Abstract
Background
Food service staff are integral to delivery of quality food in aged care homes yet measurement of their satisfaction is unable to be performed due to an absence of a valid and reliable questionnaire. The aim of this study was to develop and perform psychometric testing for a new Food Service Satisfaction Questionnaire developed in Australia specifically for use by food service staff working in residential aged care homes (Flinders FSSQFSAC).
Methods
A mixed methods design utilizing both a qualitative (in-depth interviews, focus groups) and a quantitative approach (cross sectional survey) was used. Content validity was determined from focus groups and interviews with food service staff currently working in aged care homes, related questionnaires from the literature and consultation with an expert panel. The questionnaire was tested for construct validity and internal consistency using data from food service staff currently working in aged care homes that responded to an electronic invitation circulated to Australian aged care homes using a national database of email addresses. Construct validity was tested via principle components analysis and internal consistency through Cronbach's alpha. Temporal stability of the questionnaire was determined from food service staff undertaking the Flinders FSSQFSAC on two occasions, two weeks apart, and analysed using Pearson's correlations.
Results
Content validity for the Flinders FSSQFSAC was established from a panel of experts and stakeholders. Principle components analysis revealed food service staff satisfaction was represented by 61-items divided into eight domains: job satisfaction (α=0.832), food quality (α=0.871), staff training (α=0.922), consultation (α=0.840), eating environment (α=0.777), reliability (α=0.695), family expectations (α=0.781) and resident relationships (α=0.429), establishing construct validity in all domains, and internal consistency in all (α>0.5) except for “resident relationships (α=0.429). Test-retest reliability coefficients ranged from 0.276 to 0.826 dependent on domain, with test-retest reliability established in seven domains at r>0.4; an exception was “reliability at r=0.276.
Conclusions
The newly developed Flinders FSSQFSAC has acceptable validity and reliability and thereby the potential to measure satisfaction of food service staff working in residential aged care homes, identify areas for strategic change, measure improvements and in turn, improve the satisfaction and quality of life of both food service staff and residents of aged care homes.
Key words: Reliability, validity, food, satisfaction, elderly, nursing home, questionnaire
Introduction
According to the most recent data, the number of older persons in Australia, and in residential aged care homes is increasing. The average age of residents is 84.5 years (1) and in the past 20 years, the percentage of Australians aged 85 years and over has increased by 148% (2) with the number of admissions into a permanent aged care home increasing by 4.7% in the last financial year (3). In 2014 there were 2688 aged care homes in Australia (4) providing accommodation, care assistance and the sole source of nutrition for residents, 91% of whom will live there for the remainder of their life (5).
Residents living in aged care homes are dependent on the facility to provide them with all of their meals. The rate of malnutrition in aged care homes is high, with 40-70% of residents being malnourished (6, 7). Inadequate food intake is the biggest risk factor for malnutrition (8). In aged care homes, the meals residents are provided with can greatly influence the type and amount of food they consume which in turn influences their health and longevity (9), and as such, the food service is extremely important (10). The quality of food in aged care homes is reliant on the food service that produces the meals. The satisfaction of food service staff, a multidimensional area, can influence the satisfaction, health outcomes and quality of life of residents by affecting the quality of food produced (6).
The food service industry, particularly in the aged care sector, has a high turnover of staff as well as high levels of absenteeism (11). Disengagement with their job or employer as well as burnout and stress from heavy workloads can lead to unreliable staff (12). The unpredictability of staffing can have a considerable impact on the way food services are administered as well as the morale and satisfaction of staff. Being able to establish the satisfaction or dissatisfaction of staff would enable changes to be put into place and food service staff to have the opportunity to improve satisfaction and the food served to residents (13), a beneficial outcome for staff, residents and management alike.
An effective way to measure satisfaction among food service staff in aged care homes is through a questionnaire. Questionnaires are effective tools for collecting objective information about a range of factors including people's knowledge, perspectives and behaviours (14, 15). While there are food service satisfaction questionnaires for residents in aged care homes (16, 17, 18, 19, 20), there are currently no valid and reliable food service satisfaction questionnaires for food service staff working in the residential aged care setting.
In order for a questionnaire to be useful, it needs to be both reliable and valid (21). Reliability can be defined as the amount to which measurements conducted at different time points, by different methods or by different people are similar (22) or the ability of an instrument to produce the same results when the test is repeated. One such test for reliability is internal consistency and this is the most widely used measure of reliability (23). Internal consistency measures how closely related items are in a scale or group and is measured via Cronbach's alpha. Another test of reliability is test-retest reliability (or temporal stability), whereby a test is conducted in the same population twice, after an amount of time conducive to those completing the questionnaire being unable to recall their time one response yet not have any significant events or activities affecting responses (24).
There are two common types of validity that should be determined when establishing a questionnaire: content and construct validity. Content validity is the accuracy to which the questionnaire measures what it set out to measure as determined by a group of experts in the field. The experts assess each item in a questionnaire and investigate whether the content includes all the relevant information (22) as well as the layout and format. Construct validity is the degree to which the questionnaire measures an abstract variable that cannot be directly observed, in this case, satisfaction (25).
This study aims to describe the development and psychometric testing (content and construct validity, internal consistency and temporal stability) of a new Food Service Satisfaction Questionnaire (Food Service Aged Care) developed in Australia specifically for use by food service staff working in residential aged care homes (Flinders FSSQFSAC).
