domiciliary caregivers and residential caregivers. A third non-

caregiver subject group (control) was also used as a baseline to compare caregiver

findings against.

Initially descriptives were run to find out the minimum, maximum and mean

scores from all scales within the questionnaire from each group (see Appendix.36, 37

and 38). A one way between subjects ANOVA was carried out to determine whether

there were any significant differences between the scale means (see Appendix 39) with

Tukey’s post hoc analysis to specify the statistically significant difference between the

groups (see Appendix 40). Only significant Tukey post hoc scores were reported.

Cronbach’s Alpha test was conducted onto measure the internal consistency

with participants’ HADS scores (see Appendix 10, 11 and 12), POS scores (see

Appendix 13), LOC scores (see Appendix 14), BES scores (see Appendix 16, 17 and

18) and RSES scores (see Appendix 19).

Pearson correlation test of all psychological variables was carried out (see

Appendix 41). Demographics and characteristics were excluded. For parsimonious

reasons only the variables that were significantly associated with the anxiety and

depression were put forward as predictor variables in hierarchical multiple regression

analyses. The enter method was used because this was an exploratory study wanting

to find the exact contribution of each variable to anxiety (see Appendix 42) and

depression (see Appendix 43). High comorbidity of anxiety and depression led to the

exclusion of these variables as predictors. POS was the last psychological predictor

entered into the multiple hierarchical regression analyses to examine its individual

contribution to predicting anxiety and depression over and above the other variables

because of its assumed relationship with them. Occupation type was the last variable

entered into the model to ascertain to what extent a caregiving or non-caregiving role

predicted anxiety and depression symptomology within this sample. A dom caregiver

was scored with a 1, a res caregiver was scored with a 2 and the control were scored

with a 3 (see supplementary cd). Alpha levels were set at 0.05.

Results

Analyses focused on participants’ anxiety, depression, POS, LOC, affective

empathy, cognitive empathy and self-esteem levels. Initially, descriptive analysis was

calculated including means, standard deviations of all psychological variables (see

table 1). One way analyses of variance were undertaken to ascertain if there was any

statistical difference between group scores (see table 1) with Tukey’s post hoc analysis

used to identify the statistical difference between which of the groups. Post hoc

analysis was only performed where ANOVAs were found to be statistically significant.

Thirdly, correlation analysis was performed to ascertain statistically significant

associations between variables (see table 2). Finally, enter hierarchical multiple

regression analyses were conducted to ascertain which entered variables would

statistically predict anxiety and depression symptomology (see table 3 and 4). Anxiety

and depression were omitted from the analysis when predicting each other. Cognitive

and affective empathy were entered together into the same step apart from in the

hierarchical multiple regression to predict anxiety. Only statistically associated

variables found in the correlations analysis were entered into the regression analysis.

Inferential tests

Depression scores for all three groups were under the threshold for sub clinical

depression. Table 1 shows that control group had the lowest scores for depressive

symptomology, whilst the dom group had the highest with the res group scoring

between the two. A one way ANOVA found there was a significant difference, with

Tukey’s post hoc analysis confirming that it was between the higher depression scores

of the dom group and the lower depression scores of the control group (p = .01). Table

1 shows that the res group had the highest mean scores for anxiety symptomology and

the control group had the lowest with dom group scoring between the two. The res

group mean score was just below the threshold for borderline anxiety, while both dom

and control groups mean scores were within normal range. A one way ANOVA found

there to be no significant differences between scores.

All three POS mean scores were within the higher range of the POS scale over

the central point (above 24). Table 1 shows that dom caregivers had the highest scores

for positive POS, whilst the control group had the lowest with the res group scoring

between the two. A one way ANOVA found there to be no significant differences

between scores.

All three internal LOC mean scores were within the higher range of the LOC scale over

the central point (above 17.5). Table 1 shows that the control group experienced a

greater loss of control compared to dom group who experience less loss of control with

residential scoring between the two. A one way ANOVA found there to be no

significance between the scores.

All three mean group affective empathy scores were just within the higher range

of the BES over the central point (above 35). Table 1 shows that the control group had

the highest scores for affective empathy, whilst the dom group scored the lowest for

affective empathy with the res group scoring between the two. A one way ANOVA

found there to be no significance between scores.

Similarly, all three mean group cognitive empathy scores were within the higher

range of the BES over the central point (above 27). Table 1 shows that the control

group obtained the highest mean scores for cognitive empathy symptomology, whilst

the res group scored the lowest with the dom group was scoring between the two. A

Table 1

Descriptives and one way ANOVA scores of the HADS, POS, LOC, basic empathy and self-esteem scales undertaken by dom caregivers,

res caregivers and control the control group.

Variable Caregiver type

Mean(M)

Stranded deviation (SD)

Min/maxscored

Min/maxpossible

One way ANOVA

Depression: Domiciliary 4.49 3.55 0/13 0/21Residential 3.42 2.83 0/17Control 2.85 2.29 0/10 F(2, 200) = 4.767, p = .01

Anxiety: Domiciliary 6.78 3.16 0/16 0/21Residential 7.02 3.35 1/16Control 6.76 3.28 0/16 F(2, 200) = .147, p = .86

Perceived organisational support: Domiciliary 34.98 12.94 6/48 0/48

Residential 34.65 10.89 9/48Control 31.21 8.89 9/45 F(2, 200) = 2.434, p = .09

Locus of control: Domiciliary 22.73 3.82 14/28 7/28Residential 22.19 3.47 14/29Control 21.88 3.83 7/28 F(2, 200) = .768, p = .47

Empathy affective: Domiciliary 36.69 4.55 26/48 11/55Residential 38.29 6.27 21/51Control 38.58 5.29 27/48 F(2, 200) = 1.909, p = .15

Empathy cognitive: Domiciliary 35.65 3.35 27/43 9/45Residential 35.29 4.46 24/45Control 36.94 3.64 26/45 F(2, 200) = 3.439, p = .03

Self-esteem: Domiciliary 31.31 4.72 21/40 10/40Residential 29.68 5.78 19/40Control 31.43 5.16 20/40 F(2, 200) = 2.518, p = .08

one way ANOVA found there to be a significant difference between the scores, with Tukey’s

post hoc analysis confirming that the moderate significant difference was between the higher

control group scores and the lower res group scores (p = .03).

All three mean self-esteem empathy scores were within the higher range of the

RSES over the central point (above 25). Table 1 shows that the control group had the

highest mean scores for self-esteem, while res group scored the lowest with the dom group

scoring between the two. A one way ANOVA found there was no significance between

scores.

Pearson’s correlation test. Pearson’s correlation was used to examine associations

between all variables within the study (see table 2). Table 2 indicates that there was a strong

statistically significant positive association between depression and anxiety. Furthermore,

depression was found to have a strong statistically significant association between low levels

of POS and low levels of internal LOC. Depression also had a moderate yet statistically

significant association with low levels of affective empathy, low levels of cognitive empathy

and low levels of self-esteem. Table 2 indicates that there was a strong statistically

significant association between anxiety and low levels of internal LOC. Furthermore, anxiety

was also shown to have a moderate though statistically significant association with low

levels of self-esteem and a moderately statistically significant positive association with

affective empathy.

