The Relationship of Public Service Motivation to Job Satisfaction and Job Performance of Emergency Medical Services Professionals

Abstract

Public service motivation (PSM) are motivational factors that are unique in professions that serve the public. This study examined PSM’s relationship to self-reported job satisfaction and job performance in a unique sample of emergency medical services professionals, in which little research on the PSM construct has been undertaken. The PSM factors that emerged in this study did not mirror the traditional four-factor structure. The public interest and self-sacrifice factors formed a single public service factor, and a small number of compassion factors loaded on a second factor, with the policy-making factor being fully replicated. All three factors were significantly related to job satisfaction, and none were related to job performance, while controlling for the influence of demographic and contextual factors. All of these contextual factors were significantly related to job performance, except for the length of time in the emergency medical services (EMS) field, but not job satisfaction.

Keywords

motivation theory job satisfaction workplace attitudes and behaviors

Introduction

There is probably no greater responsibility than to have another person’s life in your hands. A variety of health care professionals experience this frequently, but perhaps none more so than emergency medical services professionals (EMSPs) (Elmqvist et al., 2009), specifically those certified as paramedics and emergency medical technicians (EMTs). The pressure inherent in this lifesaving role can lead to difficulties for EMSPs in their job performance, job satisfaction, and levels of motivation (Donnelly & Siebert, 2009). EMSPs face high-stress work environments, are regularly asked to put aside their own welfare to serve others, and within this work may be primarily motivated by intrinsic concerns such as altruism and public welfare given the relatively scant availability of significant material incentives.

Currently, inadequate motivation research has been done in the field of public safety workers, specifically with EMSPs such as paramedics and EMTs. In contrast, a robust literature focusing on the concept of public service motivation (PSM) (Perry & Wise, 1990) among public sector employees exists. Most studies on PSM have dealt with employees from the federal sector (Alonso & Lewis, 2001; Brewer & Maranto, 2000; Frank & Lewis, 2004; Moynihan & Pandey, 2007; Naff & Crum, 1999; Steijn, 2008; Vandenabeele, 2009). No PSM research has examined EMSPs as research participants. Through this study, knowledge of the ways in which PSM is related to job satisfaction and perceived job performance among EMSPs is advanced. This study fills the gap in the literature that exists when considering both the construct validity of PSM among EMSPs and the relationship between PSM and job satisfaction, and job performance of EMSPs.

Literature Review

PSM Theory

The concept of PSM arises from the work of Perry and Wise (1990). Their original formulation provided a typology of motivations associated with public service that includes rational (Policy-Making), norm-based (Public Interest), and affective motives (Compassion and Self-Sacrifice). The Public Interest factor is centered on the desire to promote the common good and is fueled by loyalty and a sense of duty to public. The two affective dimensions of Compassion and Self-Sacrifice reflect a sincere belief in the importance public service has on the lives of others and that serving others is a high form of loyalty and commitment to one’s country and community (Kim, 2009). The Compassion and Self-Sacrifice factors, often viewed as the central feature of PSM, reflect an emphasis on altruism and prosocial values. The Public Interest and Policy-Making factors reflect an emphasis on public service and public institutions. Rainey and Steinbauer (1999) later described PSM as a general altruistic motivation to serve some larger group of people, whether that be an organization, community, or nation. A more recent take on the construct by Vandenabeele (2007) describes PSM as motives that transcend self-interest and involves investment and concern for larger social or political interests. Thus, PSM, regardless of whose conceptualizing it, involves concern beyond the self, a motivation to serve, and a belief that such service impacts a larger good to which the individual is attached and committed.

Perry and Wise (1990) viewed the concept of the “patriotism of benevolence” as an integral component of the Compassion dimension of PSM. At the heart of PSM, they argued, is the belief that altruism and selflessness are motives that serve both a given recipient and some greater good. These affective dimensions of PSM captures the way in which public service workers come to see their capacity for empathy, altruism, and selflessness as the emotional fuel for their pursuit of rational (Policy-Making) and normative (Public Interest) pursuits.

Measuring PSM

Perry’s subsequent research (1996) found empirical evidence for the aforementioned four-factor model of the PSM construct. Since that time several factor analytic studies have been conducted to better understand the configural validity of Perry’s four-factor model of PSM. Some studies have found support for a three-factor model (Coursey & Pandey, 2007) that includes Public Interest and Compassion factors, and the policy-making dimension. PSM is usually considered to operate both as a three- or four-dimension construct and a general PSM latent factor. For example, Kim (2009) found good evidence for a second-order PSM general factor, though policy-making was only weakly linked to it in contrast to the other three dimensions. This was also seen in a study by Camilleri (2006).

Problems have emerged with some of the proposed factors. First, attraction to Public Policy-Making has been noted to have poor conceptual fit with the other three PSM dimensions, and researchers have argued that it needs to go through additional item development (Kim, 2009). Second, the Compassion dimension of PSM has received some criticism for lacking observed reliability. Compassion had problematically low reliability coefficients (Cronbach’s α = .603 in this study), and this may be due to its use among populations of public service workers in which compassion may not be a key component of their professional worldview (Coursey et al., 2008), or in populations where beliefs about the nature and importance of compassion diverge significantly across individuals within the profession.

PSM and EMSPs

Perry and Wise (1990) argue that PSM should impact worker behaviors in three important ways: (a) As PSM rises, individuals are more compelled to seek membership in public organizations; (b) within public organizations, PSM is positively correlated with job performance; and (c) public organizations comprised of high proportions of workers who are high in PSM do not have to rely on extrinsic, utilitarian incentives to manage workers.

Strong evidence has accrued demonstrating the PSM is positively associated with job satisfaction and job performance (Naff & Crum, 1999; Vandenabeele, 2009). The effect is strongest for the Public Interest and Self-Sacrifice dimensions and in contexts where public servants are given concrete opportunities to serve the public (Homberg et al., 2015). Very little research has explored PSM’s nature and correlates such as job satisfaction and performance among EMSP or similar professionals. No studies specifically focusing on EMSP and PSM exist, but some research with arguably similar professions such as nurses, police, and soldiers have been conducted.

Research with nurses in Italy has demonstrated that PSM has a positive relationship with job satisfaction and performance and that this relationship is amplified in the presence of transformative leadership strategies that highlight for the nurse the prosocial impact they have had on those they serve (Belle, 2013). Research with police has demonstrated that PSM is positively correlated with job satisfaction, particularly when there is a good fit between the person with the job tasks and with organizational culture (Prysmakova & Vandenabeele, 2019). Qualitative research with police officers in Switzerland (Neumann et al., 2019) has found that PSM and prosocial motivation are distinct among police officers and can vary independently. Some officers value the abstract idea of dutifully serving the public, and others valued more the concrete opportunities to engage in prosocial service contact with citizens. The latter was cited by officers as a means to cope with the fact that they seldom experienced evidence of the more abstract contributions their work made to the community as an abstract whole. In terms of research with soldiers, Brænder and Andersen (2013) found that soldiers’ scores on Compassion decreased after deployment and their scores on Public Interest increased. The authors theorized that, upon exposure to the stress of deployment, soldiers dehumanize both the enemy and those they are purportedly serving. Their concrete prosocial reasons to serve specific persons decreases, but, as the authors point out, they must maintain some well of motivation. Consequently, their abstract prosocial motivation must increase, for example, their sense of serving the great public good. Finally, research with firefighters (Kim, 2011) has found PSM was linked to job satisfaction, but that the Compassion factor did not significantly correlate with the broader PSM factor. The authors speculated that firefighters in their sample did not distinguish between Compassion and Public Interest given the high bivariate correlation between the two factors.

Job Satisfaction and Job Performance in EMSPs

Many scholars and public administration experts are interested in how PSM can enhance performance by public servants in public organizations (Brewer, 2008). The effect of PSM on individual worker performance is mixed in empirical studies. Although many studies have shown PSM predicts individual performance (Bright, 2007; Lewis & Frank, 2002; Naff & Crum, 1999), some question still exist as to whether the relationship is causal.

