Comparing Passenger Satisfaction, Employees’ Perspective and Performance on Quality and Safety Indicators: A Field Study

Abstract

This paper aims to analyze the impact that different attributes related to a Regional Airport service and the socio-economic factors of the passengers have on the passenger’s overall satisfaction. The study also compared passenger and employee satisfaction in relation to the service offered by the airport, to identify possible critical areas of improvement. An Ordinal Logistic Regression (OLR) approach was used to model how the attributes considered for qualifying airport services and the socio-economic variables impact the predicted variable (i.e., passenger satisfaction). Furthermore, the results were triangulated to include quality and safety performance indicators as an objective anchor point for the performance of the company. The findings indicate interesting areas of difference between the perceptions of the passengers and airport employees regarding a company’s services and its performance. The company managers in the key areas of operation were then asked to select the main areas of improvement among the ones highlighted by the survey’s results. Quality and safety indicators were also helpful in enriching the analysis and indicating good synergy with the suggestions collected from the passengers’ and the employees’ surveys, offering yet another complementary perspective.

1. Introduction

Service quality and traveler satisfaction are subjects of high interest within the airport industry [1]. Being the result of a cognitive process, perceived service quality is both subjective and context-dependent [2], and passengers’ expectations of the services supplied by the airport may also be influenced by socio-economic factors such as gender, age, purpose of travel and annual flight frequency [3]. Although the analysis of passenger satisfaction is crucial to improve the quality of airport services, it is also essential to consider the employees’ perspective. Indeed, the business motivation of the employee plays a very important role in meeting the passenger’s needs [4]. The measurement of the quality of the airport service represents a significant performance indicator for airport operations and management [5]. Indeed, without knowing the current performance on quality and safety indicators, it is difficult to identify which aspects could be improved [6]; the commercialization of the airport industry led airlines towards performance management systems that must take into account other important aspects of airport operation, such as service delays, safety and social responsibility [7].

Furthermore, another element that should not be underestimated is the likelihood of accidental events occurring, which is a required metric in any safety management system. Human error appears as a contributing factor in 70% to 80% of all aviation accidents, and, because errors can never be eliminated completely, a culture of open reporting can foster a better understanding of the nature of the possible errors, and the required improvement strategies [8].

Only a few authors have investigated the service quality of airports by jointly considering the points of view of travelers and airport employees [9,10,11], and fewer still have triangulated them with performance indicators [12]. Indeed, many researchers explore service quality and passenger satisfaction, but few studies identified middle management choices for improvement priorities in existing airports [13]. A similar study was conducted by Leva et al. [14], providing a systematic framework for performance management with an in-depth study on safety in day-to-day operations, analyzing the day-to-day performance of the key areas of an Italian regional airport having international reach. Other authors, such as Hong et al. [15], investigated the attributes that influence passenger and employee satisfaction, while Shahzad [4] studied the impact of employee motivation on passenger satisfaction in the airline industry of Pakistan. Bezerra and Gomes [1] analyzed airport service quality attributes and socio-economic factors, examining the effects of those considerations on overall passenger satisfaction using ordinal logistic regression. Logistic regression analysis and the ordinal logit model enable the study of how certain factors affect overall subject satisfaction, not only in the air transport sector but also in different contexts [16]. Indeed, Lu [17] estimated an ordinal logit model in his study on housing satisfaction, while Lawson and Montgomery [18] analyzed customer satisfaction using ordinal logistic regression models.

In this context, the aim of this paper is twofold. Firstly, to analyze the key drivers of overall passenger satisfaction in the airport, based on the 2020 Airport Service Charter data [19], and to compare these results with employee satisfaction in relation to the quality of the service offered. Secondly, to compare the service areas identified as having an impact on customers and employee satisfaction with the company performance indicators for the same level of service. This should allow the pinpointing of more comprehensively critical aspects to be improved, considering all perspectives. Additionally, the field study is accompanied by a simple survey, based on the literature [20], to elicit from the company middle managers’ perspective which areas of improvement are actually critical among the possible alternatives resulting from the employees’ survey. The paper is structured as follows. After the introduction, in Section 2 we give a review of the literature relating to the methodological approach and the experimental design considered for the field study. Section 3 presents the profiles of the respondents, the descriptive statistics on safety and quality indicators, and the results of the analysis of passenger satisfaction alongside the employees’ perspective. Lastly, we present the survey carried out with company managers. Section 4 concludes by providing some discussion on the results obtained and possible further work.

