Gratitude, or the Positive Side of the Relationship with Patients. Development and First Validation of New Instruments: a Scale of Gratitude Perceived by Operators and a Scale of Support Offered by the Gratitude Expressed by Their Patients

The relationship with patients/users is a crucial dimension for helping professionals. It is nevertheless mostly analyzed in its negative connotation, that's to say as a possible cause, for example, of the burnout syndrome, while very few are research works that take into account the positive side of the relationship, as a resource more than a problem, and even fewer are the instruments to measure it. The aim of the present work is the development and the analysis of the characteristics of a scale to measure the gratitude expressed by patients perceived by operators and of a scale to measure the perception of support that gratitude by patients/customers offers to relieve the fatigue of daily commitment and return significance to their work. The two scales were developed and then integrated into a questionnaire including items regarding the negative side of the relationship (exceeding requests from patients) and the Customer-initiated support scale (Zimmer-mann et al., 2011). The questionnaire was filled in by 267 nurses in Piedmont, Italy. Descriptive analyses on each item, exploratory factor analysis, Cronbach's alpha, analysis of variance, and bi-variate correlations were conducted. Both scales show a one-factor solution. The independent sample t-tests highlighted differences between men and women and between people who have families to care for and people who do not. Bivariate correlations signalled a positive relationship between both the gratitude scales and the Customer-initiated support; no relationships are evident with the scale of exceeding request from patients. A limitation of the work is the dimension of the sample: a more extensive administration of the questionnaire is currently in progress that will also permit to more deeply investigate the factorial structure of the scales. The scales will allow M. Martini, D. Converso 573 filling a gap in measuring a central work issue for helping professions.