Emergency medical services (EMS) delivered outside the hospital setting are often provided in an intimidating world where darkness, hostile weather conditions, problematic terrain, and erratic dangers amplify the burden to perform efficiently (Donnelly & Siebert, 2009; Elmqvist et al., 2009; MacRitchie & Leobowitz, 2010; Vettor & Kosinski, 2000). In the face of this adversity, EMSPs are expected to be thorough, professional, maintain a strong work ethic, and continually seek new academic and technical knowledge for professional improvement (Alexander, 2006; Alexander et al., 2009). Given these high demands on job performance, an equally strong set of motivations must undergird EMSPs’ effort. This study examined whether PSM is a motivation set that is significantly related to self-reported job satisfaction and performance.

Statement of the Problem

Although significant PSM research has taken place in the preceding decades, no research has occurred with EMSPs. Moreover, the PSM construct appears rarely discussed among the leading trainers in the EMSP field, and this reflects an unmet potential for application. An additional reason why this study is needed can be seen in the high prevalence of occupational burnout and turnover among EMSPs (Boland et al., 2018). PSM holds some potential to be identified in later research as a significant factor in the development of burnout, particularly for EMSPs who may experience little opportunity to act out their affective public service values in their work. Accomplishing that task begins with assessing the PSM construct in the EMS population, evaluating its relationship with the satisfaction and performance of these workers, and the ways in which contextual factors moderate that relationship.

Research Questions

This study sought to address the following research questions utilizing the stated statistical analyses:

Research Question 1: Does the four-scale factor structure found in prior research with the PSM scale appear in this sample of EMSPs?

Given the extreme lack of PSM research with EMSPs, no prior empirical precedent exists to guide hypotheses as to what factor structure will be found. The best available evidence suggests a four-factor solution should be found. Some theoretical basis exists to assume that the Compassion dimension may display poor reliability and may not emerge as a distinct factor. Specifically, research with similar professions such as firefighters has found that such professionals may not distinguish between the normative (Public Interest) and affective (Self-Sacrifice and Compassion) dimensions of PSM (Kim, 2011). In addition, Policy-Making also may be likely to fail to emerge as a factor of PSM that is strongly related to the overall construct.

Research Question 2: Do PSM facets significantly relate to both self-reported job satisfaction and self-reported job performance among EMSPs, particularly when controlling for several relevant contextual and socialization variables pertinent to EMSPs such as gender, number of years in the EMS field, whether the worker serves a rural versus urban population, and the EMSPs’ level of education and level of EMS licensure?

Based on prior literature, a positive relationship between PSM and both job satisfaction and job performance was anticipated. In particular, based on prior research (Homberg et al., 2015; Perry, 1996) a strong positive relationship is likely between the Public Interest and Self-Sacrifice dimensions of PSM and job satisfaction and performance.

Research Question 3: What relationships exist between demographic and contextual variables and job satisfaction and performance among EMSPs?

The primary purpose of including the demographic and contextual variables of gender, tenure in the EMS profession, level of education, and level of licensure is to control for these variables so that any findings related to PSM and job performance and satisfaction can be isolated. Secondarily, the study seeks to explore the relationship between these variables in a sample of EMSPs. There is little research on this population in general, and even less on the impact of demographic and contextual variables on their working lives.

Method

Participants

Data were collected, using the web-based survey site, Survey Monkey™, over a 2-week period during the month of September 2015. The survey battery included an initial demographics questions page, and then PSM scale, job satisfaction scale, and job performance scales were set to be randomly ordered for each participant after that point. The Office of EMS and Trauma in a southeastern state provided the contact information for 12,143 EMSPs licensed in that state. Of these, 10,675 had usable email addresses. The survey instrument was sent to 10,675 EMS responders including 5,308 EMTs, 4,164 paramedics, 822 advanced EMT, 321 intermediate EMTs, and 60 emergency medical responders (EMRs) via email. Reminders were sent approximately every 3 days to the EMS responders that had not participated in the survey. A total of 1,987 responses were collected from the 10,675 for a response rate of 18.61%. No data were available for the overall demographics of the entire potential subject pool; therefore, no inferences can be made about the ultimate pool of respondents and how they compare to nonrespondents. For the purposes of the current analyses, 1,403 participants had completed response sets for all the study variables. Table 7 summarizes the number of participants in each demographic and contextual variable level. The vast majority of the sample were men (81%). Participants were split relatively evenly across rural (45%) versus urban (31%) work environments. The large majority were licensed at the paramedic level (57%) and did not hold an associate’s or bachelor’s degree (47%). Finally, 90% of the sample identified as White, with the next highest racial group being Black (3.7%).

Measures

Contextual and socialization variables

The following variables were included as control variables as they are known to be related to job performance and satisfaction. These variables include self-reported gender, level of education, the population density of the area the EMSP served (Population Served), the respondent’s level of EMS licensure (EMS Licensure), and the respondent length of time spent in the EMS profession (EMS Tenure).

PSM scale

Perry (1996) constructed the first PSM scale, and it included 44 items. He later reduced it to 24 items for efficiency. This study used the revised 24-item version of the scale. This version of the scale is subdivided into four subscales as follows: eight Self-Sacrifice subscale items, eight Compassion subscale items, five Public Interest subscale items, and three Public Policy-Making subscale items. The Perry (1996) instrument has been shown to be valid and reliable across multiple studies and authors (e.g., Frank & Lewis, 2004; Moynihan & Pandey, 2007; Naff & Crum, 1999). The four-factor structure of the PSM measure has been replicated, including in samples outside the United States and Western countries (Kim, 2009; Kim et al., 2012). However, Perry’s (1996) seminal research found potential evidence for a three-factor solution. Specifically, Perry found that the intercorrelation between the Public Interest and Self-Sacrifice was .89. When he tested a three-factor solution that combined the Public Interest and Self-Sacrifice dimensions against the four-factor solution, he found the latter to have better statistical fit to the data.

Job satisfaction

The satisfaction scale was based on four items from Naff and Crum (1999) and five items from Vandenabeele (2009). The satisfaction scale was found to be highly reliable (α = .84).

Job performance

The job performance scale for our study consisted of five items. This scale was based on four job performance items from Vandenabeele (2009) and a performance appraisal item from Naff and Crum (1999). A 5-point agreement scale was used for the items of PSM, job satisfaction, and job performance. These items ask the respondent to report the results of their most recent performance evaluation; thus, this is a measure of self-reported, rather than objective, assessment of job performance. The job performance scale for our study consisted of five items and was found to be reliable (α = .76).

Results

Exploratory Factor Analysis (EFA) of the PSM Scale

To assess the construct validity and suitability of the PSM scale with EMSPs in this study, an EFA was conducted using a principal components analysis yielding a single factor and using a promax rotation. The scree plot was inspected and suggested a four-factor solution as optimal explaining 52.59% of the variance. This was followed by an EFA using principal axis factoring extraction method and a promax rotation and set to a four-factor solution. The resulting four factors explained 43.14% of the variance. The pattern matrix was consulted to determine item loadings on each factor. These loadings are summarized in Table 1. Factor loadings of .5 or above were retained.

Table 1.

Summary of Pattern Matrix From Exploratory Factor Analysis for the Public Service Motivation Scale Using Maximum Likelihood Estimation, Principal Axis Factoring Extraction Methods, and a Promax Rotation (N = 1,403).