2. Materials and Methods

The models for examining traveler satisfaction and measuring airport service quality can be used in order to relate the service quality attribute to the overall satisfaction [21]; the various service aspects considered in this context can be expressed by qualitative variables on Likert ordinal scales, characterized by ordered categorical responses, like the judgments for the evaluation of a service: very bad, insufficient, discrete, good and excellent [21]. Logistic regression models are more commonly used in the literature to analyze passenger satisfaction and subjective qualitative measures, expressing the dependent variable as the passenger’s overall satisfaction, the predictors as the attributes related to the service quality and computing the weights of every attribute on the passenger’s overall satisfaction [21]. In this work, we used an extension of the technique referred to as ordinal logistic regression (OLR), modeled by the stepwise selection method, developed for ordinal response variables. A stepwise method is essential for this analysis because it finds the best combination of a set of attributes by automatically selecting the regressor to be added to, or removed from, the model and stopping when the variable has a significance level for entry (SLE) into the model >0.05, and the variable for removal has a significance level for staying (SLS) in the model <0.05 [22]. In this context, Eboli and Mazzulla [21] used an ordinal logistic regression (OLR) model to analyze the satisfaction of the passengers of the Lamezia Terme airport, and to identify the service aspects needing to be improved. This study was considered useful since the airport is comparable to the regional airport considered for the present study. In the OLR model, if the ordinal response variable assumes J as distinct values, the relationship with the X_k regressors can be expressed through the following formula:

log[ρ(Y ≤ j|X)/ρ(Y > j|X)] = α_j − ∑^K_k=1β_kX_k = α + Xβ,

(1)

for j varying between 1 and J-1, where α_j are the intercepts indicating the probability that the Y variable assumes low values rather than high values in case of the nullity of all the predictors, and β_k represents the log (ODDS) change corresponding to a unitary increase of the X_k variables; positive values of the β_k coefficients correspond to higher probabilities that the response variable assumes high values, and vice versa [21].

The field study was conducted from January to October 2020, and the experimental context relates to the services offered by the terminal of an Italian regional airport with international reach. Passengers were recruited by answering questions on socio-economic characteristics, after introducing them to the scope of the study and informing them that it would be conducted for research purposes and the collection of experimental data would remain completely confidential and anonymous. They were asked to express their perception of the quality of the services offered by the airport, based on the 2020 Airport Service Charter quality factors. The first experimental session was replicated by recruiting airport employees, and their views were further elicited to identify the critical aspects to be improved. The interviewed subjects expressed a judgment to each factor according to an ordinal verbal scale varying on five levels, which are “Very bad/Severely Insufficient”, “Insufficient”, “Discrete”, “Good”, and “Excellent” as required by the ENAC (Italian Civil Aviation Transport) methodology on the standard Airport Service Charter. In addition, a judgment on the overall perception of airport services was asked of the subjects, according to the same verbal scale. The second experimental session was designed as a simple survey, based on the literature [20], to elicit from the company managers in the key areas of operations their preference on the improvement areas, based on the critical aspects highlighted by the employees’ survey results. The interviewed company managers expressed their alternative of choice between the two most critical operational aspects (or other aspects) resulting from the analysis of the employees’ perspective. The areas considered during the field study are reported in Figure 1, and the various sections of the questionnaire are reported in Appendix A (Figure A1, Figure A2, Figure A3, Figure A4, Figure A5, Figure A6, Figure A7, Figure A8, Figure A9, Figure A10 and Figure A11). Based on the findings obtained from the study of Eboli and Mazzulla [21], we decided to test the hypothesis of what variables are to be considered significant or not, in relation to overall passenger satisfaction. We used a modified version of the customer satisfaction survey for the employee to enable a comparison of the two perspectives; we have also considered extra sections for the employees’ survey on operational areas such as the corporate mission and its communication, organizational structure, employee participation, and safety performance.