Table 2 indicates a strong statistical significant positive association between POS

and internal LOC. Equally, there was a strong statistically significant positive association

between cognitive empathy and internal LOC and between self-esteem with internal LOC.

Also, there was a strong statistically significant positive association between cognitive

empathy and affective empathy. All variables statistically associated with anxiety and

depression was suitably reliable for examination through hierarchical regression (see tables

3 and 4).

Table 2

Bivariate association between groups (dom caregivers, res caregivers and control group)

and HADS, POS, LOC, basic empathy and self-esteem scores

1 2 3 4 5 6 71 Depression 12 Anxiety .45** 13 Perceived organisational support

-.22** -.12 1

4 Locus of control -.37** -.31** .42** 15 Affective empathy -.15* .15* -.03 .00 16 Cognitive empathy -.16* .02 .04 .22** .35** 17 Self-esteem -.15* -.14* .02 .26** -.02 .10 1

**p < .01; *p < .05

Hierarchical multiple regression predicting Anxiety. Internal LOC variable was

entered into the first step of the hierarchical multiple regression analysis. As such, low levels

of internal LOC was found to be statistically significant explaining for 9.3% of variance in

anxiety symptomology adjusted to 9% (see table 3). Higher levels of affective empathy

increased the predictive power of this model explaining a total of 11.7% adjusted to 11%.

However, whilst higher levels of affective empathy were found to be a uniquely statistically

significant contributor, its predictive capability is weak. In step three, low levels of self-

esteem faintly increased the predictive power of the entire model explaining 11.9% of anxiety

symptomology. However, its individual contribution was not statistically significant over and

above the other variables and its influence actually decreased the predictive capability of the

model at this point, adjusted to 11%. After entry of occupation at step four the total variance

explained by the model increased to 12.2%. However, its individual contribution was not

statistically significant and decreased the predictive power of the overall model, adjusted to

10%. Low levels of internal LOC and higher levels of affective empathy remained statistically

significant and marginally stable throughout all steps of the analysis. Furthermore, they were

the only two variables that made a statistically significant contribution towards anxiety

symptomology over and above low levels of self-esteem and occupation type. However, the

unique contribution made by low levels of affective empathy was weak.

Table 3

Summary of hierarchical multiple regression analysis for variables predicting anxiety

AnxietyR-sq adj.)

F (p-value) β (p-value)

Step 1Locus of control

.09 20.684 (p<.001)-.305 (p<.001)

Step 2Locus of controlAffective empathy

.11 13.214 (p<.02)-.306 (p<.001) .154 (p<.02)

Step 3Locus of controlAffective empathySelf-esteem

.11 9.081 (p<.47)-.293 (p<.001) .148 (p<.03)-.050 (p<.47)

Step 4Locus of controlAffective empathySelf-esteemOccupation

.10 6.857 (p<.46)-.298 (p<.001) .155 (p<.02)-.047 (p<.50)-.050 (p<.46)

Hierarchical multiple regression predicting depression. Internal LOC variable

was entered into the first step of the hierarchical multiple regression analysis and was shown

to be a statistically significant contributor explaining 13.3% of variance in depression

symptomology adjusted to 13% (see table 4). Low levels of affective empathy and low levels

of cognitive empathy marginally increased the predictive power of the model at step two

explaining 15.6% of the variance on predicting depression symptomology, adjusted to

14.4%. However only lower levels of affective empathy were found to be statistically

significant (see table 4). Low levels of self-esteem marginally increased the predictive power

of the model at step 3 explaining a 16.2% of the variance of depression symptomology

adjusted 14.5%. However, its individual contribution was not found to be statistically

significant above and beyond all other variables at this step. Low levels of POS marginally

increased the predictive power of the model at step 4 explaining 16.9% of the variance of

depression symptomology adjusted to 14.8%. After entry of occupation type at step five the

total variance explained by the model was 22.2% adjusted to19.8%. The individual

contribution of occupation type was found to be a statistically significant contributor in

explaining depression symptomology over and above all other variables. The occupation

type was the domiciliary group.

Table 4

Summary of hierarchical multiple regression analysis for variables predicting depression

DepressionR-sq adj.)

F (p-value) β (p-value)

Step 1Locus of control

.13 30.874 (p<.001)-.365 (p<.001)

Step 2Locus of controlAffective empathyCognitive empathy

.14 12.289 (p<.07)-.356 (p<.001)-.135 (p<.05)-.037 (p<.61)

Step 3Locus of controlAffective empathyCognitive empathySelf-esteem

.15 9.544 (p<.26)-.338 (p<.001)-.146 (p<.04)-.030 (p<.68)-.076 (p<.26)

Step 4Locus of controlAffective empathyCognitive empathySelf-esteemPerceived organisational support

.15 8.032 (p<.18)-.295 (p<.001)-.148 (p<.04)-.033 (p<.64)-.085 (p<.21)-.097 (p<.18)

Step 5Locus of controlAffective empathyCognitive empathySelf-esteemPerceived organisational supportOccupation

.20 9.306 p<.001)-.313 (p<.001)-.129 (p<.06)-.004 (p<.95)-.076 (p<.25)-.123 (p<.08)-.235 (p<.001)

Whilst low levels of internal LOC marginally decreased as a unique predictor of depression

symptomology throughout each step of the analysis, it remained statistically significant.

Contrarily, the unique contribution of low levels of affective empathy marginally increased

throughout each step of the analysis until occupation was entered at the final stage when its

contribution became statistically insignificant. Only low levels of LOC and occupation type

were found to be strong statistically significant individual contributors towards explaining

depression symptomology within this sample. Whilst low levels of affective empathy were

significant until the final stage of the model, its significance was weak. The contribution of

low levels of cognitive empathy decreased throughout each step of the analysis.

Furthermore, the contribution of low levels of POS in explaining depressive symptomology

increased in the final stage of the analysis. However, it remained statistically insignificant

(see table 4).

Discussion

Reminder of aims and hypotheses

Based on recent reports surrounding the state of domiciliary and residential care in

the UK (for example, CQC, 2014; Franklin, 2014; Farooq, 2009; Koehler, 2014) and prior

well-being research (for example, Denton et al., 2002; Gray-Stanley et al., 2010) it was

hypothesised that dom and res caregivers would be more susceptible to anxiety and

depression symptomology than non-caregivers. However, that dom caregivers would be

more susceptible to anxiety and depression than res caregivers. It was further hypothesised

that low levels of internal LOC, low levels of affective empathy, low levels of cognitive

empathy, low levels of self-esteem and low levels of POS would predict anxiety and

depression symptomology. The results lent some strong support for the first hypothesis.

However, they showed less support for the second hypothesis.

Discussion of results and hypotheses

In line with the first hypothesis comparisons between mean HADS scores found that

both caregiver groups had higher anxiety and depression scores than the non-caregiver

groups. This result is consistent with prior research indicating that caregivers are more

susceptible to poorer MH than non-caregivers (Butterworth et al., 2010). Whilst dom

depression scores were higher than res and the control, post hoc analysis indicated that only

between dom and control group scores were statistically significant suggesting that res and

dom caregivers, within this sample, are similarly susceptible to depression symptomology.