Item	Factor loadings
Item	1	2	3	4
PI3 Meaningful public service is very important to me.	.777	.007	−.008	.025
SS7 I am one of those rare people who would risk personal loss to help someone else.	.775	−.019	.104	−.044
SS8 I am prepared to make enormous sacrifices for the good of society.	.757	−.085	.067	.020
SS4 Much of what I do is for a cause bigger than myself.	.720	.037	.004	.021
SS6 I feel people should give back to society more than they get from it.	.717	.084	.036	−.031
PI4 I consider public service my civic duty.	.683	−.027	−.045	.117
SS2 I believe in putting duty before self.	.677	−.042	.119	−.035
SS1 Making a difference in society means more to me than personal achievements.	.662	.010	.047	.089
PI2 I unselfishly contribute to my community.	.652	−.025	−.127	−.083
SS5 Serving citizens would give me a good feeling even if no one paid me for it.	.622	−.027	−.192	−.087
PP3 I don’t care much for politicians.	−.042	.853	−.077	−.043
PP1 Politics is a dirty word.	.036	.742	−.059	−.004
PP2 The compromises that are involved in public policy-making don’t appeal to me.	−.051	.534	.057	.046
C5 I seldom think about the welfare of people whom I don’t know personally.	.016	−.112	.654	.078
C1 I am rarely moved by the plight of the underprivileged.	.089	−.092	.643	−.120
SS3 Doing well financially is definitely more important to me than doing good deeds.	−.117	.015	.567	.211
PI1 It is hard for me to get intensely interested in what is going on in my community.	−.345	.158	.464	.182
C2 Most social programs are too vital to do without.	.044	−.102	.060	.626
C4 To me, patriotism includes seeing to the welfare of others.	.311	.098	−.121	.457
C6 I am often reminded by daily events about how dependent we are on one another.	.360	.041	−.010	.382
C3 It is difficult for me to contain my feelings when I see people in distress.	−.004	.029	.081	.370
PI5 I would prefer seeing public officials do what is best for the whole community even if it harmed my interest.	.393	.079	.000	.185
C7 I have little compassion for people in need who are unwilling to take the first steps to help themselves.	.112	.206	.370	−.187
C8 There are few public programs that I wholeheartedly support.	.285	.161	.362	−.210
% of variance explained	26.81	2.13	3.12	2.54

Note. Factor loadings above .5 appear in bold. PI = Public Interest subscale; SS = Self-Sacrifice subscale; PP = Public Policy-Making subscale; C = Compassion subscale.

Analysis of the factor loadings revealed that the original four-factor structure found by Perry and other authors did not fully emerge in the current sample. In line with the existing four-factor structure, the Self-Sacrifice items all loaded on the first factor, but so did three of the five Public Interest items. The second factor was comprised of all reverse-scored items from the Compassion subscale, Self-Sacrifice subscale, and Public Interest scale. This poses the possibility that this factor was a method factor due to the reverse-worded nature of the items, and this will be discussed in more detail later. The third factor replicated the Public Policy subscale from the existing Perry four-factor structure. The final fourth factor only consisted of one item from the Compassion scale and a second item that did not exceed the factor loading threshold, also from the Compassion scale.

CFA of the PSM Scale

To assess the factor structure for the PSM scale that best fits the data with the current sample, a confirmatory factor analysis (CFA) using maximum likelihood estimation was performed using the AMOS version 22 software. The original PSM scales conforming to the existing four-factor structure were used in the first CFA model. Here the chi-square statistic was statistically significant χ²(246, N = 1,403) = 2,174.95, p < .001. While this result suggests we can reject the null hypothesis that the model fits the data perfectly, the chi-square statistic is notably overly sensitive to large sample sizes, such as is the case in the present sample. Better fit indicators for these CFA analyses in this sample include the root mean square error of approximation (RMSEA), root mean square residual (RMR), and the confirmatory fit index (CFI). Along these indicators the Perry four-factor CFA exhibited a reasonably good fit on RMSEA with a statistic less than .08. However, the RMR (.084, below .08 indicating good fit) and CFI (.839, above .90 indicating good fit) statistics were in a range indicating less than acceptable fit.

Overall, the CFA for the original PSM factor structure was not a sufficient fit with the data from the current sample, and additional item loadings were considered in the following CFAs.

Given the poor fit of the existing PSM factor structure, the factor structure found in the aforementioned EFA item loadings were used to constitute the second CFA model. The EFA had yielded a factor comprised of only one item from the Compassion scale, and this item was dropped from the present CFA. This left a three-factor solution based on the EFA results to be tested in the second CFA model. The chi-square test for the three-factor model remained statistically significant, χ²(101, N = 1,403) = 862.22, p < .001. Again, while this suggests poor fit, the chi-square test is sensitive to sample size and other fit indicators should be consulted. All three fit indices suggested acceptable levels of fit to the data (RMSEA = .073, less than .08; RMR = .043, less than .08; CFI = .913, above .9). Overall, the factor structure of the three-factor CFA model derived from the EFA achieved an adequate fit to the data and was superior to the four-factor model’s fit to the data. Table 2 presents fit indices for both CFA models, and Table 3 presents factor loading regression coefficients for each item in the final three-factor model.

Table 2.

Goodness-of-Fit Indicators of CFA Models of PSM Factor Structures (N = 1,403).

Model	χ²	df	χ²/df	RMSEA	RMR	CFI	AIC	BIC
For-factor model using Perry’s item loadings	2,176***	246	8.85	.075^a	.084	.839	2,284	2,567
EFA-derived three-factor model	862***	101	8.54	.073^a	.043^a	.913^a	932	1,116

Note. CFA = confirmatory factor analysis; PSM = public service motivation; RMSEA = root mean square error of approximation; RMR = root mean square residual; CFI = comparative fit index; AIC = Akaike information criteria; BIC = Bayesian information criteria; EFA = exploratory factor analysis.

Meets established goodness-of-fit criteria.

***

p < .001.

Table 3.

Standardized and Unstandardized Coefficients for CFA of PSM Items (N = 1,403).

Observed variable	Latent construct label	β	B	SE
PI3	Public Service	.798	.657	.019
SS7	Public Service	.705	.678	.023
SS8	Public Service	.716	.701	.023
SS4	Public Service	.736	.655	.021
SS6	Public Service	.699	.605	.021
PI4	Public Service	.749	.691	.022
SS2	Public Service	.608	.589	.024
SS1	Public Service	.676	.598	.022
PI2	Public Service	.670	.598	.022
SS5	Public Service	.656	.706	.026
PP3	Public Policy	.839	.889	.030
PP1	Public Policy	.733	.848	.033
PP2	Public Policy	.541	.574	.030
C5	Compassion	.595	.569	.031
C1	Compassion	.640	.637	.033
SS3	Compassion	.513	.467	.029

Note. CFA = confirmatory factor analysis; PSM = public service motivation; PI = Public Interest subscale; SS = Self-Sacrifice subscale; PP = Public Policy subscale; C = Compassion subscale.

Finally, these are non-nested competing models and to properly compare them the Akaike information criteria (AIC) and Bayesian information criteria (BIC) should be used for model comparisons (Bowen & Guo, 2012). Table 2 includes all fit indices statistics including the AIC and BIC statistics for both models. Smaller values for both statistics indicate superior fit. On both the AIC and BIC statistics, the three-factor solution exhibited superior fit.

The first factor is comprised of positively worded items from the original Self-Sacrifice and the Public Interest scales. This factor accounted for the vast majority of variance in the initial EFA and has been labeled Public Service in this study. The second factor consists of reverse-worded items from the original Compassion scale and one item from the Self-Sacrifice scale. As will later be discussed, this factor may be a method factor due to reverse worded items. The factor has been tentatively labeled Compassion given the nature of the items and that two of the three items were from the original Compassion scale. The third factor is a replication of the Public Policy scale in the original PSM instrument and has retained this label here. Figure 1 graphically presents the results of the three-factor PSM CFA. Notably, the Policy-Making factor is inversely correlated with the Public Service factor. In contrast, the Compassion factor is positively correlated with both other factors.

Figure 1.

Final confirmatory factor analysis path diagram showing standardized regression weights for three-factor solution for the public service motivation scale (N = 1,403).