Materials and Methods
Phase 1: Initial Questionnaire Development and Expert Review
Initially, research was conducted to identify the domains and items for the questionnaire. Focus groups and interviews were conducted with relevant stakeholders to identify themes they believed were important. Stakeholders came from two sources. Twenty-nine aged care cooks and chefs from Victoria and New South Wales attended a three-day educational workshop (Maggie Beer Foundation ‘Creating an Appetite for Life' program) and participated in one of three focus groups of 9-10 participants that were run concurrently lasting 35-50 minutes. In addition, qualitative one-on-one interviews were undertaken with employees of two non-profit aged care homes in Adelaide, South Australia (n=4) to establish whether perceptions were similar between states. These interviews lasted between 38-60 minutes. Recordings were transcribed verbatim by a member of research team. Using a phenomenological approach, the focus groups and interviews were coded and extracted by two of the research team to establish nine themes. Questionnaires were also identified in the literature (16, 17, 18, 19, 20), with items and domains extracted and collated with the themes identified from the qualitative research.
A team of researchers with experience in food service, nutrition, aged care and psychology met and the Flinders FSSQFSAC was developed. The Flinders FSSQFSAC was designed to be self-administered via an online survey tool ‘SurveyMonkey' (26). The initial version of the questionnaire, before pilot testing and feedback from the expert panel, utilised a five-point Likert scale (1=very dissatisfied; 5=very satisfied) based on the Minnesota Satisfaction Questionnaire (27), which is widely used in different population groups (28, 29, 30, 31). The meaning of each of these responses was clarified at the start of the Flinders FSSQFSAC to add validity and reliability to the scale (32).
The questionnaire was tested for content validity, through the consultation of experts. The Flinders FSSQFSAC was sent out to a convenience sample of n=6, including food service dietitians (n=3), aged care food service managers (n=2) and a head chef (n=1). The experts were asked to give feedback on the content and layout of the questionnaire and the test version of the Flinders FSSQFSAC was finalised. This version was sent to the participants involved in the focus groups and interviews from phase one of the project. These participants were asked to pilot the test Flinders FSSQFSAC and complete a feedback questionnaire within seven days of receipt. Once all feedback was received final amendments were made to the Flinders FSSQFSAC and subsequent psychometric testing performed using this version.
Phase 2: Psychometric testing
The Flinders FSSQFSAC that underwent testing for construct validity included the items from the above development phase, an importance scale and demographic questions. The participants could answer “very dissatisfied, “dissatisfied, “neutral, “satisfied, “very satisfied, “not applicable and “I don't know. For the purpose of analyses, the “not applicable and “I don't know responses were treated as missing data. The questionnaire was estimated to take 20-30 minutes to complete.
For psychometric testing of the Flinders FSSQFSAC (construct validity, internal consistency) participants were recruited from a database of Australian residential aged care homes obtained from a reputable marketing agency as well as through a newsletter promoted on the Maggie Beer Foundation website. To be included, participants had to be ≥18 years old, be working at an aged care home, have worked there for at least 6 months, speak fluent English and have computer and internet access.
To assess temporal stability, participants were recruited from a two day educational workshop (Maggie Beer Foundation ‘Creating an Appetite for Life' program) and were required to complete the Flinders FSSQFSAC on two occasions, 14 days apart, prior to attending the event in line with the literature (33).
Statistical analysis
For the purpose of statistical analysis the importance scale and demographic questions were not included. To establish construct validity, exploratory factor analysis (principle components analysis) with varimax rotation was used to group items together that correlated highly with each other, forming ‘factors' (34). Varimax rotation was chosen as it reduces the complexity of the factors by making the large loadings larger and the smaller loadings smaller, resulting in data that is easier to interpret and report. This minimises the number of variables that have high loadings on each factor (24). The data was deemed suitable via consultation of the correlation matrix, a Kaiser-Meyer-Olkin value above 0.6 (35) and a significant Bartlett's test of sphericity (at p <0.05) (24). The number of factors to retain was based on eigenvalues <1.0 (24) and through consultation of the scree plot using Catell's scree test (36).
As part of the analysis, any statements that were grouped within a factor but had a high correlation with one or more factors were omitted (34). Also, any statements that crossloaded on two factors with factor loadings greater than 0.4 were removed (34). Finally, any statements loading individually on a factor were removed as per researcher judgement (34).
The reliability of the questionnaire was assessed through Cronbach's alpha and test-retest reliability. Cronbach's alpha coefficients of α≥0.5 was considered reliable for a questionnaire in the development phase (37). Coefficients of α≥0.7 were considered excellent for the Flinders FSSQFSAC as this is the recommendation for acceptable reliability of an established questionnaire (37). The temporal stability was estimated using Pearson's correlation, performed between the mean score of time 1 (the first time partipants completed the Flinders FSSQFSAC) and time 2 (the second time participants completed the Flinders FSSQFSAC) for each domain. Correlation coefficients r>0.8 were considered strong, coefficients r=0.4-0.8 were considered moderate and coefficients r<0.4 were considered poor (38).
All methods for this project were approved by the Flinders University Social and Behavioural Research Ethics Committee (Project number 6929) and participants provided written informed consent.
Results
Questionnaire development
Deductive analysis of data collected from the focus groups with stakeholders produced a total of six themes which were considered for inclusion in the Flinders FSSQFSAC. Qualitative themes are illustrated below using quotes from focus group stakeholders, who are referred to using code numbers. Subsequent interviews with food service staff from South Australia determined no new themes and saturation was considered to be achieved. A literature review determined that there was no existing questionnaire that measured food service satisfaction for food service staff in aged care homes although some were identified for completion by residents. Themes unique to this study and not identified in questionnaires for use by residents were ‘meeting expectations of others', ‘attitudes towards food services' and ‘relationships with residents'. Themes common to this study and questionnaires identified for residents were ‘dining room atmosphere', ‘meal/ food quality' and ‘choice'. Unique themes from questionnaires identified for residents included ‘overall satisfaction', ‘work simplification' and ‘using meal schedules' (Table 1). Informed by the above themes, a total of nine questionnaire domains, plus one ‘overall' food service satisfaction item were created for the Flinders FSSQFSAQ. At the end of each of the nine domains an ‘overall' satisfaction item was included to measure participant's overall satisfaction with that domain of food service. At this stage of development the questionnaire consisted of a total of 65 items in addition to 11 demographic questions and an overall importance item (rated as not important at all, not very important, neutral, important, extremely important) for each of the nine domains.