However, results from the hierarchical analysis indicated that depression symptomology is

strongly attributed to working as a dom caregiver. Yet, this result should be viewed with

caution as all three mean depression scores were within normal depression range, thus not

clinically recognised for immediate intervention.

Contrary to the anxiety hypothesis, ANOVA yielded no significant differences

between all three groups’ mean anxiety scores. Also, occupation type was not found to be a

significant predictor of anxiety in the final step of the regression analysis. This suggests

anxiety symptomology is not necessarily attributed to working as a dom or res caregiver

compared to a non-caregiving role. Additionally, res caregiver mean anxiety scores were

higher than dom caregiver scores, just below the threshold for borderline anxiety. Whilst this

result conflicts with the hypothesis, it offers some moderate support for prior research

indicating that res caregivers are likely to experience anxiety (Mutkins et al., 2011). Yet,

these findings were used to interpret anxiety and depression’s relationship to emotional

exhaustion and depersonalisation for which only depression was found to be a significant

predictor.

In line with the second hypothesis, low levels of internal LOC were found to have a

statistically significant contribution to anxiety and depression over and above the other

psychological variables. This result supports prior research that higher levels of internal LOC

are related to better quality of life and fewer depressive symptoms (Gibson et al., 2013) and

provides evidence to the contrary of the belief that only external LOC predicts anxiety (for

example, Molinari & Khanna, 1981) and depression (Presson & Benassi, 1996). Whilst the

review of literature yielded no previous association between empathy and depression, low

levels of affective empathy were found to be a moderate predictor of depressive

symptomology above and beyond low levels of all other psychological variables. However, in

the final step of the hierarchical multiple regression analysis, low levels of affective empathy

was no longer found to be statistically significant suggesting that occupation was a much

greater predictor.

In contradiction to the hypothesis, higher levels of affective empathy were found to be

a moderate predictor of anxiety symptomology. This remained relatively stable throughout

each step of the analysis, suggesting that it has a greater contribution to anxiety

symptomology over all other psychological variables, apart from low levels of LOC. However,

whilst it was posited that there may be a relationship between empathy and anxiety, even

though a systematic review of literature yielded no direct association, the direction of this

result was unexpected. According to Austin et al., (2009) the ability to empathise would

decrease if a caregiver was subject to decrease in well-being. However, this was assumed

to be in relation to stress, burnout and emotional contagion not anxiety. As such this result

should be viewed with caution considering that the predictive power of high levels of

affective empathy on anxiety was only moderate.

Res and dom type characteristics reflects both national and past research caregiver

characteristics. Specifically, the disproportion of female to male employees, average age,

employment type and type of working hours (Crown et al., 1995; Skills for Care, 2015;

Zeytinoglu, et al., 2009), validating comparison to past caregiver research.

Discussion of POS variable. No other psychological variables were found to

contribute towards anxiety and depressive symptomology. Non-caregivers recorded higher

mean POS questionnaire scores than both caregiver groups. Furthermore, analysis of

variance reported no significant difference between group mean scores, suggesting that

caregivers feel equally valued by their organisation than non-caregivers. This sample then

conflicts with the accounts from recent UK reports (for example, CQC, 2014; Franklin, 2014;

Koehler, 2014) that caregivers, particularly dom caregivers, are operationally, financially and

socially unsupported by their organisations.

POS had no significant correlation relationship with any of the psychological

variables apart from its positive association with internal LOC and a negative association

with depression. As covered, it was not found to be a significant predictor of depression

symptomology. Whilst this is contrary to previous findings that an employee with low POS

will generally suffer from anxiety symptomology (Rhoades & Eisenberg, 2002), it suggests

that this sample’s well-being may not be affected by negative organisational cultures as seen

in prior caregiver research (Denton et al., 2002). This may be inferred by the quality of

organisations that participated in the questionnaire who in turn may support and value their

employees. The plausibility of this interpretation is related to the poor response rates from

organisations willing to participate. Those who consented may already recognise their

contribution towards protecting their employees’ well-being, subsequently reflecting those

outstanding services recognised in the Care Quality Commission report (2014).

Discussion of self-esteem variable. No statistical differences were found between

groups RSES scores even though res caregivers recorded lower mean scores compared to

dom and res caregivers. This suggests that the role of caregiving may not contribute to lower

levels of self-esteem when compared to non-caregivers. Results of the correlation analysis

contribute to the literature associating low self-esteem with feeling anxious (Lee & Hankin,

2009; Riketta, 2004) and depressed (Riketta, 2004; Sowislo, et al., 2014; Sowislo et al.,

2014; Zunzunegui et al., 2002). However, correlations were found to be weak and did not

statistically predict anxiety and depression symptomology. A professional perspective could

suggest that caregivers socioemotional need from their organisations are met, proposed by

POS theory (Rhoades & Eisenberg, 2002). If the caregivers’ self-esteem is dependent on the

amount of organisational support then this may clarify its lack of accountability in explaining

depression and anxiety symptomology. As such, this may suggest that the recommendations

by the Burnstow report to improve self-esteem amongst caregiver (for example, greater

training, free influenza vaccinations and key-worker status) are already enforced within these

participated organisations (Koehler, 2014). However, the correlation analysis failed to show

a statistical association between POS and self-esteem.

Caregiver, self-esteem scores could be related to job satisfaction. In particular the

positive aspects of caregiving through maintaining healthy mutual relationships with care

receivers’ (Kramer, 1997; Lopez et al., 2005; Neufeld & Harrison, 1995). Potentially by

having adequate time to provide care and by using effective coping strategies to deal with

the potential challenging behaviours that provoke stress (Denton et al., 2002 ) known to

impact on self-esteem (Orth et al., 2009). This is inferred from the high proportion reported

care receivers that are either elderly, comorbid and or have dementia. These are both

associated with challenging behaviours (for example, Molyneux, et al., 2008; Rose et al.,

2013). However, the findings do not support this assumption. It is merely a postulation. As

such, self-esteem like mood and happiness fluctuates (Campbell et al., 1991; Galambos et

al., 2015) thus the differences in scores may be related to how the participants felt during the

questionnaire completion as a consequence of external life events or even the time of day.

Discussion of affective and cognitive empathy variables. Both affective and

cognitive BES mean scores were within a similar range suggesting that caregivers and non-

caregivers are equally capable at recognising other people’s emotions and mental states as

well as their own (Carré et al., 2013; Jolliffe & Farrington, 2006). Correlation analysis

confirmed their positive relatability in line with the development and validation of the scale

(Jolliffe & Farrington, 2006). Whilst between group affective empathy scores yielded no

significant difference, post hoc analysis found a moderate statistical difference between dom

caregiver and control mean cognitive scores in favour of control. Contextually, this result

conflicts with the assumptions that caregivers may be more adapt to feelings of empathy, a

prerequisite for providing compassionate care for vulnerable adults (Globe et al., 2009;

Koehler, 2014). Similarly, the disproportion in female participants between the caregiver

groups (84.3%, 76.5%) and non-caregiver group (50.7%) conflicts with Joliffe and

Farrington’s (2006) findings that females had higher affective and cognitive empathy scores

than males and that females were more likely to respond prosocially to an individual’s

distress. However, this was not related to a caregiving context. This result could be

interpreted as compassion fatigue known to impact on empathy triggered by stress and

burnout as found in nurses (Austin et al., 2009). However, correlation results found no

negative association between affective and cognitive empathy and other variables apart from

a moderate association between affective empathy and depression. This suggests if the lack

of difference in empathy scores between caregiver and non-caregiver groups is attributable

to a psychological impact, then POS, LOC, self-esteem and anxiety are not one them,

according to these findings. Thus, additional research is required to quantify this assumption

using other variables such as stress and burnout.