Introduction
The significance of the relationship with their patients is often mentioned by health professionals among the reasons for having chosen their job (Grandey, Fisk, & Steiner, 2005;Homburg & Stock, 2004;Wright & Pandey, 2008).
Even if largely criticized for its rhetoric allusion to the relationship with patients, the expression "helping profession" itself stringently originates from the centrality that the relational aspects of the socio-sanitary and educational professions encompass.However, peculiarly or paradoxically, the positive impact and the value as support of such relationship is hardly ever cited in the literature, while countless works have focused on its negative sides.
Specifically, several studies (Cox, Kuk, & Leiter, 1993, among others) take into account the outcomes of an over-demanding and hence negative relationship with patients (stress, burnout, intention to leave; Lee & Ashforth, 1996;Jourdain & Chênevert, 2010).As Maslach (1982) has highlighted since her early studies, burnout is the result of an imbalance between giving and receiving, that's to say, an unequal relationship between the people who offer help, support, caring, without adequate back reward, and the users of the services offered, who are often-and paradoxically-increasing their requests.This is what emerges from Dorman and Zapf's ( 2004) work, focused on exceeding requests from the users of several services.
In contrast with such unidirectional trend, we underline that the relationship with patients, students, clients, can be regarded at not only as a "request" but also as a "resource".
Requests are all the characteristics of the job and of the organizational context (such as, for example, high workload or fast paces) that require physical or psychological efforts, sometimes well beyond the worker's capacity.On the other hand, resources are all the aspects of the job and of the organizational context (a such as, for example, social support from colleagues and autonomy) that support workers in performing their job and in their professional growth.As Hobfoll (1989) states in his Conservation of Resources (COR) theory, psychological resources, in particular, are built throughout positive experiences (Zimmermann, Dormann, & Dollard, 2011).
The theoretical framework of this study, the Job Demand-Resources (JD-R) model (Bakker, Demerouti, & Schaufeli, 2003), describes the complex interactions between the resources available for workers and the requests from their job.The authors specifically outline a "dual process" that is induced by both requests and resources.The first process, due to exceeding requests, is a progressive exhaustion of workers health and the consequent reduction in productivity and increase of absenteeism.The second one-the consequence of availability of resources for the worker-is an improvement in the ability to cope with diverse demands and to achieve professional goals as well as an increase in involvement at work.
Even if the JD-R model has previously been used to analyse experiences at work of health professionalsalso in Italian studies (Lo Presti, 2013), the second, positive side of the model, has rarely considered the relationship with the patients (Martini & Converso, 2012) among resources and its possible protective function (Converso, Gattino, & Loera, 2009).Among the few research works on this topic we can mention Zimmermann and colleagues' (2011), focused on the support from the customers' positive behaviour in a service.
According to the approach of Positive Psychology (Mills & Fleck, 2013), investigating aspects that are resources for workers and thus contribute to the increase in personal and professional workers' well-being, is therefore valuable.It can have positive consequences, such as increased productivity, performance improvement, reduction of the risk of clinical errors, and better service quality and users' satisfaction (Ferrara, Converso, & Viotti, 2013).
Recently, gratitude has become a subject of interest for several Positive Psychology scholars (among others, Emmons, 2007).Gratitude is usually considered, also according to common sense, as a psychological positive characteristic related to a feeling of well-being (Toussaint & Friedman, 2008) and it is seen as a strength typical of some individuals (Peterson & Seligman, 2004).It has been regarded at and analysed as a trait, a mood, a moral virtue, an emotion.In his "The Stress of life" (1956), Selye underlined that the emotion that can highly reduce or foster stress in human relationships is precisely the feeling of gratitude.People who adopt behaviors that elicit gratitude, also receive approval thus increasing their wellbeing and self-esteem.Gratitude is, in other words, an "emphatic emotion" (Lazarus & Lazarus, 1994), that can be defined as a response to behaviors that other people perform to contribute to the wellbeing of someone and that may in turn activate analogous behaviors (McCullough, Kilpatrick, Emmons, & Larson, 2001;McCullough, Emmons, & Tsang, 2002).
No study, however, has ever addressed gratitude expressed to professionals of the services by their users or patients, i.e. gratitude received by professionals, in other words, the positive counterpart of exceedingly demanding behaviors by patients and users, the so-called Customer-related social stressors described by Dormann and Zapf (2004).
During the qualitative phase (interviews and focus groups) of the research that preceded the development of the scales presented in this work, several health professionals have testified gratitude as above described.Moreover, several health operators, when asked to describe what aspects assist them in performing their work, reported expressions of gratitude by their patients and stated that they feel rewarded and supported by such expressions.
In this way, gratitude expressed by patients is a form of social support that, as Zimmermann and colleagues (2011) remind, may be defined as a resource to buffer stress.Specifically, for the health operators we interviewed, gratitude expressed by patients is a resource because it relieves the fatigue of daily commitment and returns significance to their work.
The activity of professionals working in helping professions does not, in fact, produce any "tangible" result, that might itself be a source of satisfaction, but it is characterized by the repetition of the same gestures over and over (it is sometimes defined a "reproductive" job; Balbo, 1982), often without any manifest outcome.For health operators, not seeing the result of their work can lead to progressive disinvestment in their job, or, sometimes, even to the decision to leave (Hasselhorn et al., 2008).Therefore, the gratitude expressed by patients may be both a reward for the efforts in working and a confirmation of the expected result (i.e. the wellbeing of the patients who have expressed their gratitude).

Aim
Since the literature lacks in references and instruments focused on these aspects of the relationship with patientsaspects that can no longer be overlooked taking into account the ever-growing amount of requests and resources for health professionals-the present work is aimed at developing and starting to validate two new scales: the first one (Gratitude perceived by operators) measuring the gratitude perceived by operators as the positive side of the relationship with patients/customers; the other one (Perception of support offered by patients/customers' gratitude) measuring the perception of support that gratitude expressed by patients/customers offers to relieve the fatigue of daily commitment and to return significance to professionals' work.

Procedure and Instruments
In order to develop these instruments, 12 interviews and 2 focus-groups with 24 health operators-who are usually in relationship with patients or customers during their working hours-were conducted.The transcripts of the subjects were analyzed to identify the contents and the issues of their relationships with patients/customers, particularly focusing on the positive aspects of the relationship.As a result, we identified the theme of the perception of gratitude expressed by patients/customers towards operators and the theme of gratitude as a source of "relief" for the effort at work and of significance for their work and their commitment.Some sayings expressed by the operators interviewed were used to develop the items of patients/users' gratitude.We generated a first version of the two scales, respectively composed of 6 and 8 items, with the intent to highlight the several facets of gratitude expressed by patients or users.We chose to develop quite brief scales that can more easily be inserted into instruments to investigate several constructs.For both scales we decided to use a five-point scale ranging from 1 = I completely disagree to 5 = I completely agree, in line with Zimmermann and colleagues (2011).
We administered the scales to 40 nurses for a comprehension pre-test.Their feedback induced us to eliminate 1 item from each scale because they seemed controversial (Rattray & Jones, 2005).
The scale of Gratitude perceived by operators is now composed of 5 items and the scale of Perception of support offered by patients/customers' gratitude is composed of 7 items.They were inserted in a questionnaire that included a personal data section and the following two scales: • The Dorman & Zapf's ( 2004) Customer-related social stressors (Italian adaptation; Taddei & Vanni, 2008): Likert scale ranging from 1 = Not true at all to 5 = Completely true (e.g.: "Our patients are unable to wait").• The Zimmermann and colleagues' (2011) Customers-initiated support: Likert scale ranging from 1 = I completely disagree to 5 = I completely agree (e.g.: "The customer trusted in my competencies").