Table 1.
Collation of foodservice satisfaction themes sourced from literature and focus groups with stakeholders used to create the content of the Flinders Foodservice Satisfaction Questionnaire (Flinders FSSQFSAC)
| Themes regarding foodservice satisfaction | Literature | Stakeholder focus groups |
|---|---|---|
| Meeting expectations of others | X | |
| Staff attitudes towards foodservices | X | |
| Relationships with residents | X | |
| Dining room/Eating Environment | X | X |
| Meal/food quality | X | X |
| Choice | X | X |
| Overall satisfaction | X | |
| Using meal schedules | X | |
| Work simplification | X |
The first domain, ability to provide high quality meals (1FOOD) sourced items from both qualitative data and the literature (Table 2). Items regarding quality definitions such as temperature, taste and presentation were taken from the qualitative research, particularly the ‘Meal/food quality' theme, which involved stakeholder definitions of quality such as “the goodness in it (FG3), and also their ability to accommodate dietary needs and ensure resident's “[felt] comfortable with the plate in front of [them] (FG1).
Table 2.
Overview of the three sources of content for the Flinders Test- Foodservice Satisfaction Questionnaire (Flinders FSSQFSAC); existing literature, qualitative research, and consultation with experts and focus group participants
| Test-FSSQ Content Domain | Item total (n) | Literature themes | SourceStakeholder focus group themes | Items added from consultation with stakeholders and expert panel (n) |
|---|---|---|---|---|
| 1FOOD | 10 | ‘Enjoying food and food service’ ‘Meal quality and enjoyment’ ‘Food quality’ ‘Using aesthetics’ ‘The meals taste good ’ | • | - |
| 2EATE | 8 | ‘Providing good food service ’ ‘Satisfaction with the physical environment’ ‘Creating dining room atmosphere’ | • | 2 |
| 3CHOI | 5 | ‘Exercising Choice’ ‘Enjoying food and foodservice’ ‘Autonomy’ ‘Understanding resident rights’ ‘Choice’ | • | 2 |
| 4RELI | 8 | ‘Exercising choice’ ‘Using meal schedules’ ‘Meals are served on time ’ item | • | 2 |
| 5OPPO | 5 | - | • | - |
| 6EXPE | 9 | - | • | 2 |
| 7RELA | 14 | ‘Providing good food service’ ‘Staff consideration’ ‘Staff/service issues’ | • | 1 |
| 8WORK | 10 | ‘Work simplification’ | • | 4 |
| 9TRAI | 10 | ‘Providing good food service – positive view’; item: ‘the kitchen staff here care about the food they serve’ ‘Staff/service issues’; item ‘the staff who deliver my menus are helpful’ | • | 2 |
| Overall foodservice satisfaction | 1 | ‘Overall satisfaction’ items | - | - |
The ability to provide a high quality eating environment (2EATE) domain was adapted largely from the ‘Dining room/ Eating environment theme' from this study, as well as from existing questionnaires for completion by residents (Table 2). Items focused on both physical and social aspects of the dining experience as focus groups with stakeholders identified that both were important for satisfaction, but social aspects are usually ignored; “you got a negative attitude, the resident's gonna (sic) have a negative attitude (FG3).
The ability to provide enough choice (3CHOI) domain takes into account the ability to provide food choice as well as where residents eat. Items were developed using the ‘choice' theme from the focus groups with stakeholders and existing questionnaires for completion by residents (Table 2). This domain measures the food service's ability to provide choice which gives resident's “security (FG5), but not put pressure on the aged care home; as one focus group participant explained, “as long as they've got a choice… because you don't have a choice all the time at home either (FG10).
The domain ability to provide a reliable food service (4RELI) was expanded from the ‘choice' theme and items took into account food service dependability. Focus group participant explained that they would be in “troule" (FG11) if they changed the menu without notifying residents. The ability to provide a reliable food service was supported by existing questionnaires (Table 2).
The five items on the opportunity for consultation about the food service (5OPPO) domain were adapted from the focus groups with stakeholders; particularly the ‘Staff attitude towards food service' theme, as a focus group participant stated, “they don't consider us care providers (FG13) (Table 2).
The meeting expectations of others (6EXPE) domain in the Flinders FSSQFSAC was split into expectations of (i) families and (ii) management as one participant explained, “whenever they make decisions it's always for the clinical side - always (FG23) in regards to their ability to make resident's families, and aged care home management happy with the food service (Table 2).
The relationships with others (7RELA) domain was split into relationships with (i) management; (ii) other staff and (iii) residents. This domain focused on how valued and supported staff felt in their aged care home by management and staff. As focus group participants explained, “its catering [staff] versus care [staff] (FG2) and food services were not considered “important (FG4) by management. This affected their ability to provide high quality food services as, “there's always problems about the nurses - how they slop [food] on the plate (FG13). The items from the relationships with residents section were influenced by the ‘relationships with residents theme' from the focus groups with stakeholders. Staff were satisfied if the food service was resident centred and promoted “independence (FG23). This domain was also influenced by domains from existing questionnaires (Table 2).
The work (food service) environment (8WORK) domain was created using an existing questionnaire and data from the focus group participants (Table 2). It assessed satisfaction with job role, tasks and resources provided in their aged care home; as one focus group participant explained, “unfortunately that is the reality of it... it's all about the budget (FG2).