Testing for a General PSM Second-Order Factor

The final analysis to establish the best fit model of PSM items to the current data set is to test a second-order general PSM factor. To do this three CFAs were conducted, summarized in Table 4. The first CFA included a second-order PSM factor that is comprised of all 3 first-order factors—Public Service, Compassion, and Public Policy. The second CFA included a second-order PSM factor that was comprised of the Public Service factor and the Compassion factor, with the Public Policy factor left to covary with the second-order factor. The third model included the Public Service and Public Policy factors forming a second-order factor, with the Compassion factor left to covary with the second-order factor. The model with a second-order factor including all first-order factors was statistically equivalent to the model with a second-order factor comprised of Public Service and Public Policy. Both the three-factor model and the model including the Public Service and Compassion second-order factor achieved adequate fit indices along RMSEA, RMR, and CFI statistics, though the three-factor model has slightly better fit indices. Overall, the evidence for a general PSM second-order factor was insufficient to warrant collapsing any of the three factors into a single score.

Table 4.

Goodness-of-Fit Indicators of CFA Models for Second-Order Public Service Motivation Factor Structure (N = 1,403).

Model	χ²	df	χ²/df	RMSEA	RMR	CFI	AIC	BIC
Three-factor CFA	862_a***	101	8.54	.073^†	.043^†	.913^†	932	1,115
Three-factor CFA with second-order factor	1,022_b***	103	9.93	.08^†	.072^†	.895	1,088	1,261
Second-order factor (Public Service and Compassion)	962_c***	102	9.43	.078^†	.630^†	.902^†	1,029	1,208
Second-order factor (Public Service and Public Policy)	1,021_b***	102	1.01	.08^†	.073^†	.895	1,089	1,267

Note. Chi-square statistics sharing the same alphabetical subscript are not statistically significantly different from one another. CFA = confirmatory factor analysis; RMSEA = root mean square error of approximation; RMR = root mean square residual; CFI = comparative fit index; AIC = Akaike information criteria; BIC = Bayesian information criteria.

†

Denotes that the statistic meets established goodness-of-fit criteria.

***

p < .001.

The Public Service factor accounted for the vast majority of variance in the initial EFA and does not appear to combine meaningfully with either of the other two factors. Thus, each will be treated as separate scale scores in subsequent analyses. These CFA models examining second-order PSM factors are non-nested and, again, the AIC and BIC statistics are more appropriate for comparison between non-nested models than is the chi-square statistic (Bowen & Guo, 2012). In addition to other fit indices, Table 4 includes the AIC and BIC statistics for both models, and the best scores on these indicators were observed on the three-factor solution without a second-order factor. This, again, suggests that in the current sample a general PSM factor comprised of two or more of the three latent factors observed is not the best fit to the data.

Common Method Bias (CMB) Analysis

This study utilized a mono-method approach to gathering data from EMSPs—an anonymous online survey. These leave open the possibility that some of the resulting response patterns are influenced due to a single data collection method being utilized.

To test for CMB, the Harmann single-factor EFA approach and the common factor CFA tests were performed. Both procedures produced latent factor indicators of CMB that were below the thresholds considered to reflect this problem. Specifically, the Harmann single-factor CFA only explained 26.56% of variance in responding. This falls below the generally accepted 50% threshold used for this analysis and does not suggest the presence of CMB (Podsakoff et al., 2003). For the common factor CFA test, the percentage of variance attributable to common method variance was 16.81%. Again, this falls well below the 50% variance explained threshold that would suggest a strong likelihood of unacceptable levels of CMB (Podsakoff et al., 2003).

Possible Reverse-Scored Method Factor

The final CFA from the original analysis yielded three factors. Factor 2 was comprised of two items from the Compassion scale (C1 and C5) and one item from the Self-Sacrifice Scale (S3). All of these were reverse-scored items. This pattern suggests the possibility of a method factor induced by the presence of reverse-scored items. To assess the source of the reverse-scored method factor, a CFA analysis was conducted (Table 5 summarizes these results), with all of the PSM items were included, but all of the reverse-scored items were included in one latent factor. This included the reverse-scored items that loaded on to the method factors in the final three-factor CFA, items that did not load significantly on any factor in the initial EFA, and the three items of the Public Policy scale, which are all reverse-scored. This CFA model yielded superior goodness-of-fit indicators than the four-factor CFA model and the three-factor CFA model. Specifically, the AIC and BIC statistics for comparisons of fit between non-nested models such as these exhibited the most superior fit for the three-factor solution. This suggests that the scale’s reverse-scored items, when taken as a whole, did not form a compelling single latent factor that fit the data better than the final three-factor solution. This casts some doubt upon the possibility that the final three-item scale of reverse-scored items is due solely or primarily to careless responding to such items. If these scenarios were the case, such a response pattern would emerge across all reverse-scored items, and a clear reverse-scored method factor would emerge when examining all such items on the scale. Still, this does not definitely rule out the possibility that the three-item scale would be a response style artifact.

Table 5.

Summary of CFA Goodness-of-Fit Indicators for Factor Structures Assessing the Presence of a Reverse-Scored Method Factor.

Model	χ²	df	χ²/df	RMSEA	RMR	CFI	AIC	BIC
Four-factor model with all PSM items	2,176***	246	8.85	.075^a	.084	.839	2,284	2,567
EFA-derived three-factor model	862***	101	8.54	.073^a	.043^a	.913^a	932	1,116
Two-factor model with a reverse-scored method factor	2,214***	151	14.66	.099	.109	.793	2,292	2,497

Note. CFA = confirmatory factor analysis; RMSEA = root mean square error of approximation; RMR = root mean square residual; CFI = comparative fit index; AIC = Akaike information criteria; BIC = Bayesian information criteria; PSM = public service motivation; EFA = exploratory factor analysis.

Denotes that the statistic meets established goodness-of-fit criteria.

p < .05, **p < .01, ***p < .001.

Descriptive Statistics

The descriptive statistics for the revised three-factor scales of the PSM instrument, the nine-level EMS Tenure variable that is treated as a continuous variable, as well as the descriptive statistics for the job satisfaction and job performance dependent variables are illustrated in Table 6. All variables exhibit acceptable levels of skewness, with two variables, Public Service and Job Performance, exhibiting large kurtosis. Given the large sample size, this deviation from normality is not likely to adversely affect the interpretability of results. Table 7 summarizes descriptive statistics for the categorical variables.

Table 6.

Descriptive Statistics for Continuous Variables (N = 1,403).

Variable	Minimum	Maximum	M	SD	Skewness	SE	Kurtosis	SE
Public Service	12	60	46.08	8.21	.981	.065	2.21	.131
Compassion	3	15	1.76	2.14	.370	.065	0.445	.131
Public Policy	3	15	7.37	2.66	.370	.065	0.067	.131
EMS Tenure	1	9	3.63	2.01	.513	.065	0.505	.131
Job Satisfaction	9	45	32.79	6.30	.592	.065	0.382	.131
Job Performance	5	25	2.58	2.60	.588	.065	2.03	.131

Note. EMS = emergency medical services.

Table 7.

Frequency Counts for Categorical Variables (N = 1,403).

Variable label	Variable levels	Count
Population served	Rural area or small town	626
	Medium town	341
	Large town or mid-sized city	436
Licensure level	EMR or EMT	430
	I-EMT or A-EMT	177
	Paramedic	796
Gender	Female	266
Gender	Male	1,137
Level of education	Some college or less	664
	Associate’s degree	397
	Bachelor’s degree or higher	342

Note. EMR = emergency medical responder; EMT = emergency medical technician; I-EMT = intermediate EMT; A-EMT = advanced EMT.

Bivariate Correlations

Before conducting the primary MANCOVA (multivariate analysis of covariance) analysis, bivariate Pearson correlations between the continuous and dichotomous variables in the analysis were conducted and summarized in Table 8 along with Cronbach alpha reliability coefficients in parentheses for each scale. This was conducted to test the MANCOVA assumption that the dependent variables are, in fact, correlated with one another. Job Satisfaction and Job Performance were correlated at .227 (p < .01). Significant correlations were also observed between all of the PSM revised subscales and job satisfaction and performance, except between Compassion and job performance, nor between Public Policy and job performance. The PSM revised subscales exhibited a significant degree of intercorrelation, namely the Public Service and Compassion scale were mildly positively correlated (r = .297, p < .01), and the Compassion and Public Policy were positively mildl, positively correlated r = .187, p < .001. In contrast, Public Policy was inversely correlated with Public Service (r = −.123, p < .001). Note that the Public Policy factor utilizes reverse-scored items, with higher scores reflecting more interest in Policy-Making.