A lack of statistical difference between caregiver and non-caregiver mean empathy

scores may be attributable to the appropriateness of the scale. Designed for adolescents,

contextually it is not designed for the work place. Therefore questions including ‘‘It is hard for

me to understand when my friends are sad’’ (Jolliffe & Farrington, 2006, p. 593), may not be

contextually relevant as caregivers may be able to differentiate how they control or interpret

their emotions between a friend who maybe in distress compared to a care receiver who is in

distress. To quantify this assumption a replication study may use the BES-A (Carré, 2013).

Whilst questions including “Other people’s feelings don’t bother me at all” (Carré, 2013, p.

690) are not directly related to an occupational setting they are more generic thus may give

a truer representation of empathy between a caregiver and non-caregiver.

Discussion of LOC variable. LOC results show that caregivers scored higher than

the non-caregiver group. Yet, no statistical difference between LOC means scores were

found, suggesting that there is no remarkable difference in how caregivers and non-

caregivers within this sample perceive themselves to be in control of their own life events. All

scores were in the higher range of the scale suggesting that all groups have above average

internal LOC in relation to the numerical scoring of the scale. Also, correlation analysis found

a statistically significant positive association between LOC and POS. This conflicts with the

assumption that caregivers maybe more susceptible to lower levels of internal LOC than

non-caregivers inferred from the lack of social, financial and operational support for

caregiver , particularly dom caregiver, in the workplace (for example, CQC, 2014; Denton et

al., 2002; Franklin, 2014; Koehler, 2014). As such, POS theory indicates that feeling valued

as an employee is related to aspects of social support in the workplace (Rhoades &

Eisenberg, 2002) not just operational support. Thus the positive association between POS

and LOC may have been mediated by aspects of social support from the organisation. The

plausibility of this assumption reflects the research by Lefcourt et al. (1984) who believed

that being socially supported can lead to positive coping strategies for those with an internal

LOC that in turn deter depressive symptomology (Chan, 2000). This interpretation reflects

the negative association between internal LOC and POS as an outcome of the correlation

analysis mirroring the belief that a lack of social support is related to depressive

symptomology (Rodakowski et al., 2012) in a caregiving capacity. However this was in

relation ageing adults with spinal cord injuries cared for by paid family members. Similarly,

correlation analysis revealed a strong positive association between internal LOC and self-

esteem. This association is consistent with prior research indicating that high internal LOC is

related to coping strategies that manage not just self-esteem (Anderson, 1998) but stress

(Hay & Diehl, 2010) and psychological adjustment (Elfstöm & Kreuter, 2006) in non-

caregivers and in the caregiving context also (Pelletier et al., 1994).

Limitations

A number of limitations of the current research should be noted, including the small

sample sizes, which appear problematic in caregiver research (Mutkins et al., 2011). Of all

three groups dom caregivers had the poorest response rate. This was in part due to the

nomadic nature of the work. It was communicated from dom care managers that caregiver

visits to the office were brief and infrequent impacting on the ability of managers to distribute

and collect the questionnaires. Furthermore, a large number of dom care organisations were

retendering for contracts during the data collection period. This further impacted on

questionnaire response rate as dom caregivers were moving companies for fear of job

losses if contracts were not secured. In one instance, a consented dom care organisation

disbanded prior to questionnaire distribution. Similarly, participant response rates between

organisations were extremely varied, impacting on an accurate measurement of the

variables. This was interpreted by the sensitive and personal nature of the questions and the

length of the questionnaire during feedback. However, inconsistent response rates are seen

to be a common occurrence within caregiver samples (Gray-Stanley et al., 2010).

Furthermore, whilst questionnaire feedback was taken from two regions it is not nationally

representative thus the results must be interpreted with some caution.

Future research

A complete replication of this study would require a larger sample size from all three

groups. This would preferably be a national sample thus requiring a longer period for data

collection. Furthermore, an incentive might be used to entice participation from those less

forthcoming organisations. However, if this research reflected the lack of statistical

difference between caregivers and non-caregiver group scores found by this study, then

perhaps using different well-being measures would be more appropriate. For example,

subjective burden and dysfunctional coping are seen to be related to anxiety (del‐Pino‐Casado et al, 2014). Similarly, challenging behaviour and cognitive functioning of elderly

dementia care receivers are known to have had a direct impact on depression levels

(Molyneux et al., 2008). However these are hard to qualify as individual predictors of anxiety

and depression because of their relatability to other psychological variables recognised as

umbrella terms including burnout and ‘work stress’, which have seen to be related to

depression (Gray-Stanley et al., 2010).

Whilst this was an explorative study, future research may endeavour to differentiate

between the impacts natural stressors related to the caring role have on anxiety and

depression and the related significance that poor organisational support has on anxiety and

depression as a means to provide the correct interventions. One method adopted by Denton

et al. (2002) was the use of focus groups. By gathering specific information on the

caregiver’s perceptions of their work environment they were able to develop a self-

administrative questionnaire in conjunction with findings from a literature review. Caregivers

MH and well-being were then measured by three independent variables (job satisfaction,

stress and job stress) by ordering focus group outcomes (for example, working conditions)

and coding questions into Likert point scales. As such, Denton et al. (2002) were able to

pinpoint working characteristics of the caregiving role towards well-being measures. For

example, job-related stress was related to excessive workload and difficult clients. This

approach would enable this current research to pin point with greater accuracy the negative

perceptions of working for an organisation with specific practices that may lead to anxiety

and depression symptomology. Findings dependent, this research will be able to make

specific recommendations to organisations as a means for positive interventions.

Conclusion

This research was aimed to quantify and compare res and dom caregiver’s

susceptibility to anxiety and depression compared with a non-caregiver group and to

ascertain what extent anxiety and depression were predicted by low levels of LOC, low

levels of POS, low levels of self-esteem, low levels of affective empathy and low levels of

cognitive empathy. This research has shown moderate support for the attribution of anxiety

and depressive symptomology in res and dom caregivers compared to non-caregivers within

this sample. Whilst the role of a dom caregiver was shown to be a statistically significant

predictor of depression, the role of a res caregiver did not. This partially supports the first

hypothesis. Furthermore, results suggest that whilst a res caregiver is most likely to

experience anxiety symptomology when compared to a dom caregiver role and to the non-

caregiver group, res caregiving occupation type did not statistically predict anxiety

symptomology within this sample. These results must be viewed with caution considering

that all three groups’ means scores fell within normal anxiety and depression range. Similarly

caution should be raised considering the lack of statistical difference between the groups’

anxiety and depression mean scores.