The training in provision of high quality food service (9TRAI) domain was separated into staff that were (i) directly and (ii) indirectly involved (e.g. nurses) in food services and had eight items. This domain stemmed from the ‘staff attitudes towards food service staff' theme where lack of training was identified as a barrier in focus groups to good quality food services. This was supported by the existing questionnaires for completion by residents (Table 2).
Finally, one overall food service satisfaction item was included at the end of the questionnaire consistent with the literature and as a method to assess relationships between each domain and overall food service satisfaction (Table 2).
Feedback from the expert panel indicated that the Flinders test- FSSQFSAC items made sense and were relevant and that no items were unnecessary. It was reported that it should be made available online and in printed format and four experts would use the Flinders test- FSSQFSAC in their own setting. Based on feedback, an introduction with definitions was added to the start of the Flinders test- FSSQFSAC and 11 additional items were added to the questionnaire itself; two items to domain two; one item to domain three; two items to domain four; two items to domain six; one item to domain seven; one item to domain eight and two items to domain nine - see Table 2.
Final feedback from nine stakeholders (from the phase one focus group group) was positive in terms of comprehension and format. Based on feedback, a further four items were added to the Flinders test- FSSQFSAC making it a total of 80 items; one to domain three and three items to domain eight - see Table 2. This version of the Flinders test- FSSQFSAC was then completed by 18 stakeholders (from the original participants of the focus groups) as a final pilot prior to psychometric testing and all nine domains were rated as ‘important' or ‘very important' by 91-100% of stakeholders. Two domains ‘5OPPO' and ‘7RELA' were rated as ‘neutral' by a total of one or two stakeholders respectively. Six items were ranked as ‘N/A' or TDK by a maximum of two stakeholders, thus all items were valid for 88-100% food service staff.
Psychometric testing
Three hundred and eighty one people participated in the testing of the 80 item Flinders test- FSSQFSAC, with 265 meeting the inclusion criteria. Sixty-six per cent were male and 52% were between 40 and 54 years old. The job titles of the participants included: food service managers (34%), food service managers with chef/cook training (31%), head chefs (15%), chefs (4%), cooks (15%) and kitchen hands (2%). A summary of the sample can be found in Table 3. Twentynine participants completed the test-retest component. For the purposes of psychometric testing all 80 items minus the overall food service satisfaction item were included.
Table 3.
Characteristics of participants contributing data to the psychometric testing of the Flinders Foodservice Satisfaction Questionnaire (Flinders FSSQFSAC)
| Characteristic | n | % |
|---|---|---|
| Gender | ||
| Male | 90 | 34 |
| Female | 175 | 66 |
| Age | ||
| 18-24 years | 3 | 1 |
| 25-39 years | 50 | 19 |
| 40-54 years | 138 | 52 |
| 55-64 years | 67 | 25 |
| >65 years | 7 | 3 |
| Education level | ||
| Secondary school (not completed) | 38 | 14 |
| Secondary school (completed) | 36 | 14 |
| TAFE/Trade qualification | 149 | 56 |
| University degree | 26 | 10 |
| Postgraduate qualification | 16 | 6 |
| Job title | ||
| Food service /hotel services/catering manager/supervisor | 90 | 34 |
| Food service /hotel services/catering manager/supervisor | 81 | 31 |
| with cook/chef training | ||
| Head chef | 40 | 15 |
| Chef | 11 | 4 |
| Cook | 39 | 14 |
| Kitchen hand | 4 | 2 |
| Experience in a RAC home | ||
| <12 months | 13 | 5 |
| 1-5 years | 71 | 27 |
| >5 years | 181 | 68 |
| Facility type | ||
| Not for profit | 165 | 62 |
| Private | 80 | 30 |
| Government | 20 | 8 |
Abbreviations: n=number; TAFE=technical and further education; RAC=residential aged care.
Construct validity
Seventeen factors were identified based on eigenvalues >1.0, however a nine factor solution was consistent with Catell's scree test, therefore nine domains were extracted from the principle components analysis. A nine-factor solution explained a total of 61% of the variance however due to a poor test-re test score of the item on the ninth factor (0.010, p=0.958), this item was removed which resulted in an eight-factor solution. This eight-factor solution resulted in the deletion of 18 items from the Flinders test- FSSQFSAC, culminating in a 62-item questionnaire with eight new factors exploring job satisfaction, food quality, staff training, consultation, eating environment, reliability, family expectations and resident relationships, explaining 59% of the variance. This version of the questionnaire, the Flinders FSSQFSAC, was used in the remainder of the data analysis. Factor correlations as well as the individual variations of each factor are provided in Table 4.
Table 4.