Table 8.

Bivariate Correlations Between Continuous and Dichotomous Variables (N = 1,403).

Variable	1	2	3	4	5	6	7
1. Gender (1 = female, 2 = male)	—
2. EMS Tenure	.141**	—
3. Public Service	.021	.042	(.905)
4. Compassion	−.183**	.045	.297**	(.602)
5. Public Policy	−.065*	.034	−.123**	.187**	(.739)
6. Job Satisfaction	.003	.014	.430**	.227**	.036	(.838)
7. Job Performance	.078**	.019	.382**	.048	.110**	.287**	(.759)

Note. Numbers in parentheses reflect Cronbach alpha reliability coefficients for the designate scale. EMS = emergency medical services.

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

Given the very low Cronbach’s alpha observed for the Compassion factor, bivariate intercorrelations between the items comprising this factor were reviewed to better understand the nature of this resulting factor. C1–I am rarely moved by the plight of the underprivileged was correlated with C5–I seldom think about the welfare of people whom I don’t know personally moderately (r = .409, p < .01). C5 was correlated with SS3–Doing well financially is definitely more important to me than doing good deeds moderately (r = .284, p < .01). Finally, C1 was correlated with SS3 moderately (r = .310, p < .01). Overall, the Self-Sacrifice item exhibited the weakest intercorrelations with other items in this factor. The Cronbach’s alpha analysis was run again, and the change in alpha if an item was deleted was consulted. The scale alpha did not rise with the deletion of any of the three items; for example, the alpha of .60 only reduces further if any of the three items are removed. To some degree, this low alpha is unsurprising as the statistic is known to vary directly with the number of items on the scale (Kottner & Streiner, 2010), and other research has had similar findings (Coursey et al., 2008).

MANCOVA Predicting Job Satisfaction and Job Performance

A MANCOVA was conducted to test for the relationship between several categorical independent variables: gender, level of education, licensure level, and type of population served (rural vs urban) and the job satisfaction and job performance dependent variables, as well as between four continuous covariates consisting of the identified latent factors among the PSM scale items, and a measure of time spent in the EMS profession with the two dependent variables. Box’s test of equality of covariance matrices had an observed p value of .001. To correct for this inequality of variances, the Pillai’s trace (V) multivariate statistic will be used instead of Wilks’s lambda. Also, given the large sample sizes in the present analysis, Multivariate analysis of variance (MANOVA) is robust to violations of homogeneity of variance. Levene’s test of equality of variances was nonsignificant for both dependent variables indicating equivalent levels of variance across both dependent variables.

Several statistically significant multivariate effects were found for the following independent variables/covariates: gender, population served (rural vs. urban), level of EMS license, Public Service, Compassion, and Public Policy scales of the PSM instrument. See Table 9 for a summary of these multivariate effects. The largest observed effect size for the multivariate tests was for the Public Service factor, which explained 23% of the total variance among the dependent variables ( $η_{p}^{2}$ = .234), which is considered a strong effect size. All other multivariate tests yielded a mild effect size.

Table 9.

MANCOVA Multivariate Effects.

Independent variable	Pillai’s trace	F	df	Error df	η22P
Gender	.005	3.675	2	1,390	.005
Education	.017	5.899	4	2,782	.008
Population serviced	.007	2.498	4	2,782	.004
EMS licensure level	.007	2.572	4	2,782	.004
EMS tenure	.002	1.488	2	1,390	.002
Factors 1—Public service	.231	209.146	2	1,390	.231
Factor 2—Compassion	.012	8.595	2	1,390	.012
Factor 3—Public policy	.009	6.425	2	1,390	.009

Note. MANCOVA = multivariate analysis of covariance; EMS = emergency medical services.

The between-subjects effects for each independent variable/covariate further clarified which dependent variable’s variance was significantly explained within each significant multivariate effect. Table 10 summarizes these findings. Scores on the Public Service factor were significantly related to both job satisfaction and job performance, and this was the only similarity between the two dependent variables. Gender, population served, and EMS licensure level were significantly related to job performance. Level of education, scores on the Compassion factor, and scores on the Public Policy factor were significantly related to job satisfaction. As was the case when examining the multivariate findings, the largest effect size was seen for the Public Service factor for both dependent variables. The total model for both dependent variables explained a mild degree of variance, 18% for job performance and 20% for job satisfaction.

Table 10.

MANCOVA Between-Subjects Effects.

Independent variable	Dependent variables
	Performance			Satisfaction
	F	df	η22P	F	df	η22P
Gender	7.35**	1	.005	0.138	1	.000
Level of education	9.87**	2	.014	0.885	2	.001
Population served	4.52**	2	.006	0.572	2	.001
EMS licensure level	4.94**	2	.007	0.144	2	.000
EMS tenure	2.25	1	.006	0.344	1	.002
Public service	233.36***	1	.144	255.49***	1	.155
Compassion	2.65	1	.002	12.16***	1	.009
Public policy	2.74	1	.002	8.16*	1	.006
	Adjusted R²		.176			.196

Note. MANCOVA = multivariate analysis of covariance; EMS = emergency medical services.

p < .05. **p < .01. ***p < .001.

Post Hoc Analyses

Comparisons between means of categorical variables with significant multivariate and between-subjects effects using a Bonferroni correction were conducted. These comparisons are summarized in Table 11. All of the findings for categorical independent variables are in reference to the job performance dependent variable. Here, men scored significantly higher on job performance than women. There was also a linear effect for level of education, with less educated respondents reporting lower job performance then those with higher levels of education. Respondents working in rural areas reported lower job performance than respondents working in urban areas. Finally, respondents with lower levels of EMS licensure reported lower job performance than those with higher levels of licensure.

Table 11.

Summary of Marginal Means and Simple Effects for Contextual and Socialization Categorical Variables and the Job Performance Dependent Variable (N = 1,403).

Categorical independent variable	M	SE	F statistic and p value
Gender			F = 7.07 (1, 1387), p < .01
Female	20.245_a	.154
Male	20.687_b	.089
Level of education			F = 9.77 (2, 1387), p < .05
Some college or less	20.100_a	.114
Associate’s degree	20.518_b	.138
Bachelor’s or graduate degree	20.780_b	.146
Population served			F = 4.223 (2, 1387), p < .01
Rural area or small town	20.234_b	.116
Medium town	20.515_abc	.146
Large town or mid-sized city	20.650_c	.132
Level of licensure			F = 4.88 (2, 1387), p < .01
EMR or EMT	20.297_b	.134
I-EMT or A-EMT	20.361_abc	.185
Paramedic	20.740_c	.101

Note. Mean scores sharing the same alphabetical subscript are not statistically significantly different from one another using Bonferroni post hoc comparisons. EMR = emergency medical responder; EMT = emergency medical technician; I-EMT = intermediate emergency medical technician; A-EMT = advanced emergency medical technician.

Discussion

PSM Factor Structure

This study found item loadings that diverged significantly from Perry’s four-dimension model. The Self-Sacrifice and Public Interest items fused into a single factor explaining the majority of the variance in items responses and the largest amount of variance in job satisfaction. The domain was also the only PSM dimension to significantly predict job performance and explained the largest share of the variance here as well. Compassion was positively and moderately related to the Public Service and Public Policy factors. In contrast, Public Policy was mildly and inversely related to Public Interest. This was not without precedent as Perry’s (1996) own research found Public Interest and Self-Sacrifice to be highly intercorrelated, and research with firefighters (Kim, 2011) found the Compassion and Public Interest dimensions to be highly intercorrelated.