Furthermore, this research found low levels of internal LOC to be a statistically

significant predictor of anxiety and depression symptomology contrary to the belief that only

external LOC predicts anxiety (for example, Molinari & Khanna, 1981; Presson & Benassi,

1996). Similarly, low levels of affective empathy were found to be a weak predictor of

depression symptomology. However, it was found not to be statistically significant when

taking into account occupation type. An unexpected outcome was the statistical significance

of high levels of affective empathy in predicting anxiety symptomology. A review of literature

research found no direct association thus future research will be beneficial in verifying and

determining this relationship. However the significance of this contribution was fairly weak

and must be interpreted with caution. Within this population sample no support was found for

the assumption that caregiver-well-being may negatively influenced by POS considering

both caregivers had higher mean POS scores than the control. The lack of negative

association between POS and psychological well-being does not provide additional evidence

to the reports suggesting that some res and dom care organisations are failing to financially,

socially, and operationally support their staff in the UK (for example, CQC, 2014; Franklin,

2014; Koehler, 2014). However, these findings must be viewed with caution considering the

small size of the sample. Additionally, the size of the response from the large number of

organisations approached to take part in the research must be considered. It may be inferred

that those organisations unwilling to participate may not have wanted to draw additional

attention to their service recognising operational flaws within. As such, a large proportion of

the consented organisations may reflect those outstanding services known to operate in the

UK (CQC, 2013, 2014), therefore attributing to the lack of significance in these findings.

However, this interpretation must be viewed with caution.

Future research must focus on extending the distribution range of questionnaires to

achieve a better representative sample. Furthermore, in finding a truer representation of the

impact POS may have on res and dom caregivers, a method such as focus groups (Denton

et al., 2002) to develop a self-report questionnaire based on caregivers views of support in

the workplace related to the quality assurance ideals presented by Eustis et al. (1993) and

well-being must be considered. This will enable the research to differentiate between what

impacts the role of caregiving and what impact the lack of organisational support has on

anxiety and depression as a means to develop specific methods of intervention.

References

Aggar, C., Ronaldson, S., & Cameron, I. D. (2011). Self-esteem in carers of frail older

people: Resentment predicts anxiety and depression. Aging & Mental Health, 15(6),

671–678. doi: 10.1080/13607863.2011.562176

Anderson, A. L. (1998). Strengths of gay male youth: An untold story. Child and Adolescent

Social Work Journal, 15(1), 55–71. doi: 10.1023/A:1022245504871

Austin, W., Globe, E., Leier, B., & Bryne, P. (2009). Compassion fatigue: the experiences of

nurses. Ethics and Social Welfare, 3(2), 195-214. doi: 10.1080/17496530902951988

Bartoldus, E., Gillery, B., & Sturges, P. J. (1989) Job-related stress and coping among

home-care workers with elderly people. Health and Social Work, 14(3), 204-210.

doi:10.1093/hsw/14.3.204

Blanco, V., Rohde, P., Vázquez, F. L., & Torres, A. (2014). Identification of caregivers at

greatest risk of major depression in two prevention studies. Psychotherapy

Research, 24(5), 578-593, doi: 10.1080/10503307.2013.847989

Bjelland, I., Dahl, A. A., Haug, T. T., & Neckelmann, D. (2002). The validity of the Hospital

Anxiety and Depression Scale An updated literature review. Journal of

Psychosomatic Research, 52(2), 69 –77. doi:10.1016/S0022-3999(01)00296-3.

Blumenthal, S., Lavender, T., & Hewson, S. (1998). Role clarity, perception of the

organization and burnout amongst support workers in residential homes for people

with intellectual disability: a comparison between a National Health Service trust and

a charitable company. Journal of Intellectual Disability Research, 42(5), 409-417. doi:

10.1046/j.1365-2788.1998.00150.x

Bowling, A. (2005). Measuring health: A review of quality of life measurement scales (3rd

ed.). Maidenhead: Open University Press.

Brady, T. J. (2003). Measures of self-efficacy, helplessness, mastery, and control. Arthritis &

Rheumatism (Arthritis Care & Research), 49(5), 147–16. doi: 10.1002/art.11413

Brouwer, W., van Exel, N., van den Berg, B., van den Bos, G., & Koopmanschap, M. (2005).

Process utility from providing informal care: The benefit of caring. Health Policy,

74(1), 85–99. doi: 10.1016/j.healthpol.2004.12.008

Butterworth, P., Pymont, C., Rodgers, B., Windsor, T, D., & Antsey, K. J. (2010). Factors

that explain the poorer mental health of caregivers: Results from a community survey

of older Australians. Australian and New Zealand Journal of Psychiatry, 44(7), 616-

624. doi: 10.3109/00048671003620202

Campbell, J. D., Chew, B., & Scratchley, L. S. (1991). Cognitive and emotional reactions to

daily events: The effects of self-esteem and self– complexity. Journal of Personality,

59(3), 473–505. doi:10.1111/j.1467-6494.1991.tb00257.x

Carré, A., Stefaniak, N., D’Ambrosio, F., Bensalah, L., & Besche-Richard, C. (2013). The

basic empathy scale in adults (BES-A): factor structure of a revised form.

Psychological Assessment, 25(3), 679-691. doi: 10.1037/a0032297

Castle, N. G., & Engberg, J. (2007). The influence of staffing characteristics on quality of

care in nursing homes. Health and Research, 45(5), 1822-1847. doi: 10.1111/j.1475-

6773.2007.00704.x

Chappell, N., & Dujela, C. (2008). Caregiving: Predicting at-risk status. Canadian Journal on

Aging, 27(2), 169–179. doi: 10.3138/cja.27.2.169

Chan, R. C. K., (2000). Stress and coping in spouses of persons with spinal cord injuries.

Clinical Rehabilitation, 14(2), 137-144. doi: 10.1191/026921500675826560

Chester, H, Hughes, J., & Challis, D. (2014). Commissioning social care for older people:

influencing the quality of direct care. Ageing and Society, 34(6), 930-950. doi:

10.1017/S0144686X12001249

Cleveland, J. N., & Shore, L. M. (1992). Self-perspectives and supervisory perspectives on

age and work attitudes and performance. Journal of Applied Psychology, 77(4), 469–

484. doi: 10.1037/0021-9010.77.4.469

Cohen, C., Colantonio, A., & Vernich, L. (2002). Positive aspects of caregiving: Rounding out

the caregiver experience. International Journal of Geriatric Psychiatry, 17(2), 184–

188. doi: 10.1002/gps.561

Colton, M., & Roberts. S. (2007). Factors that contribute to high turnover among residential

child care staff. Child and Family Social Work, 12(2), 133–142. doi: 10.1111/j.1365-

2206.2006.00451.x

CQC, (2013). Not just a number home care inspection programme: National overview.

Retrieved from: http://www.cqc.org.uk/sites/default/files/documents/9331-cqc-

home_care_report-web_0.pdf.