Correlations between each questionnaire statement and factors of the Flinders Foodservice Satisfaction Questionnaire (Flinders FSSQFSAC)
| Statements | Factors | ||||||||
|---|---|---|---|---|---|---|---|---|---|
| Jobsatisfaction | Food quality | Staff training | Consultation | Eatingenvironment | Reliability | Familyexpectations | Residentrelationships | Positivepromotion | |
| With how supported you feel in your food service role by management | 0.773 | ||||||||
| That overall, there is a positive relationship between management and food service staff | 0.728 | ||||||||
| With your role in the food service | 0.727 | ||||||||
| With the tasks you have to perform on a daily basis | 0.718 | ||||||||
| With the way management communicates decisions made about the food service | 0.708 | ||||||||
| With the way management makes decisions about the food service | 0.699 | ||||||||
| That there is enough guidance to provide high quality food services | 0.679 | ||||||||
| With the way management handles any concerns about the food service | 0.677 | ||||||||
| That you have enough time to complete your tasks | 0.673 | ||||||||
| With the resources available to you to perform tasks to the best of your ability | 0.671 | ||||||||
| Overall with the environment you work in | 0.644 | ||||||||
| That resources are shared appropriately between departments (e.g catering, clinical) | 0.571 | ||||||||
| That, overall, staff other than management have a positive relationship with food service staff | 0.565 | ||||||||
| That you are able to work with dietitian/s when required | 0.500 | ||||||||
| That food safety regulations are easily understood | 0.424 | ||||||||
| That the food service environment is efficient | 0.408 | ||||||||
| That meals served taste good | 0.750 | ||||||||
| That meals served look presentable | 0.709 | ||||||||
| That meals served smell good | 0.691 | ||||||||
| That overall, your food service provides high quality meals | 0.681 | ||||||||
| That meals served meet residents’ food preferences (e.g likes and dislikes or cultural preferences) | 0.623 | ||||||||
| That the food service provides adequately for any special diets of residents (e.g gluten free, diabetic, texture modified) | 0.612 | ||||||||
| That meals served are a suitable texture | 0.548 | ||||||||
| That meals served are a suitable temperature | 0.507 | ||||||||
| That the meals served are adequately nourishing for residents | 0.502 | ||||||||
| That meals served are the suitable portion size | 0.476 | ||||||||
| That food meets cultural and/or religious needs of residents | 0.440 | ||||||||
| That residents have enough choice for foods that might not be on the menu (e.g mid- meal snacks) | 0.413 | ||||||||
| That overall, staff indirectly involved have enough training to provide high quality food service | 0.823 | ||||||||
| … be knowledgeable about the menu | 0.815 | ||||||||
| … promote the food positively | 0.812 | ||||||||
| …safely handle food | 0.782 | ||||||||
| … serve high quality meals | 0.765 | ||||||||
| … assist residents when needed | 0.548 | ||||||||
| That your opinions about the menu are listened to | 0.778 | ||||||||
| Job satisfaction | Food quality | Staff training | Consultation | Eating environment | Reliability | Family expectations | Resident relationships | Positive promotion | |
| That you have enough input into other aspects of the food service (e.g eating environment, the timing of meals) | 0.636 | ||||||||
| That your opinions about other aspects of the food service are listened to (e.g eating environment, the timing of meals) | 0.618 | ||||||||
| That your facility has an up-to-date system for recording resident’s dietary needs? That the eating environment allows residents to eat independently | 0.462 | 0.681 | |||||||
| That residents have enough choice about where they eat (e.g dining room or own room) | 0.609 | ||||||||
| That residents have enough choice about with whom they eat with | 0.603 | ||||||||
| That the eating environment has a good atmosphere (e.g peaceful, nice music, not clinical) | 0.596 | ||||||||
| That the eating environment is visually appealing | 0.557 | ||||||||
| That the staff assist residents where needed | 0.541 | ||||||||
| That the meal delivered is the meal listed on the menu | 0.651 | ||||||||
| That ‘menu’ traditions are observed (e.g fish on Friday, roast on Sunday) | 0.598 | ||||||||
| That residents are given enough warning about changes to the menu | 0.557 | ||||||||
| That almost always the food is delivered on time | 0.555 | ||||||||
| That the food service operates within budget expectations | 0.507 | ||||||||
| That overall, your food service meets the expectations of management? | 0.445 | ||||||||
| That the food service provides food which is safe (e.g free from allergens, appropriate temperature and adequately stored) | 0.442 | ||||||||
| That overall, your food service meets the expectations of families | 0.793 | ||||||||
| That families of residents’ are happy with the meals provided | 0.754 | ||||||||
| That families of resident’s are happy with other aspects of the food service (e.g the eating environment, the timing of meals) | 0.705 | ||||||||
| That families of resident’s have adequate opportunity to have input into the food | 0.678 | ||||||||
| service | |||||||||
| With the feedback the food service staff receive from residents | 0.634 | ||||||||
| That food service staff are able to interact and engage directly with residents (e.g to encourage sharing of ‘stories’) | 0.559 | ||||||||
| That overall, residents have a positive relationship with food service staff | 0.477 | ||||||||
| That overall, the food service provides enough choice for residents | 0.473 | ||||||||
| That staff directly involved in the food service have enough training to …promote the food positively | 0.525 | ||||||||
| % Of variance | 32 | 7.04 | 5.25 | 4.13 | 3.36 | 2.72 | 2.43 | 2.37 | 1.7 |
*Factor loadings >0.4 are shown
Internal consistency Internal consistency reliability (Cronbach's alpha) coefficients ranged from a=0.429 "resident relationships" to α=0.932 "job satisfaction" (see Table 5). Seven of the eight domains achieved α>0.5, while six domains achieved α>0.7, the more stringent recommendation for established questionnaires.
Table 5.
Internal consistency for all eight factors of the Flinders Foodservice Satisfaction Questionnaire (Flinders FSSQFSAC)
| Domain | Cronbach’s alpha (α) |
|---|---|
| Job satisfaction | 0.932 |
| Food quality | 0.871 |
| Staff training | 0.922 |
| Consultation | 0.840 |
| Eating environment | 0.777 |
| Reliability | 0.695 |
| Family expectations | 0.781 |
| Resident relationships | 0.429 |
| Positive promotion | n/a |
Temporal stability
Two-week test-retest coefficients ranged from r=0.267 "Reliability" to r=0.826 "Job satisfaction" (Table 6). "Job satisfaction" and "staff training" were the only scales to reach r>0.7, while "reliability" fell below r=0.4, indicating poor reliability. The "reliability" domain was the only domain not to reach statistical significance of P<0.01. As explained above, the item on the ninth factor scored r=0.010 (p=0.958) and was therefore removed.
Table 6.