3. Results

3.1. The Passenger and Employee Profiles and Descriptive Statistics on Safety and Quality Indicators

Of the 378 passengers recruited, descriptive statistics revealed a relative balance between male and female travelers: 57% versus 43%, respectively. The age distribution of travelers appeared to be heterogeneous: 33% of them were concentrated in the third age class (35–44 years old), 22% of them belonged to the fourth age class (45–54 years old), 19% of them were 25–34 years old, 14% of them were concentrated in the fifth age class (>54 years old) and 12% of them belonged to the first age class (<25 years old). However, the distributions of the annual flight frequency and the purpose of the trip appeared to be homogeneous: 67% of passengers listed “leisure” as their trip purpose and 33% of them answered “business” as their trip purpose, while 56% of them answered “6–10 times a year” as their annual flight frequency, 18% of them were concentrated in the fourth annual flight frequency class (11–15 times a year), 14% of them belonged to the second annual flight frequency class (1–5 times a year) and 12% of them answered “less than once a year” as their annual flight frequency. Of the 26 employees recruited for the survey, 42% of them were male, 4% of them belonged to the second and fifth age class and 46% of them were concentrated in the third and fourth age class, whereas the employees’ average period of service was 16 years.

Table 1. Triangulation between passengers’ perception, employees’ perception and performance on quality indicators.

Quality Factor	Indicator	Passengers’ Perception	Employees’ Perception	Target of Metrics for 2020
Journey Security	Security screening	0.92	1.00	0.90
Personal safety and security	Personal safety and security	0.92	1.00	0.90
Reliability and punctuality	Punctuality	0.91	1.00	0.90
Airport cleanliness	Toilet tidiness	0.77	0.92	0.89
	Terminal tidiness	0.89	0.90	0.88
Overall airport comfort	Luggage trolley availability	0.92	0.81	0.90
	Air conditioning	0.91	0.92	0.90
	Overall comfort	0.91	0.91	0.90
Additional service	Wi-Fi	0.88	0.70	0.84
	Recharge points	0.84	0.77	0.80
	Quality and prices of shops	0.89	0.90	0.89
	Bar and restaurant	0.90	0.92	0.90
	Vending machines	0.73	0.88	0.84
Information to customers	Website	0.90	0.92	0.90
	Information points	0.91	0.92	0.90
	Internal sign-posting	0.87	0.89	0.90
	Staff skills	0.91	0.96	0.90
	Public information	0.86	0.88	0.89
Counter and gate services	Ticket counter	0.90	1.00	0.90
	Check-in	0.91	1.00	0.91
	Checkpoint	0.90	1.00	0.90
Modal integration	External sign-posting	0.91	0.92	0.90
	Airport links	0.89	0.92	0.89

Source: Authors’ elaboration.

shows the percentage frequency of satisfied passengers and employees, grouped by quality factors, compared to the target metrics set by the 2020 Airport Service Charter according to the ENAC (Italian Civil Aviation Transport) indications given by the airport operators. The relative frequency distributions of passengers’ vs. employees’ perceptions, related to the main indicators showing significant differences between the two groups of subjects, are reported in Figure 2, Figure 3, Figure 4, Figure 5, Figure 6, Figure 7, Figure 8, Figure 9, Figure 10, Figure 11, Figure 12, Figure 13, Figure 14, Figure 15, Figure 16 and Figure 17.