Participants
The questionnaire was completed by 267 nurses and health operators in three hospitals in a North-Western region of Italy.The participants were 85.9% women, 14.1% men; 50.6% had children and 20.8% had other family members to care for.The average age was slightly higher than 42 years (SD 8.10) and the average job tenure in the health sector was about 17 years (SD 9.04).Their average working time was of 45.5 hours per week (SD 13.4).

Data Analysis
Data were analyzed by means of the statistical package SPSS 20.After descriptive analysis (M, SD; Asymmetry, Kurtosis; Normality tests) of single items of the two scales, explorative factor analysis was conducted in order to start to study the psychometric characteristics of the scales.Moreover, we checked the reliability of each factor by Cronbach's alpha.T-test for independent samples was later carried out to test the ability of the two scales to discriminate between groups of subjects: grouping variables were socio-demographic characteristics (such as: being women, having children and having other people to care for) that might be sources of request or resources for health operators (Converso et al., 2009;Colombo, Zito, & Ghislieri, 2012).
In order to start to verify the validity of the two scales (Campbell & Fiske, 1959;Hinkin, 1998) we also calculated correlations between them and other constructs.In particular, to test divergent validity, we checked if correlation with Customer-related social stressors (Dorman & Zapf, 2004), was low or zero.Testing convergent validity was more difficult, because of the lack, to our knowledge, of an Italian validated scale to measure the positive side of the relationship with patients/costumers or similar issues.Therefore, we decided to calculate correlations between the two new scales and the Customers-initiated support scale (Zimmermann et al., 2011), even if not yet validated in Italian.
We then calculated correlations between the two scales and age, job tenure and with hours worked for week that are most common requests from workers (Colombo et al., 2012).

Gratitude Perceived by Operators Scale
Descriptive analyses (see Table 1) show that the items of the scale of Gratitude perceived by operators do not have a strictly normal distribution.All items are lightly negatively asymmetric.Besides, item 1a has a positive kurtosis value; all the others have negative values.The values of both skewness and kurtosis, however, range between −1.00 and 1.00, which are the threshold values to assume normality of the distribution (Muthén & Kaplan, 1985).In spite of this, the normality tests both of Kolmogorov-Smirnov and of Shapiro-Wilk (more powerful; Razali & Wah, 2011) induce, for all items, to reject the hypothesis of normality of the distributions.

Perception of Support Offered By Patients/Customers' Gratitude Scale
Similarly to the items of the Gratitude perceived by operators scale, descriptive analysis for the Perception of support offered by patients/customers' gratitude scale (see Table 2) shows that items do not have a strictly normal distribution.Item 4b shows a slightly positive asymmetry while all the others are lightly negatively asymmetric.Also in this case the values both of skewness and kurtosis range between −1.00 and 1.00, but the two

Gratitude Perceived by Operators Scale
As the items of the scale have no a normal distribution, we conducted exploratory factor analysis with the Generalized Least Squares extraction method.By the "Eigenvalues-greater-than-1" criterion, it resulted in a one factor solution (Table 3).The calculation of Cronbach's alpha suggested, then, to eliminate one item.We thus obtained a one-factor-solution (explained variance 69.81%) composed of 4 items, which show quite high saturations.The index of reliability, Cronbach's alpha, for this sample is .90.

Perception of Support Offered by Patients/Customers' Gratitude Scale
Exploratory factor analysis (Generalized Least Squares extraction method) points out a one-factor solution also for this scale (Table 4).The factor shows high saturations corresponding to the 7 items of the scale.Explained variance is 73.09%.Cronbach's alpha, for this sample, is .95.