Temporal stability test-retest coefficients for all nine factors of the Flinders Foodservice Satisfaction Questionnaire (Flinders FSSQFSAC)
| Domain | Test-retest coefficient(r) | P value |
|---|---|---|
| Job satisfaction | 0.826 | <0.001 |
| Food quality | 0.473 | 0.010 |
| Staff training | 0.708 | <0.001 |
| Consultation | 0.632 | <0.001 |
| Eating environment | 0.600 | 0.001 |
| Reliability | 0.276 | 0.147 |
| Family expectations | 0.615 | <0.001 |
| Resident relationships | 0.586 | 0.001 |
| Positive promotion | 0.010 | 0.958 |
The only sub-domain no longer represented by an item in the 62-items Flinders FSSQFSAC was ‘That staff directly involved in food services have enough training to provide high quality food service' Therefore, to maintain content validity, the overall item from this sub-domain was added to the final questionnaire making a final 63 item FSSQ.
Discussion
In the aged care setting it is recognised that food is the most important predictor of meal satisfaction for residents (39), translating directly to an increase in food intake and fewer negative health outcomes, ultimately improving quality of life for those in aged care homes (20). In order to have effective food service, producing good food for residents, food service staff in aged care homes need to be satisfied with their food service. The research reported in this paper confirms that the development of the Flinders FSSQFSAC, the first tool of its kind, is sufficiently valid and reliable for use in establishing the food service satisfaction of food service staff working in aged care homes. Staff satisfaction tools allow employers, in this case aged care homes, to monitor the satisfaction of their employees, and put in place changes to maintain or improve their satisfaction.
The development of the Flinders FSSQFSAC used robust qualitative methods which add to the credibility and dependability of results (40). Further, the use of qualitative research with food service staff, while of established value, is not a commonly adopted method to develop questionnaire content, only two other research groups used interviews with food service staff to develop items for perceptionbased questionnaires (41, 42). When compared with existing literature, the qualitative data collected in this study exposed three unique themes to food service satisfaction supporting the validity of utilizing this study design.
In terms of validity, this research established content validity through the consultation of a panel of experts. Further to this, construct validity was established in all domains, based on results from the factor analysis. Consequently, the questionnaire was shortened to 63 items across eight domains. As evidenced by the number and variability of the domains included in the questionnaire, this study confirms previous research findings that food service is multi-dimensional in nature (39, 43). As construct validity has been established in all domains, in the future it would be possible to extract discrete domains from the Flinders FSSQFSAC to use in aged care homes. Management would be able to use domains individually to assess selected areas of food service satisfaction among their food service staff, rather than administer the entire 63 item Flinders FSSQFSAC.
In terms of internal consistency, seven of the eight domains had Cronbach's alphas ranging between α=0.695 and α=0.932. Importantly, the high Cronbach's alpha for job satisfaction supports previous research findings that general job satisfaction is the main predictor of staff satisfaction (44). This domain also explained the highest amount of variance in the factor analysis at 32%. Factors with a smaller number of items generally have a lower reliability (24). This can be seen in this study in the "resident relationships" domain, with an α=0.429, indicating low internal consistency. To improve the internal consistency of the "resident relationships" domain, the number of items could be adjusted up or down, although adding more items to the questionnaire could increase respondent fatigue and burden. Alternatively, it might be that an increase in the sample size for the reliability testing could improve the overall findings for this domain.
The test-retest values were high in six of the eight domains. The value for the "reliability" domain was extremely low at r=0.276, while the "food quality" domain also had a low to moderate value of r=0.473, possibly as a result of the participants meeting socially prior to the second completion of the Flinders FSSQFSAC. At that time, the participants may have been influenced through discussions of their work environments, ultimately affecting their opinions from time one and therefore impacting the test-retest reliability results. It is also possible that a single item in the "reliability" and "food quality" domains had poor test-retest reliability, bringing the score down for the whole domain. To establish whether this is the case, further testing should be done on individual items, rather than each domain as a whole, possibly leading to an item being deleted, ultimately improving the test-retest reliability of the Flinders FSSQFSAC.
The sample size of this study was relatively small when compared to that of studies testing food service satisfaction questionnaires. That being said, this is the first questionnaire of its kind for food service staff in aged care and therefore there is no study for comparison using the same population. This study was able to complete a valid factor analysis as well as achieve internal consistency and test-retest reliability in most of the domains, thus the sample size was deemed adequate. Furthermore, the ideal sample size to determine internal consistency in a measurement tool is 50 participants (45), which this study exceeded.
Although this study makes an important contribution to the understanding of staff satisfaction in aged care food service, it is not without limitations. First, all participants who met the inclusion criteria were treated as food service staff during data analysis and no sub-group analysis was conducted based on job title. This was necessary to obtain a large enough sample size to undertake a factor analysis, however it is possible that the issues important to the satisfaction of staff differ depending on job title. Job titles included food service manager, head chef, chef, cook and kitchen hand and while it is acknowledged that issues impacting on satisfaction may differ, it is likely that all these staff would inform strategic changes in aged care homes and likely be pooled from an organisational perspective. Therefore, it makes sense to determine statistics on the Flinders FSSQFSAC from the combined sample, rather than by job title.
A second limitation is the possibility of social desirability bias, which has the possibility of affecting the validity of the results (46). The researchers attempted to reduce the likelihood of this occurring by making the Flinders FSSQFSAC self administered and anonymous. The length and online format of the questionnaire may have been a barrier to recruitment of participants, who may not have had access to a computer or internet, or found the 20-30 minutes it took to complete the questionnaire too burdensome. Lastly, the Flinders FSSQFSAC is only suitable for food service staff that work in food service within aged care homes; it is not suitable for individuals working in the community, home care setting or commissary kitchens, such as Meals on Wheels. This would be an area for future research, as the satisfaction of staff in these settings also likely affects the quality of food that is produced.