Satisfaction indicators were calculated by dividing the total number of satisfied respondents who answered “Discrete”, “Good”, or “Excellent” for the total number of respondents according to the ENAC (Italian Civil Aviation Transport) methodology on standard Airport Service Charter. Passengers’ perceptions elicited were in line with the 2020 target metrics, except for “Toilet tidiness and services” (as the number of available toilets seemed to be quite limited), “Vending machines” (the amount of vending machines also seems to be quite limited, and they need to be replenished more often), “Internal sign-posting” (few signs to direct the flow of people), and “Public information services” (limited availability of information points for passengers, with only the ticket service inside the terminal). On the other perspective, the employees’ perceptions elicited were in line with the 2020 target metrics, except for “Luggage trolley availability” (which is judged to be limited), “Wi-fi connectivity” (wi-fi connectivity inside the terminal for employees appears to be poor), “Recharge points” (recharging points for employees also appear to be scarce), “Internal sign-posting” and “Public information services”. The relative frequency distribution of passengers’ perceptions ranking is comparable with that expressed by the employees in relation to the same service attributes. Figure 2 reports the passengers’ and employees’ perception of the security of airport operations: 77% of employees perceived the level of security within the airport to be “Good” or “Excellent”, while only 47.7% of passengers judged this service to be “Good” or “Excellent”, 44.3% of them rated it as “Discrete” and 8% of them rated it below that level. Figure 3 shows the passengers’ and employees’ perception of the airport service punctuality: 85% of employees perceived the level of airport service punctuality to be “Good” or “Excellent”, while 46% of passengers judged this service to be “Good” or “Excellent”, 45.7% of them rated it as “Discrete” and 8.3% of them rated it below that level. The passengers’ and employees’ perception of the terminal tidiness is shown in Figure 4: 48% of employees perceived the level of terminal tidiness to be “Good” or “Excellent”, while only 37.3% of passengers judged this service to be “Good” or “Excellent”, 51.1% of them rated it as “Discrete” and 11.6% of them rated it below that level. Figure 5 shows the passengers’ and employees’ perception of the availability of luggage trolleys: 45% of employees perceived the availability of luggage trolleys to be “Good” or “Excellent”; while 42.3% of passengers considered this service to be “Good” or “Excellent”, 50% of them rated it as “Discrete” and 7.7% of them rated it below that level. Figure 6 shows the passengers’ and employees’ perception of the air conditioning: 50% of employees perceived the air conditioning available in the airport to be “Good” or “Excellent”, while 47% of passengers judged this service to be “Good” or “Excellent”, 44% of them rated it as “Discrete” and 9% of them rated it below that level. Figure 7 shows the passengers’ and employees’ perception of Wi-fi connectivity inside the terminal: only 30% of employees perceived the level of wi-fi connectivity inside the terminal to be “Good” or “Excellent”, while 36.9% of passengers judged this service to be “Good” or “Excellent”, 50.6% of them rated it as “Discrete” and 12.5% of them rated it below that level. Figure 8 shows the passengers’ and employees’ perception of recharge points for mobile devices: 46% of employees perceived the availability of recharge points for mobile devices in public areas to be “Good” or “Excellent”, while 42.9% of passengers judged this service to be “Good” or “Excellent”, 40.9% of them rated it as “Discrete” and 16.2% of them rated it below that level. Figure 9 reports the passengers’ and employees’ perception of the prices in bars and restaurants: 50% of employees perceived the prices in bars and restaurants to be “Good” or “Excellent”, while 42.6% of passengers judged this service to be “Good” or “Excellent”, 47.2% of them rated it as “Discrete” and 10.2% of them rated it below that level. Figure 10 shows the passengers’ and employees’ perception of the availability of vending machines: 57% of employees perceived the availability of vending machines to be “Good” or “Excellent”, while 27.3% of passengers judged this service to be “Good” or “Excellent”, 45.7% of them rated it as “Discrete” and 27% of them rated it below that level. Figure 11 reports the passengers’ and employees’ perception of the airport website: 65% of employees perceived the airport website to be “Good” or “Excellent”, while 38% of passengers judged this service to be “Good” or “Excellent”, 52.3% of them rated it as “Discrete” and 9.7% of them rated it below that level. Figure 12 shows the passengers’ and employees’ perception of airport information points: 50% of employees perceived the level of airport information points to be “Good” or “Excellent”, while 44% of passengers judged this service to be “Good” or “Excellent”, 47% of them rated it as “Discrete” and 9% of them rated it below that level. Figure 13 reports the passengers’ and employees’ perception of the internal sign-posting: 58% of employees perceived the level of internal sign-posting to be “Good” or “Excellent”; while 39.8% of passengers judged this service to be “Good” or “Excellent”, 47.7% of them rated it as “Discrete” and 12.5% of them rated it below that level. Figure 14 shows the passengers’ and employees’ perception of staff skills: 65% of employees perceived the level of staff skills to be “Good” or “Excellent”, while 55.7% of passengers judged this service to be “Good” or “Excellent”, 34.9% of them rated it as “Discrete” and 9.4% of them rated it below that level. Figure 15 reports the passengers’ and employees’ perception of ticket counter services: over 60% of employees perceived the level of ticket counter services to be “Good” or “Excellent”, while 47.5% of passengers judged this level to be “Good” or “Excellent”, 42.6% of them rated it as “Discrete” and 9.9% of them rated it as “Very bad” or “Insufficient”. Figure 16 shows the passengers’ and employees’ perception of check-in waiting times: over 80% of employees perceived the level of check-in waiting times to be “Good” or “Excellent”, while 44.8% of passengers judged this service to be “Good” or “Excellent”, 45.2% of them rated it as “Discrete” and 10% of them rated it below that level. Figure 17 reports the passengers’ and employees’ perception of the airport surface links: over 50% of employees perceived the level of airport surface links to be “Good” or “Excellent”, while 34.1% of passengers judged this service to be “Good” or “Excellent”, 54% of them rated it as “Discrete” and 11.9% of them rated it below that level.