CQC, (2014).The state of health care and adult social care in England: 2013/2014. Retrieved

from: http://www.cqc.org.uk/sites/default/files/state-of-care-201314-full-report-1.1.pdf

Crown, W., Ahlburg, D., & McAdam, M. (1995). The demographic and employment

characteristics of home care aides: a comparison with nursing home aides, hospital

aides, and other workers. The Gerontologist 35(2), 162-170. Retrieved from:

http://gerontologist.oxfordjournals.org/content/35/2/162.full.pdf

CSSIW. (2014). National review of commissioning for social services in Wales 2014:

overview report. Retrieved from:

http://cssiw.org.uk/docs/cssiw/report/140416overviewen.pdf

Cummings, S. M., & Kropf, N. P. (2015). Predictors of depression among caregivers of older

adults with severe mental illness. Journal of Gerontological Social Work, 58(3), 253-

271. doi: 10.1080/01634372.2014.978927

Daly, M. (2008). Concerns over private care firms. Retrieved from:

http://news.bbc.co.uk/1/hi/scotland/7455711.stm

D’Ambrosio, F., Oliver, M., Didon, D., & Besche, C. (2009). The basic empathy scale: a

French validation of a measure of empathy in youth. Personality and Individual

Differences, 46(2), 60-165. doi: 10.1016/j.paid.2008.09.020

Davis C. M. (1990). What is empathy, and can empathy be taught? Physical Therapy,

70(11), 707–711. Retrieved from: http://www.physio-pedia.com/images/5/5b/Davis_-

_1990_-_What_is_empathy,_and_can_empathy_be_taught-annotated.pdf

Davis, M. H. (1983). The effects of dispositional empathy on emotional reactions and

helping: A multidimensional approach. Journal of Personality, 51(2), 167–184. doi:

10.1111/j.1467-6494.1983.tb00860.x

del‐Pino‐Casado, R., Pérez‐Cruz, M., & Frías‐Osuna, A. (2014). Coping, subjective burden

and anxiety among family caregivers of older dependents. Journal of Clinical

Nursing, 23(23-24), 3335-3344. doi: 10.1111/jocn.12561

Denton, M. A., Zeytinoğlu, I, U., & Davies, S. (2002). Working in clients' homes: the impact

on the mental health and well-being of visiting home care workers. Home Health

Care Services Quarterly, 21(1), 1-27. doi: 10.1300/J027v21n01_01

Dijkstra, M. T.M., Beersma, B., & Cornelissen, R. A. W. M. (2012). The emergence of the

activity reduces conflict associated Strain (ARCAS) model: a test of a conditional

mediation model of workplace conflict and employee strain. Journal of Occupational

Health Psychology, 17(3), 365-375. doi: 10.1037/a0028550

Edwards, A. B., Zarit, S. H., Stephens, M. A. P., & Townsend, A. (2002). Employed family

caregivers of cognitively impaired elderly: an examination of role strain and

depressive symptoms. Aging and Mental Health, 6(1), 55-61. doi:

10.1080/13607860120101149

Eisenberger, R., Huntington, R., Hutchison, S., & Sowa, D. (1986). Perceived organizational

support. Journal of Applied Psychology, 71(3), 500–507. doi: 10.1037/0021-

9010.71.3.500

Elfström, M. L., & Kreuter, M. (2006). Relationships between locus of control, coping

strategies and emotional well-being in persons with spinal cord lesion. Journal of

Clinical Psychology in Medical Settings, 13(3), 89–100. doi:10.1007/s10880-005-

9001-8

Eustis, N. N., & Fischer, L. (1991). Relationships between home care clients and their

workers: Implications for quality care. The Gerontologist, 31(4), 447-456. doi:

10.1093/geront/31.4.447

Eustis, N. N., Kane, R. A., & Fischer, L. R., (1993). Home care quality and the home care

worker: beyond quality assurance as usual. The Gerontologist, 33(1), 64-73. doi:

10.1093/geront/33.1.64

Farooq, A. (2009). Britain's Homecare Scandal. Retrieved from:

http://news.bbc.co.uk/panorama/hi/front_page/newsid_7986000/7986482.stm

Franklin, B. (2014). The future care workforce. Retrieved from:

file:///C:/Users/Alex/Downloads/Future_Care_Workforce_Report%20(1).pdf

Galambos, N. L., Fang, S., Krahn, H. J., Johnson, M. D., & Lachman, M. E. (2015). Up not

down: the age curve in happiness from early adulthood to midlife in two longitudinal

studies. Developmental Psychology, 1-8. doi: 10.1037/dev0000052

George, J. M., Reed, T. F., Ballard, K. A., Colin, J., & Fielding, J. (1993). Contact with AIDS

patients as a source of work-related distress: Effects of organizational and social

support. Academy of Management Journal, 36(1), 157–171. doi: 10.2307/256516

Geng, Y., Xia, D., & Qin, B. (2012). The basic empathy scale: Chinese validation of a

measure of empathy in adolescents. Child Psychiatry & Human Development,

43(40), 499-510. doi: 10.1007/s10578-011-0278-6

Gibson, R. C., Morgana, K. A. D., Abela, W. D., Sewella, C. A., Martina, J. S., Lowea, G. A.,

Hayea, W. De La., E dwardsb, C, L., O’Garoc, K. N., Reidd, M. E., & Asnanid, M. R.

(2013). Locus of control, depression and quality of life among persons with sickle cell

disease in Jamaica. Psychology, Health & Medicine, 18(4), 451–460, doi:

10.1080/13548506.2012.749353

Gray-Stanley, J. A., Muramatsu, N., Heller, T., Hughes, S., Johnson, T. P., & Ramirez-

Valles, J. (2010). Work stress and depression among direct support professionals:

The role of work support and locus of control. Journal of Intellectual Disability

Research, 54(8), 749–761. doi: 10.1111/j.1365-2788.2010.01303.x

Harvey, S., & Keashley, L. (2003). Predicting aggression in the workplace: risk factors, self-

esteem, and time at work. Social behaviour and Personality, 31(8), 807-814. doi:

10.2224/sbp.2003.31.8.807

Hatch, D. J., DeHart, W. B., & Norton, M. C. (2014). Subjective stressors moderate

effectiveness of a multi-component, multi-site intervention on caregiver depression

and burden. International Journal of Geriatric Psychiatry, 29(4), 406-413. doi:

10.1002/gps.4019

Hay, E. L., & Diehl, M. (2010). Reactivity to daily stressors in adulthood: The importance of

stressor type in characterizing risk factors. Psychology and Aging, 25(1), 118–131.

doi: 10.1037/a0018747

Hunter, C. E. A., Ward, L., & Camp, C. J. (2012). Transitioning spaced retrieval training to

care staff in an Australian residential aged care setting for older adults with dementia:

a case study approach. Clinical Gerontologist, 35(1), 1-14. doi:

10.1080/07317115.2011.626513

Jolliffe, D., & Farrington, D. P. (2006). The development and validation of the Basic Empathy

Scale. Journal of Adolescence, 29(4), 589-611. doi:10.1016/j.

adolescence.2005.08.010.

Kernis, M. H., Whisenhunt, C. R., Waschull, S. B., Greenier, K. D., Berry, A. J., Herlocker, C.

E., & Anderson, C. A. (1998). Multiple facets of self-esteem and their relations to

depressive symptoms. Personality and Social Psychology Bulletin, 24(6), 657– 668.

doi:10.1177/ 0146167298246009

Koehler, I. (2014). Key to care: Report of the Burstow Commission on the future of the

homecare workforce. Retrieved from

http://www.lgiu.org.uk/wp-content/uploads/2014/12/KeyToCare.pdf

Kramer, B. J. (1997). Gain in the caregiving experience: Where are we? What next?