Conclusion
The Flinders FSSQFSAC has acceptable validity and reliability and therefore has potential to be used to measure the food service satisfaction of food service staff working in RACF. In terms of use in practice, the Flinders FSSQFSAC has the potential to influence the quality of food in aged care homes through changing the way management assess satisfaction among food service staff. Through understanding the issues experienced by food service staff, changes can be made, creating a working environment that staff want to remain in, reducing the high absentee and turnover rate common in food service (11). Also, in the future it would be possible to extract certain domains from the Flinders FSSQFSAC (available from the authors on request) to use in aged care homes as the majority of domains have been shown to be valid and reliable in their own right, hence the questionnaire is versatile to suit the needs of administrators.
In terms of future research, the Flinders FSSQFSAC should be retested to comply with the changing environment of aged care. The aged care environment is expected to change in the near future, with the ageing of the baby boomer generation. Recent research shows that ‘baby boomers' (born between 1946 and 1964) are more likely to utilise community care and either stay in their homes longer or live with adult children rather than enter into aged care homes (47). This shift to an increase in community care could mean a change in aged care and therefore a change in food services provided.
Conflict of Interest
None of the authors have any conflict of interest to declare.
Ethical standard
This study was approved by the institutional research and ethics committee.
References
- 1.Australian Government Productivity Commission. Report on government services: aged care services. 2015 [cited 2016 7th April]; Available from: http://www.pc.gov.au/research/ongoing/report-on-government-services/2015/community-services/agedcare-services/rogs-2015-volumef-chapter13.pdf.
- 2.Australian Bureau of Statistics. Feature article: Population by age and sex, Australia, States and Territories. 2015 [cited 2016 3rd February]; Available from: http://www.abs.gov.au/ausstats/abs@.nsf/Latestproducts/3101.0Feature%20Article2Jun%202015?opendocument&tabname=Summary&prodno=3101.0&issue=Jun%202015&num=&view=.
- 3.Australian Institute of Health and Welfare. Admission into aged care. 2016 [cited 2016 3rd February]; Available from: http://www.aihw.gov.au/aged-care/residentialand-home-care-2013-14/admissions/.
- 4.Australian Institute of Health and Welfare. Residential aged care and home care 2013–14 supplementary data. 2014 [cited 2016 3rd February]; Available from: http://www.aihw.gov.au/WorkArea/DownloadAsset.aspx?id=60129553641.
- 5.Australian Institute of Health and Welfare. Use of aged care services before death. 2015 [cited 2016 8th August]; Available from: http://www.aihw.gov.au/WorkArea/DownloadAsset.aspx?id=60129550411.
- 6.Gaskill D, et al. Malnutrition prevalence and nutrition issues in residential aged care facilities. Aust J Ageing. 2008;27(4):189–194. doi: 10.1111/j.1741-6612.2008.00324.x. 10.1111/j.1741-6612.2008.00324.x [DOI] [PubMed] [Google Scholar]
- 7.Dietetics Association of Australia. Preventing and treating malnutrition in older Australians pre-budget submission 2013–2014. 2013 [cited 2016 3rd February]; Available from:http://daa.asn.au/wp-content/uploads/2013/08/DAA_PreBudget-Submission_Malnutrition_Jan-2013.pdf.
- 8.Paquet C, et al. A blueprint-based case study analysis of nutrition services provided in a midterm care facility for the elderly. J Am Diet Assoc. 2003;103(3):363–368. doi: 10.1053/jada.2003.50047. PubMed PMID: 12616261. [DOI] [PubMed] [Google Scholar]
- 9.Hao Z, et al. Association between longevity and element levels in food and drinking water of typical Chinese longevity area. The journal of nutrition, health & aging. 2016;20(9):897–903. doi: 10.1007/s12603-016-0690-5. 10.1007/s12603-016-0690-5 [DOI] [PubMed] [Google Scholar]
- 10.Brown J, Nutrition through the lifecycle, 2011
- 11.Pizam A, Thornburg S. Absenteeism and voluntary turnover in Central Florida hotels: a pilot study. Int J Hosp Manage. 2000;19(2):211–217. 10.1016/S0278-4319(00)00011-6 [Google Scholar]
- 12.Deery M. Talent management, work-life balance and retention strategies. Int J Contemp Hospit Manag. 2008;20(7):792–806. 10.1108/09596110810897619 [Google Scholar]
- 13.W-H., K. The relationships among professional competence, job satisfaction and career development confidence for chefs in Taiwan. Int J Hospit Manag. 2012;31(3):1004–1011. 10.1016/j.ijhm.2011.12.004 [Google Scholar]
- 14.Oppenheim A. Questionnaire design, interviewing, and attitude measurement. St. Martin’s Press; New York: 1992. [Google Scholar]
- 15.Sapsford R. Survey Research. SAGE Publications; London: 1999. [Google Scholar]
- 16.Crogan N, Evans B. The shortened food expectations—Long-term care questionnaire: Assessing nursing home residents’ satisfaction with food and food service. J Gerontol Nurs. 2006;32(11):50–59. doi: 10.3928/00989134-20061101-08. PubMed PMID: 17112138. [DOI] [PubMed] [Google Scholar]