These findings suggest that EMSPs have distinct beliefs about the general motivation to serve the greater good—the fusion of Self-Sacrifice and Public Interest items—versus the motivation to serve specific concrete human needs—compassion. This is similar to findings by Brænder and Andersen (2013) with soldiers and by Neumann et al. (2019) with police officers, in the sense that both the current sample and these authors’ samples display distinct views of abstract and rational versus concrete and affective motivations to serve others. The EMSPs in this study may have two distinct sets of motivational beliefs. One is focused on an abstract commitment to the greater good embodied in the Public Service factor that emerged. A distinct, but separate motivation is the desire, or lack thereof, to help specific others as reflected in the Compassion factor items. This interpretation must be offered very tentatively given the small number of items and the poor internal consistency of the Compassion factor found in this study. These professions—soldiers, police, and EMSPs—have in common the fact that they all deploy in a sense to a field setting, all involve serving and protecting in different ways, and all may face contact with members of the public that are stressful, not always safe, and often antagonistic. Given that the bifurcating of abstract public service versus concrete human service may be a means in all three populations to cope with the vagaries of their work. In such a work environment, a broad motivation to serve is sustainable, but a concrete motive with each given recipient may frequently wither in the face of situations that have no clear closure or success, for example, patients in an EMSP are, by definition, always headed somewhere else for someone else to complete the service task.

Another issue that begs explanation in regard to the Compassion scale is that it largely failed to emerge here, save two reverse-worded items. These two reverse-scored Compassion items sit alongside a reverse-scored Self-Sacrifice item, and, based on analyses, these items do not appear to merely reflect a method factor linked to reverse-worded items. The resulting factor exhibited poor internal consistency and did not combined with other factors to form a second-order general PSM. This all speaks to its potential reflecting, a unique artifact of the particular population being surveyed, rather than a domain underneath the larger construct of PSM among EMSPs. In other words, EMSPs may not view Compassion motives as part and parcel of their sense of duty and professionalism, and thus did not respond to them in ways that strongly covaried with their responses to Self-Sacrifice and Public Interest items. This failure of the responses of the EMSPs in this study to cohere around the Compassion items may reflect a tendency to cope with the stresses of the job through the use of gallows humor and/or cynicism, and as such, the way in which Compassion items were viewed here may be a product of burnout and a dearth of empathic reserves for some of these professionals. For example, research strongly suggests that empathy among EMSPs is inversely related to indicators of burnout, particularly emotional exhaustion and depersonalization (Williams et al., 2017). In addition, paramedic students have been shown to report less empathy than do members of related health care professions (Williams et al., 2016). For EMSPs, there is a cost to empathy and compassion, and they may more easily maintain their motivation through the more abstract commitments to sacrifice and meeting their normative public service duties. Thus, while EMSPs seemingly benefit from low empathy as an inoculative against burnout, they also may have less empathic reserve than other health care professionals, and factors such as empathy and abstract rational motives to serve the public good may not strongly covary with empathy and compassion in this population.

The final domain of Public Policy, as in previous research (Kim, 2009), did not appear to be a fundamental aspect of the PSM, and may be better viewed as a distinct, separate construct. This is further evidence that Public Policy-Making may not be a key component for PSM for many professions including EMSPs, but may still be an important occupational interest in its own right.

PSM, Job Satisfaction, and Job Performance

The Public Service factor was ultimately the most powerful predictor in the study of both job satisfaction and job performance, while controlling for other study variables. This suggests that the most powerful motivational source of job satisfaction is an EMSP’s belief in selfless service toward an abstract good greater than themselves. Much less salient in this study were the affective motivators (Compassion) nor the rational motivators (Public Policy). EMSPs appear to derive satisfaction and pleasure from fulfilling this sense of duty—normative motive. Similar to the case of soldiers and police, EMSP’s motivational options may be relegated to an abstract, normative motive, in that opportunities to derive motivation and satisfaction from concrete acts of compassion and service are either limited in scope or are even negative experiences. The combination of Public interest and Self-Sacrifice items predicting job satisfaction is precisely what was found in Homberg et al.’s (2015) meta-analysis. Thus, this specific relationship between job satisfaction and these domains of PSM is by no means unique to EMSPs and likely reflects the fact that EMSPs have a motive to serve the public good and their work provides a venue to act this normative motive out in a satisfying manner.

A second purpose that beliefs in the Self-Sacrifice and Public Interest items in particular on the Public Service factor may serve for EMSPs is that the work of EMSPs is often relatively low paying, stressful, and sometimes thankless. The disconnect between any material or fiscal motives among EMSPs and the realities of the job likely induces significant cognitive dissonance. Self-Sacrifice and Public Interest motives, in particular, may serve to rationalize working in a job that may limit their upward mobility and financial gain, thereby leading to a positive relationship between Self-Sacrifice motivation and higher levels of satisfaction.

Contextual and Socialization Variables

Based on the current model, the prototypical high-performing EMSP was male, with a bachelor’s degree, serving in urban area, and holding a paramedic license and who espouses a strong motivation to engage in selfless service to the public. This replicates findings by Alonso and Lewis (2001), who found that women reported lower job performance levels than comparably educated and qualified men. However, research by Perry (1997) and Vandenabeele (2009) did not find differences in job performance between men and women. Given that the performance variable here is self-reported, this finding may reflect actual, objective higher performance levels of men, but may also reflect only men’s differentially more positive appraisal of their performance when compared with women. Female gender roles often emphasize self-denial and self-negation or outright self-rebuke, and to the degree those internalized roles are at work in this finding, men’s apparent better job performance is possibly a function of women downgrading their performance as opposed to men either actually performing better or rating their performance better.

Second, a predicted positive relationship between level of education and job performance was observed. This replicates prior findings (Alonso & Lewis, 2001; Kim, 2005). EMSPs who have upgraded their skills quite logically likely perceive, if not actually are told by supervisors, that their performance is higher. EMS is a field in which high levels of education are not required for entry, and professionals who achieve college degrees are likely to bring a wider perspective and skill set to the job.

The population size of the area in which the EMSP works was significantly related to job performance as well. Previous research in Texas found that urban EMSPs are typically better paid, better educated, younger, male, and less burned-out, and likely to have less than a paramedic level of licensure (Chng et al., 2012). In this study, EMSPs in urban areas reported higher job performance than those in rural areas. Similarly, those with paramedic licensure reported higher job performance than those with EMR or EMT licenses. Work in urban areas may offer more opportunities to practice skills more frequently leading to performance improvements. Similar to education, those achieving the paramedic licensure level have received more in-depth training and higher performance is to be expected.

These findings as a whole must be viewed with caution given the self-report nature of the job performance variable. It is possible that the profile of high-performing EMSPs found here may just as easily reflect cognitive dissonance processes as much as it may reflect actual differences in performance. Variables such as education and licensure level reflect sustained effort on the part of the worker. Workers may rationalize their investment in these accomplishments by being biased toward viewing themselves as accomplished and performing well. The finding for rural versus urban may simply reflect the larger sample of patient contact that urban EMSPs accrue during their work which leads to increased opportunities for improvement.

Limitations

This is a cross-sectional study, and the temporal direction of causality between demographic/contextual variables, PSM, and job satisfaction and performance cannot be stipulated. It is possible that PSM develops from positive job experiences and positive job performance, rather than the reverse.

In addition, the study had a relatively low response rate of 18.61%, which leaves open the possibility of nonresponse bias. However, an increasing number of studies have collected survey data in successive waves and are able to reach population saturation in later waves. Such studies have observed that findings are similar regardless of the response rate observed (Meterko et al., 2015; Rindfuss et al., 2015; Smith, 2009). That said, the risk of nonresponse bias in this study cannot be completely ruled out and, ultimately, is unknown.