Gerontologist, 37(2), 218–232. Retrieved from: file:///G:/Dissertation/self

%20esteem/self-esteem%20krammer.pdf

Lance Ferris, D., Lian, H., Brown, D. J., & Morrison, R. (2015). Ostracism, self-esteem, and

job performance: When do we self-verify and when do we self-enhance? Academy of

Management Journal, 58(1), 279-297. doi: 10.5465/amj.2011.0347

Lefcourt, H. M., Martin, R. A., & Saleh, W. E. (1984). Locus of control and social support:

Interactive moderators of stress. Journal of Personality and Social Psychology, 47(2),

378–389. doi: 10.1037/0022-3514.47.2.378

Lefcourt, H. M. (1991). Locus of control. In J. P. Robinson, P. R. Shaver, & L. S. Wrightsman

(Eds.), Measures of personality and social psychological attitudes (2nd ed., pp. 413–

498). San Diego, CA: Academic Press.

Lee, A., & Hankin, B. L. (2009). Insecure attachment, dysfunctional attitudes, and low self-

esteem predicting prospective symptoms of depression and anxiety during

adolescence. Journal of Clinical Child and Adolescent Psychology, 38(2), 219–231.

doi:10.1080/15374410802698396

Li, A., Shaffer, J., & Bagger, J. (2015). The psychological well-being of disability caregivers:

examining the roles of family strain, family-to-work conflict, and perceived supervisor

support. Journal of Occupational Health Psychology, 20(1), 40-49. doi:

10.1037/a0037878

Lobchuk, M. M. (2006). Concept analysis of perspective-taking: meeting informal caregiver

needs for communication competence and accurate perception. Journal of Advanced

Nursing, 54(3), 330-341. 10.1111/j.1365-2648.2006.03815.x

Lopez, J., Lopez-Arrieta, J., & Crespo, M. (2005). Factors associated with the positive

impact of caring for elderly and dependent relatives. Archives of Gerontology and

Geriatrics, 41(1), 81–94. doi: 10.1016/j.archger.2004.12.001

Lyons, K.S., Sayer, A.G., Archbold, P.G., Hornbrook, M.C., & Stewart, B.J. (2007). The

enduring and contextual effects of physical health and depression on care-dyad

mutuality. Research in Nursing and Health, 30(1), 84–98. doi: 10.1002/nur.20165

Marsh, H. W., Scalas, L. F., & Nagengast, B. (2010). Longitudinal tests of competing factor

structures for the Rosenberg Self-Esteem Scale: Traits, ephemeral artifacts, and

stable response styles. Psychological Assessment, 22(2), 366–381. doi:

10.1037/a0019225

McMullen, T., & Resnik, B. (2013). Self-esteem among nursing assistants: reliability and

validity of the Rosenberg self-esteem scale. Journal of Nursing Measurement, 21(2),

335-344. doi: 10.1891/1061-3749.21.2.335 .

Meurer, S. T. Luft, C. B., Benedetti , G. Z., & Mazo (2012). Validade de construto e

consistência interna da escala de autoestima de Rosenberg para uma população de

idosos brasileiros praticantes de atividades físicas (Construct validity and reliability in

Rosenberg’s self-esteem scale for Brazilian older adults who practice physical

activities). Motricidade, 8(4), 5-15. doi: 10.6063/motricidade.8(4).1548

Michelson, W., & Tepperman, L. (2003). Focus on home: what time-use data can tell us

about caregiving to adults. Journal of Social Issues, 59(3), 591-610. doi:

10.1111/1540-4560.00079

Minnes, P., Woodford, L., Passey, J. (2007). Mediators of well-being in ageing family carers

of adults with intellectual disabilities. Journal of Applied Research in Intellectual

Disabilities, 20(6), 539–552. doi: 10.1111/j.1468-3148.2007.00364.x

Moen, P., Robinson, J. & Dempster-McClain, D. (1995). Caregiving and women's well-being:

A life course approach. Journal of Health and Social Behavior, 36(3), 259-273. doi:

10.2307/2137342

Molinari, V., & Khanna, P. (1981). Locus of control and its relationship to anxiety and

depression. Journal of Personality Assessment, 45(3), 314–319.

doi:10.1207/s15327752jpa4503_14

Molyneux, G.J., McCarthy, G.M., McEniff, S., Cryan, M., & Conroy, R. M. (2008). Prevalence

and predictors of carer burden and depression in of patients referred to an old age

psychiatric service. International Psychogeriatrics, 20(6,), 1193–1202.

doi:10.1017/S1041610208007515

Mutkins, E., Brown, R. F., & Thorsteinsson, E. B. (2011). Stress, depression, workplace and

social supports and burnout in intellectual disability support staff. Journal of

Intellectual Disability Research, 55(5), 500–510. doi: 10.1111/j.1365-

2788.2011.01406.x

National Audit Office. (2014). Adult social care in England: overview. Retrieved from:

http://www.nao.org.uk/wp-content/uploads/2015/03/Adult-social-care-in-England-

overview.pdf

Neufeld, A., & Harrison, M. J. (1995). Reciprocity and social support in caregiver’s

relationships: Variations and consequences. Qualitative Health Research, 5(3), 348 –

365. doi: 10.1177/104973239500500306

Orth, U., Robins, R. W., & Meier, L. L. (2009). Disentangling the effects of low self-esteem

and stressful events on depression: Findings from three longitudinal studies. Journal

of Personality and Social Psychology, 97(2), 307–321. doi:10.1037/a0015645

Orth, U., Robins, R. W., & Widaman, K. F. (2012). Life-span development of self-esteem and

its effects on important life outcomes. Journal of Personality and Social Psychology,

102(6), 1271–1288. doi: 10.1037/a0025558

Otero, P., Smit, F., Cuijpers, P., Torres, A., Blanc, V., & Vázquez, F. L. (2015). Long-term

efficacy of indicated prevention of depression in non-professional caregivers:

randomized controlled trial. Psychological Medicine, 45(7), 1401-1412. doi:

10.1017/S0033291714002505

Pearlin, L. I., & Schooler, C. (1978). The structure of coping. Journal of Health and Social

Behavior, 19(2), 2-21. doi: 10.2307/2136319

Pelletier, P. M., Alfano, D. P., & Fink, M.P. (1994). Social support, locus of control and

psychological health in family members following head or spinal cord injury. Applied

Neuropsychology, 1(1-2), 36-44. doi: 10.1207/s15324826an0101&2_8

Phillips, A. C., Gallagher, S., Hunt, K., Der, G., & Carroll, D. (2009). Symptoms of depression

in non-routine caregivers: The role of caregiver strain and burden. British Journal of

Clinical Psychology, 48(4), 335–346.doi:10.1348/014466508X397142

Presson, P. K., & Benassi, V. A. (1996). Locus of control orientation and depressive

symptomatology: A meta-analysis. Journal of Social Behavior and Personality, 11(1),

201–212. Retrieved from: http://content.ebscohost.com/ContentServer.asp?