- 17.Crogan N, Evans B. Food expectations-long term care Spanish questionnaire: Pilot Testing with Older Mexican American Nursing Home Residents. Research in Gerontological Nursing. 2010;3:282–290. doi: 10.3928/19404921-20100330-02. 10.3928/19404921-20100330-02 PubMed PMID: 20438049. [DOI] [PubMed] [Google Scholar]
- 18.Crogan N, Evans B, Velasquez D. Measuring nursing home resident satisfaction with food and food service: initial testing of the FoodEx-LTC. J Gerontol A Biol Sci Med Sci. 2004;59(4):370–377. doi: 10.1093/gerona/59.4.m370. 10.1093/gerona/59.4.M370 PubMed PMID: 15071081. [DOI] [PubMed] [Google Scholar]
- 19.Lengyel C, et al. A questionnaire to examine food service satisfaction of elderly residents in long-term care facilities. Journal of Nutrition for the Elderly. 2004;24(2):5–18. doi: 10.1300/J052v24n02_02. 10.1300/J052v24n02_02 PubMed PMID: 15778154. [DOI] [PubMed] [Google Scholar]
- 20.Wright O, Capra S, Connelly L. Foodservice satisfaction domains in geriatrics, rehabilitation and aged care. Journal of Nutrition, Health & Aging. 2010;14:775–780. doi: 10.1007/s12603-010-0123-9. 10.1007/s12603-010-0123-9 [DOI] [PubMed] [Google Scholar]
- 21.Bailey K. Methods of social research. The Free Press; New York: 2008. [Google Scholar]
- 22.Streiner D, Norman G, Cairney J. Health Measurement Scales: A Practical Guide to Their Development and Use. Oxford University Press; Oxford, United Kingdom: 2015. 10.1093/med/9780199685219.001.0001 [Google Scholar]
- 23.Tavakol M. Making sense of Cronbach’s alpha. Int J Med Educ. 2011;2:53–54. doi: 10.5116/ijme.4dfb.8dfd. 10.5116/ijme.4dfb.8dfd PubMed PMID: 28029643, PMCID 4205511. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Pallant J, SPSS Survival Manual., 2013
- 25.Streiner D. A guide for the statistically perplexed: selected reading for cinical researchers. University of Toronto Press; Canada: 2013. [Google Scholar]
- 26.SurveyMonkey Inc., Survey Monkey, 2016
- 27.Weiss D, et al, Manual for the Minnesota Satisfaction Questionnaire., 1967
- 28.DeMicco F. The relationship of work satisfaction, organizational commitment and retirement intention of older workers in institutional foodservices. J Am Diet Assoc. 1987;47(9-B):3711. [PubMed] [Google Scholar]
- 29.Dienhart J, Gregoire M. Job Satisfaction, Job Involvement, Job Security, and Customer Focus of Quick-Service Restaurant Employees. 1993;16(2):29–43. [Google Scholar]
- 30.GhiselleB. Bai Job satisfaction, life satisfaction, and turnover intent: Among food-service managers. Cornell Hotel Restaur Adm Q. 2001;42(2):28–37. [Google Scholar]
- 31.Hancer M, George R. Job Satisfaction Of Restaurant Employees: An Empirical Investigation Using The Minnesota Satisfaction Questionnaire. J Hosp Tour Res. 2003;27(1):85–100. 10.1177/1096348002238882 [Google Scholar]
- 32.Krosnick J, Presser S, Question and questionnaire design: Handbook of survey research., 2010
- 33.Condrasky M, et al. Chefs’ Opinions of Restaurant Portion Sizes. 2007;15(8):2086–2094. doi: 10.1038/oby.2007.248. [DOI] [PubMed] [Google Scholar]
- 34.Tabachnick B, Fidell L, Using Multivariate Statistics, 2013
- 35.Kaiser H. A second generation little jiffy. Psychomietrika. 1970;35:401–415. 10.1007/BF02291817 [Google Scholar]
- 36.Catell R. The Scree Test For Number Of Factors. Multivariate Behav Res. 1966;1:245–276. doi: 10.1207/s15327906mbr0102_10. 10.1207/s15327906mbr0102_10 [DOI] [PubMed] [Google Scholar]
- 37.Nunnally J, Bernsten I, Psychometric theory, 1994
- 38.Streiner DLNG. Health measurement scales: a practical guide to their development and use. Oxford University Press; Oxford, New York: 2003. [Google Scholar]
- 39.Dube L, Trudeau E, Belanger M. Determining the complexity of patient satisfaction with food services. J Am Diet Assoc. 1994;94:394–398. doi: 10.1016/0002-8223(94)90093-0. 10.1016/0002-8223(94)90093-0 PubMed PMID: 8144806. [DOI] [PubMed] [Google Scholar]
- 40.Liamputtong P, Qualitative Research Methods., 2013
- 41.Ryan C, Ghazali H, Mohsin A. Determinants of intention to leave a nonmanagerial job in the fast-food industry of West Malaysia. Int J Contemp Hosp Manage. 2011;23(3):344–360. 10.1108/09596111111122523 [Google Scholar]
- 42.Zopiatis A, Kyprianou G, Pavlou I. Occupational challenges facing chefs: The case of Cyprus. J Qual Assur Hosp Tour. 2011;12:104–120. 10.1080/1528008X.2011.541822 [Google Scholar]
- 43.Gregoire M. Quality of patient meal service in hospitals: delivery of meals by dietary employees vs delivery by nursing employees. J Am Diet Assoc. 1994;94:1129–1134. doi: 10.1016/0002-8223(94)91132-0. 10.1016/0002-8223(94)91132-0 PubMed PMID: 7930316. [DOI] [PubMed] [Google Scholar]
- 44.Thompson P. A brief index of affective job satisfaction. GOM. 2012;37(3):275–307. [Google Scholar]
- 45.Javali S, Gudaganavar N, Raj S. Effect of varying sample size in estimation of coefficients of internal consistency. Biostatistics. 2011;2(2):1–8. [Google Scholar]
- 46.Van de Mortel T. Faking it: social desirability response bias in self-report research. Aust J Adv Nurs. 2008;25(4):40–48. [Google Scholar]
- 47.Robison J, et al. Long-term supports and services planning for the future: implications from a statewide survey of baby boomers and older adults. The Gerontologist. 2014;54(2):297–313. doi: 10.1093/geront/gnt094. 10.1093/geront/gnt094 PubMed PMID: 23990621. [DOI] [PubMed] [Google Scholar]