Gratitude Perceived by Operators Scale
The T-test for independent samples highlights that people who have families to care for experience significantly more Gratitude perceived by operators (M 15.36, SD 3.32), compared to those who do not (M 13.94, SD 3.43) [t(266) = 2.55, p < .05]; the effect size measured with Cohen's d was .42.No differences emerged, on the contrary, between women and men, or between people who have children and people who have not.

Perception of Support Offered by Patients/Customers' Gratitude Scale
The T-test for independent samples highlights that women have significantly higher Perception of support offered by patients/customers' gratitude (M 23.

Discussion and Conclusion
Aim of this work was the development and the analysis of two scales devoted to measure the gratitude expressed by patients and the perception of support that gratitude by patients/customers offers, in order to fill the gap of the absence, to our knowledge, of an instrument to measure the positive side of the relationship between patients and health operators: gratitude for received service and caring is an aspect that seems central in these workers' experience.
In line with expectations, the first analysis to validate the instruments suggested a one-factor solution for both scales, with good internal reliability.Moreover, as the analysis of variance points out, both of them can discriminate between groups of subjects.Proceeding to verify the validity of the instruments, bivariate correlations have given away a low and non-significant relationship with the construct of customer-related social stressors and a high significant correlation with the analogous construct of customers-initiated support (even if this scale has no Italian validated version yet): in this way, we obtained a first confirmation of both divergent and convergent validity.The two scales show good psychometric properties and a new administration of the instruments to a larger sample of helping professionals in several sectors (nurses, front-office workers and educators) is now being performed to overcome the limitations of the present work (the sample was composed of a small number of subjects from the same geographical area; we only conducted initial analyses of the validation process).The ongoing research work will consent to confirm the internal structure of the Gratitude perceived by operators and of Perception of support offered by patients' gratitude scales and to investigate more thoroughly their psycho- metric characteristics.Therefore, it will allow us to examine the relationships of the scales with several other constructs (as e.g.burnout) relative to care and service professions.Using the data collected, we will moreover suggest an Italian adaptation of the Zimmerman and colleagues' (2011) Customer-initiated support scale.
Helping professions are specifically characterized by the relationship with "others", a relationship that can be perceived either positively or negatively.It would, therefore, be interesting-for both theoretical research and research applied in organizations-to offer the possibility of measuring both these aspects by ad hoc instrumentsbrief and one-factor structured-able to be easily inserted in large instruments that analyze several dimensions, such as the Gratitude perceived by operators scale and the Perception of support offered by patients' gratitude scale, joint to the Social support scale (Converso et al., 2009), on one side, and of the Customer-related social stressors (Dormann & Zapf, 2008) on the other.Assessing the absence or presence of the positive side of the relationship with patients, especially in contexts where such relationship lasts for a long time and is central to the process of caring, may have the significance-in organizational practice-of guiding managers to ponder resources and to develop organizational interventions and/or actions.In this way, workers' motivational processes and job satisfaction may be supported.Furthermore, devices-functional to face efforts and requests of a daily physically and emotionally demanding and involving job-may become available for helping professionals, such as nurses and health operators.

Table 1 .
Descriptive analysis of single items of the gratitude perceived by operators scale.

Table 2 .
Descriptive analysis of single items of the perception of support offered by patients/customers' gratitude scale.Smirnov and of Shapiro-Wilk suggest rejecting the hypothesis of normality of the distributions.

Table 3 .
Factorial solution for the gratitude perceived by operators scale.

Table 4 .
Factorial solution for the perception of support offered by patients/customers' gratitude scale.'s d was .76.Also for this scale, no differences emerged, on the contrary, between people who have children and people who do not.shows, Gratitude perceived by operators highly correlates with both Perception of support offered by patients' gratitude and with Customers-initiated support.On the contrary, no correlations with Customer-related social stressors are shown, nor with age, job tenure in the sector or worked hours per week.

Perception of Support Offered by Patients/Customers' Gratitude Scale
Similarly, the scale of Perception of support offered by patients/customers' gratitude shows a high correlation both with Gratitude perceived by operators and with Customers-initiated support.Also in this case, no correlations are shown with Customer-related social stressors, nor with age, job tenure in the sector or worked hours per week.