T=P&P=AN&K=9610284953&S=L&D=s3h&EbscoContent=dGJyMNXb4kSeqLE40dv

uOLCmr02eprBSsa24TK

%2BWxWXS&ContentCustomer=dGJyMPGtt0qzprFNuePfgeyx7oHu1%2B56

Rodakowski, J., Skidmore, E. R., Rogers, J. C., & Schulz, R. (2012). Does social support

impact depression in caregivers of adults ageing with spinal cord injuries? Clinical

Rehabilitation, 27(6) 565–575. doi: 10.1177/0269215512464503.

Randall, D. M., & O’Driscoll, M. P. (1997). Affective versus calculative commitment: Human

resource implications. Journal of Social Psychology, 137(5), 606–617. Retrieved

from: http://content.ebscohost.com/ContentServer.asp?

T=P&P=AN&K=9711110260&S=R&D=pbh&EbscoContent=dGJyMMvl7ESep7U4v

%2BvlOLCmr02ep7JSsaa4SLOWxWXS&ContentCustomer=dGJyMPGtt0qzprFNue

Pfgeyx7oHu1%2B56

Reda, A. A. (2011). Reliability and Validity of the Ethiopian Version of the hospital anxiety

and depression scale (HADS) in HIV infected patients. PLoS ONE, 6(1), 1-6. doi:

10.1371/journal.pone.0016049

Rezende, T. C. B., Coimbra, A. M. V., Costallat, L. T. L. & Coimbra, I. B. (2010). Factors of

high impacts on the life of caregivers of disabled elderly. Archives of Gerontology and

Geriatrics 51(1). 76–80. doi: 10.1016/j.archger.2009.08.003

Rhoades, L., Eisenberger, R., & Armeli, S. (2001). Affective commitment to the organization:

The contribution of perceived organizational support. Journal of Applied Psychology,

86(5), 825–836. doi: 10.1037/0021-9010.86.5.825

Rhoades, L., & Eisenberg, R. (2002). Perceived organizational support: a review of the

literature. Journal of Applied Psychology, 87(4), 698-714. doi: 10.1037//0021-

9010.87.4.698

Riketta, M. (2004). Does social desirability inflate the correlation between self-esteem and

anxiety? Psychological Reports, 94(3), 1232–1234. doi: 10.2466/PR0.94.3.1232-

1234

Roddenberry, A., & Renk, K. (2010). Locus of control and self-efficacy: Potential mediators

of stress, illness, and utilization of health services in college students. Child

Psychiatry & Human Development, 41(4), 353–370. doi:10.1007/s10578-010-0173-6

Rose, J., Mills, S., Silva, D., & Thompson, L. (2013). Client characteristics, organizational

variables and burnout in care staff: The mediating role of fear of assault. Research in

Developmental Disabilities, 34(3), 940–947. doi: 10.1016/j.ridd.2012.11.014

Rosenberg, M. (1965). Society and the adolescent self-image. Princeton, NJ: Princeton

University Press.

Rotter, J. B. (1966). Generalized expectancies for internal versus external control of

reinforcement. Psychological Monographs, 80(1), 1–28. Retrieved from:

http://www.soc.iastate.edu/sapp/soc512Rotter.pdf

Sánchez-Pérez, N., Fuentes, L. J., Jolliffe, D., & González-Salinas, C. (2014). Assessing

children’s empathy through a Spanish adaption of the basic empathy scale: parents

and child’s report forms. Frontiers in Psychology, 5, 1-13. doi:

10.3389/fpsyg.2014.01438

Seligman, M. E. P, (1991). Learned optimism. Knopf, New York

Shore, L. M., & Shore, T. H. (1995). Perceived organizational support and organizational

justice. In R. S. Cropanzano & K. M. Kacmar (Eds.), Organizational politics, justice,

and support: Managing the social climate of the workplace (pp. 149–164). Westport,

CT: Quorum

Silva, A. L., Teixeria, H. J., Teixeira, M. J. C., & Freitas, S. (2013). The needs of informal

care givers of elderly people living at home: an integrative review. Scandinavian

Journal of Sciences, 27, 792-803. doi:10.999/scs.12019.

Skills for Care. (2015). The state of the adult social car sector and workforce in England.

Retrieved from: http://www.skillsforcare.org.uk/Document-library/NMDS-SC,-

workforce-intelligence-and-innovation/NMDS-SC/State-of-2014-ENGLAND-WEB-

FINAL.pdf

Sowislo, J. F., & Orth, U. (2013). Does low self-esteem predict depression and anxiety? A

meta-analysis of longitudinal studies. Psychological Bulletin, 139(1), 213–240.

doi:10.1037/a0028931

Sowislo, J. F., Orth, U., & Meier, L. (2014). What constitutes vulnerable self-esteem?

Comparing the prospective effects of low, unstable, and contingent self-esteem on

depressive symptoms. Journal of Abnormal Psychology, 123(4), 737-753. doi:

10.1037/a0037770

Taggart, L., Truesdale-Kennedy, M., Ryan, A., & McConkey, R. (2012). Examining the

support needs of ageing family carers in developing future plans for a relative with an

intellectual disability. Journal of Intellectual Disabilities, 16(3), 217-234. doi:

10.1177/1744629512456465

Taylor, B., & Barling, J. (2004). Identifying sources and effects of carer fatigue and burnout

for mental health nurses: a qualitative approach. International Journal of Mental

Health Nursing, 13(2), 117–125. doi: 10.1111/j.1445-8330.2004.imntaylorb.doc.x.

The National Care Forum. (2015). Personnel statistics report: survey of NCF member

organisations one of the largest workforce surveys in the care sector. Retrieved from:

file:///C:/Users/Alex/Downloads/PS%20report%202015.pdf

Triggle, N. (2014). Home care services 'close to crisis'. Retrieved from:

http://www.bbc.co.uk/news/health-30278034.

Triggle, N. (2015). Home care 'being short changed'. Retrieved from:

http://www.bbc.co.uk/news/health-31700664

Wang, Y. N., Shyu, Y. I, L., Chen, M. C; & Yang, P. S. (2011). Reconciling work and family

caregiving among adult-child family caregivers of older people with dementia: effects

on role strain and depressive symptoms. Journal of Advanced Nursing, 67(4), 829-

840. doi: 10.1111/j.1365-2648.2010.05505.x

Wibberley, G. (2013). The problems of a ‘dirty work place’ in domiciliary care. Health and

Place, 21, 156-162. doi: 10.1016/j.healthplace.2013.01.014

Zeytinoglu, I. U., Denton, M., Davies, S., & Plenderleith, J. M. (2009). Casualized

employment and turnover intention: Home care workers in Ontario, Canada. Health

Policy, 91(3), 258-268. doi: 10.1016/j.healthpol.2008.12.004

Zigmond, A.S., & Snaith, R. P (1983). The Hospital Anxiety and Depression Scale. Acta

Psychiatrica Scandinavica, 67(6), 361–70.

Zunzunegui, M. V., Llácer, A., & Béland, F. (2002). The role of social and psychological

resources in the evolution of depression in caregivers. Canadian Journal on Aging,

21, 357–369. doi:10.1017/S0714980800001689