Recommendations for Future Research

This study examines PSM among a unique population of EMSPs, which has not been undertaken in prior research. Although the role of PSM, particularly the Self-Sacrifice and Public Interest items, with job satisfaction and job performance was clear in the results, these findings do not necessarily generalize to the broader range of professionals often referred to as “first responders.” For example, some of the participants in this study worked in fire departments, but not enough to generalize the findings to firefighters as a subspecialty. In addition, the present results cannot be assumed to generalize to police officers, including officers who are trained as EMT/paramedic, who are often on the scene and providing initial medical care similar to the work undertaken by EMT/paramedics. Further research with the full range of first responders is needed to fully understand the role of PSM in this population of workers, though the present findings make a strong initial case that the construct is salient and of practical utility among these particular public sector workers.

It would also be helpful if future research was longitudinal and utilizing more objective measures of professional competence. This would involve following PSM, job satisfaction, and job performance levels of EMS responders over time as changes are made in organizations. Job performance would be best operationalized as official evaluations by those in the workers’ line of supervision, or perhaps using feedback from patients/consumers of EMS services, as well as objective indicators of patient outcomes. Another useful follow-up study would move us past the prediction approach used in this study. We recommend that future research focus on developing theoretical causal models showing both precursors and effects of PSM. In such research, PSM would be a mediating endogenous variable that explains linkages between early professional socialization and preparation variables and later professional competence and functioning.

Finally, the question of mediating variables between PSM and job satisfaction remains a largely understudied domain. One of the best-known mediators in this relationship is the fit between the individual’s PSM-based desire to serve the public and a particular job providing such opportunities. EMSPs are likely to fit this description, and future research can specify that the nature of EMS work in terms of its public service opportunities mediates the relationship between PSM and job satisfaction and/or performance. Another potential mediating variable is alluded to the phenomenon of cynicism or gallows humor, which may mediate the relationship between the PSM dimension of Compassion and job satisfaction and/or performance. To the degree EMSPs motivation is seated in abstract emotional and normative motives to serve the public, this may be a means of staving off burnout and cynicism accrued through negative and disappointing experiences with those they serve. While a reliance on abstract PSM is at least some form of motivation, this may, ultimately, harm EMSPs as they do not cultivate a means of regularly experiencing satisfaction from more concrete acts of empathy and compassion with those they serve. In this study, high scorers on the Compassion factor reported more job satisfaction, and this may be due to a lack of the use of this purported coping mechanisms to deal with the challenges of this work. Moreover, EMSPs in this study did not seem to fully perceive and respond to the various Compassion items, and more research is needed to properly understand the content domains of empathy, compassion, and concrete human service motives among EMSPs. The current items may not effectively tap this domain. Related to this is the need to study issues such as compassion fatigue and vicarious traumatization as well as formal PTSD (post-traumatic stress disorder) in EMSPs and how PSM factors into this. A particularly interesting question is whether PSM thwarts the development of negative beliefs such as moral injury, which can further the progression of burnout and traumatic reactions.

Implications for Policy and Practice

Some practical implications for determining the levels of PSM and job satisfaction are to use the PSM framework to develop a more motivated and competent EMS. These improvements could be accomplished by using the PSM framework as a template for training, incentivizing, and sustaining of EMSPs.

In terms of training, the present results suggest that PSM is linked to EMSP’s satisfaction and performance. Training programs and continuing education in the EMS field may have the most impact when it explicitly instills, targets, and leverages PSM among learners. EMSP training that includes both technical skills learning but also a healthy dose of professional socialization and role modeling along the PSM dimensions may have more impact over a longer term by promoting retention and ongoing intrinsic investment in the work by EMS trainees. EMS scholars have argued that the role of these professionals has become increasingly professionalized over the recent decades, and this has increasingly driven the development of clear professional identities for EMSPs. This has occurred parallel to changes in their work profiles, moving from protocol-driven work to work that involves decision-making and professional collaboration (O’Meara, 2009).

The present results point toward the utility of responses by EMS leadership that highlight worker’s degree of abstract investment in the public or common good, their selflessness or even their interest in larger policy issues as ways to positively reinforce EMSPs’ behaviors. For EMSPs with motives in the policy domain, this may involve creating structures through which they can connect with and have an impact upon entities outside the EMS agency. This concept conforms closely to the concept of clinical governance put forth by Woollard (2009) as a vision for the professional development of the EMS professions. Clinical governance refers to, among other factors, systemic learning from critical incidents, continuous quality improvement, leadership, and creating organizational cultures that value clinical excellence. Such activities are seen by Wollard as fundamental to the growth and development of EMS professions.

Footnotes

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

ORCID iD

Tres Stefurak

Author Biographies

Tres Stefurak, PhD, is an associate professor of Counseling Psychology and the Department Chairperson of the Department of Counseling & Instructional Sciences at the University of South Alabama. He is a core faculty member of the Clinical & Counseling Psychology PhD program and his research is in the area of child welfare, juvenile offending, trauma psychology, police psychology and mixed methods research. His most recent work focuses on issues of burnout and vicarious trauma amon.

Ron Morgan, PhD, is the emergency medical sciences coordinator at Mississippi Gulf Coast Community College. He is formerly an assistant professor at the University of Western Illinois, former coordinator of the Emergency Medical Services training program at the University of south Alabama, and a former District Fire Chief for the City of Mobile, Alabama. He holds a doctorate in Instructional Design and Development and his interests are in the training and development of emergency medical services professionals.

R. Burke Johnson, PhD, is a professor of Educational Research Methods and Instructional Design & Development at the University of South Aalabama. He is an internationally recognized expert in the area of research methods in general and mixed methods research and program evaluation in particular. His most recent activity consists of outreach in recent years to international scholarly organizations to disseminate best practices in mixed methods research and recently serving as the executive director of the Mixed Methods International Research Association.

References

Alexander

M. A.

(2006). Foundations for the practice of EMS education. Pearson.

Alexander

M. A.

Weiss

Braude

Ernst

Fullerton-Gleason

(2009). The relationship between paramedics’ level of education and degree of commitment. The American Journal of Emergency Medicine, 27, 830–837. https://dx-doi-org.web.bisu.edu.cn/10.1016/j.ajem.2008.06.039

Alonso

Lewis

G. B.

(2001). Public service motivation and job performance evidence from the federal sector. The American Review of Public Administration, 31, 363–380. https://dx-doi-org.web.bisu.edu.cn/10.1177/02750740122064992

Belle

(2013). Leading to make a difference: A field experiment on the performance effects of transformational leadership, perceived social impact, and public service motivation. Journal of Public Administration Research and Theory, 24, 109–136.

Boland

L. L.

Kinzy

T. G.

Myers

R. N.

Fernstrom

K. M.

Kamrud

J. W.

Mink

Stevens

A. C.

(2018). Burnout and exposure to critical incidents in a cohort of emergency medical services workers from Minnesota. The Western Journal of Emergency Medicine, 19(6), 987–995.

Bowen

Guo

(2012). Structural equation modeling. Oxford University Press.

Brænder

Andersen

L. B.

(2013). Does deployment to war affect public service motivation? A panel study of soldiers before and after their service in Afghanistan. Public Administration Review, 73(3), 466–477. https://dx-doi-org.web.bisu.edu.cn/10.1111/puar.12046

Brewer

G. A.

(2008). Employee and organizational performance. In Perry

J. L.

Hondegham

Perry

J. L.

Hondegham

(Eds.), Motivation in public management (pp. 136–156). Oxford University Press.

Brewer

G. A.

Maranto

R. A.

(2000). Comparing the roles of political appointees and career executives in the U.S. federal executive branch. American Review of Public Administration, 30(1), 69–86.

10.

Bright

(2007). Does person-organization fit mediate the relationship between public service motivation and the job performance of public employees? Review of Public Personnel Administration, 27, 361–379. https://dx-doi-org.web.bisu.edu.cn/10.1177/0734371X07307149

11.

Camilleri

(2006). Towards developing an organizational commitment—Public service motivation model for Maltese public service employees. Public Policy and Administration, 21, 63–83.

12.

Chng

C. L.

Collins

Eaddy

(2012). A comparison of rural and urban emergency medical system (EMS) personnel: A Texas study. Prehospital and Disaster Medicine, 16(3), 159–165.

13.

Coursey

D. H.

Pandey

S. K.

(2007). Public service motivation measurement: Testing an abridged version of Perry’ s proposed scale. Administration & Society, 39(5), 547–568. https://dx-doi-org.web.bisu.edu.cn/10.1177/0095399707303634

14.

Coursey

D. H.

Perry

J. L.

Brudney

J. L.

Littlepage

(2008). Psychometric verification of Perry’s public service motivation instrument: Results for volunteer exemplars. Review of Public Personnel Administration, 28(1), 79–90.

15.

Donnelly

Siebert

(2009). Occupational risk factors in the emergency medical services. Prehospital and Disaster Medicine, 24, 422–429. https://dx-doi-org.web.bisu.edu.cn/10.1017/S1049023X00007251

16.

Elmqvist

Brunt

Frilund

(2009). Being first on the scene of an accident: Experiences of “doing” prehospital emergency care. Scandinavian Journal of Caring Sciences, 24, 266–273. https://dx-doi-org.web.bisu.edu.cn/10.1111/j.1471-6712.2009.00716.x

17.

Frank

Lewis

(2004). Government employees: Working hard or hardly working? The American Review of Public Administration, 34, 36–51. https://dx-doi-org.web.bisu.edu.cn/10.1177/0275074003258823

18.

Homberg

McCarthy

Tabvuma

(2015). A meta-analysis of the relationship between public service motivation and job satisfaction. Public Administration Review, 75(5), 711–722.

19.

Kim

(2005). Individual-level factors and organizational performance in government organizations. Journal of Public Administration Research and Theory, 15, 245–261. https://dx-doi-org.web.bisu.edu.cn/10.1093/jopart/mui013

20.

Kim

(2009). Revising Perry’s measurement scale of public service motivation. The American Review of Public Administration, 39(2), 149–163.

21.

Kim

(2011). Testing a revised measure of public service motivation: Reflective versus formative specification. Journal of Public Administration Research and Theory, 21, 521–546. https://dx-doi-org.web.bisu.edu.cn/10.1093/jopart/muq048

22.

Kim

Vandenabeele

Wright

B. E.

Andersen

L. B.

Cerase

F. P.

Christensen

R. K.

Desmaris

Koumenta

Leisink

Liu

Palidauskaite

Pedersen

L. H.

Perry

J. L.

Ritz

Taylor

De Vivo

(2012). Investigating the structure and meaning of public service motivation across populations: Developing an international instrument and addressing issues of measurement invariance. Journal of Public Administration Research and Theory, 23, 79–102. https://dx-doi-org.web.bisu.edu.cn/10.1093/jopart/mus027

23.

Kottner

Streiner

D. L.

(2010). Internal consistency and Cronbach’s α: A comment on Beeckman et al. (2010). International Journal of Nursing Studies, 47(7), 926–928.

24.

Lewis

Frank

(2002). Who wants to work for the government? Public Administration Review, 62, 395–404. https://dx-doi-org.web.bisu.edu.cn/10.1111/0033-3352.00193

25.

MacRitchie

Leobowitz

(2010). Secondary traumatic stress, level of exposure, empathy and social support in trauma workers. South African Journal of Psychology, 40, 149–158. https://dx-doi-org.web.bisu.edu.cn/10.1177/008124631004000204

26.

Meterko

J. D.

Restuccia

K. S.

Mohr

Brennan

Glasgow

P. K.

(2015). Response rates, nonresponse bias, and data quality: Results from a national survey of senior healthcare leaders. Public Opinion Quarterly, 79(1), 130–144.

27.

Moynihan

D. P.

Pandey

S. K.

(2007). The role of organizations in fostering public service motivation. Public Administration Review, 67, 40–53. https://dx-doi-org.web.bisu.edu.cn/10.1111/j.1540-6210.2006.00695.x

28.

Naff

K. C.

Crum

(1999). Working for America: Does public service motivation make a difference? Review of Public Personnel Administration, 19(4), 5–16. https://dx-doi-org.web.bisu.edu.cn/10.1177/0734371X9901900402

29.

Neumann

Baertschi

Ritz

(2019). Public service motivation, prosocial motivation and altruism: Towards disentanglement and conceptual clarity. International Journal of Public Administration, 42, 1200–1211. https://dx-doi-org.web.bisu.edu.cn/10.1080/01900692.2019.1588302

30.

O’Meara

(2009). Paramedics marching toward professionalism. Journal of Emergency Primary Health Care, 7(1), 990339.

31.

Perry

J. L.

(1996). Measuring public service motivation: An assessment of construct reliability and validity. Journal of Public Administration Research and Theory, 6, 5–22. https://dx-doi-org.web.bisu.edu.cn/10.1093/oxfordjournals.jpart.a024303

32.

Perry

J. L.

(1997). Antecedents of public service motivation. Journal of Public Administration Research and Theory, 7, 181–197. https://dx-doi-org.web.bisu.edu.cn/10.1093/oxfordjournals.jpart.a024345

33.

Perry

J. L.

Wise

(1990). The motivational bases of public service. Public Administration Review, 50, 367–373. https://dx-doi-org.web.bisu.edu.cn/10.2307/976618

34.

Podsakoff

P. M.

MacKenzie

S. B.

Lee

J.-Y.

Podsakoff

N. P.

(2003). Common method biases in behavioral research: A critical review of the literature and recommended remedies. Journal of Applied Psychology, 88(5), 879–903. https://https://dx-doi-org.web.bisu.edu.cn/10.1037/0021-9010.88.5.879

35.

Prysmakova

Vandenabeele

(2019). Enjoying police duties: Public service motivation and job satisfaction. Journal of Police and Criminal Psychology, 1–14. https://dx-doi-org.web.bisu.edu.cn/10.1007/s11896-019-09324-7

36.

Rainey

H. G.

Steinbauer

(1999). Galloping elephants: Developing elements of a theory of effective government organizations. Journal of Public Administration Research and Theory, 9(1), 1–32.

37.

Rindfuss

R. R.

Choe

M. K.

Tsuya

N. O.

Bumpass

L. L.

Tamaki

(2015). Do low survey response rates bias results? Evidence from Japan. Demographic Research, 32, 797–828.

38.

Smith

H. L.

(2009, December 1). Double sample to minimize bias due to non-response in a mail survey (PSC Working Paper Series, No. 09-05). Population Studies Center, University of Pennsylvania.

39.

Steijn

(2008). Person-environment fit and public service motivation. International Public Management Journal, 11, 13–27. https://dx-doi-org.web.bisu.edu.cn/10.1080/10967490801887863

40.

Vandenabeele

(2007). Toward a public administration theory of public service motivation. Public Management Review, 9(4), 545–556.

41.

Vandenabeele

(2009). The mediating effect of job satisfaction and organizational commitment on self-reported performance: More robust evidence of the PSM—Performance relationship. International Review of Administrative Sciences, 75, 11–34. https://dx-doi-org.web.bisu.edu.cn/10.1177/0020852308099504

42.

Vettor

Kosinski

(2000). Work-stress burnout in emergency medical technicians and the use of early recollections. Journal of Employment Counseling, 37, 216–228. https://dx-doi-org.web.bisu.edu.cn/10.1002/j.2161-1920.2000.tb01028.x

43.

Williams

Boyle

Howard

(2016). Empathy levels in undergraduate paramedic students: A three-year longitudinal study. Nurse Education in Practice, 16(1), 86–90.

44.

Williams

Lau

Thornton

Olney

L. S.

(2017). The relationship between empathy and burnout—Lessons for paramedics: A scoping review. Psychology Research and Behavior Management, 10(27), 329–337.

45.

Woollard

(2009). Professionalism in UK paramedic practice. Australasian Journal of Paramedicine, 7(4), 9. https://dx-doi-org.web.bisu.edu.cn/10.33151/ajp.7.4.181