In summary, looking at Figure 2, Figure 3, Figure 4, Figure 5, Figure 6, Figure 7, Figure 8, Figure 9, Figure 10, Figure 11, Figure 12, Figure 13, Figure 14, Figure 15, Figure 16 and Figure 17, the employees expressed a more positive judgment on the service of the airport than did the passengers, while they seemed to be aligned in their perception of airport terminal tidiness and they also seemed to be more critical in aspects related to “Luggage trolley availability”, wi-fi connectivity inside the terminal and recharge points for mobile devices in public areas. The judgment most frequently expressed by the passengers recruited was “Discrete”, followed by “Good”: therefore, they were considered not to be very satisfied, on the whole, with the services offered by the airport, but not particularly dissatisfied either.

Performance on the overall flights regarding punctuality, grouped monthly for 2019, is shown in Figure 18, comparing the largest low-cost carrier (Ryanair) and the full-service carrier (Alitalia) operating from the regional airport considered. We found that in July and August 2019, the performance regarding punctuality was below the target, and this was at a time when the traffic of the airport increased, maintaining the same amount of personnel; the overall punctuality of Ryanair flights was on average higher than the punctuality target, compared to the overall punctuality of Alitalia flights. It is, however, worth noting that the Ryanair punctuality target is lower, at 90%, while the target of the full-service airline Alitalia is 98%.

On the other hand, performance regarding safety indicators on the trend of voluntary and mandatory events occurring during the last eight years is shown in Figure 19. The results of Figure 19 are important, as the ratio between voluntary and mandatory reporting can be used as an indicator of the organization’s safety culture [23]. A voluntary occurrence is submitted by the reporter without any legal, administrative, or financial requirement to do so [24], whereas in mandatory reporting systems, operational personnel are required to report accidents and certain types of incidents specifically by the regulator [25]. Failure to do so can result in legal prosecution. However, learning from near-misses and operational incidents is considered a cornerstone of good safety management practices, as far back as the Heinrich [23] hierarchy; it is recommended that a good safety management system should be sustained by information provided by voluntary reports on minor events, as for every major incident there were many cases of smaller incidents, and behind those smaller incidents there were many near-misses that could occur within the organization, the knowledge and follow up of which can help prevent major occurrences [26]. The lack of voluntary reporting, therefore, in favor of only mandatory reporting is considered to be an index of under-reporting and lack of a good reporting culture within an organization, as pointed out by previous industrial studies [27,28,29], as the opportunity to use reporting to reduce the level of risk within an organization may be missed. Therefore, there could be a false sense of security due to an underestimation of actual occurrences within the organization, as voluntary events have not been reported since 2017.

3.2. Results of the Analysis of Passenger Satisfaction: The Ordinal Logistic Regression (OLR) Model

The estimated results obtained from Stata for the OLR stepwise model are reported in Figure 20. We used the ordinal logistic regression (OLR) model, instead of the ordinal probit regression model, because the OLR model gave us better results in terms of statistical significance of the parameters. The analysis included those attributes significantly different from zero at the 5% significance level, using the stepwise selection method. In the first column: “Overall satisfaction” indicates the dependent variable; “Airport punctuality”, “Terminal tidiness”, “Air conditioning”, “Recharge points”, “Bar restaurant”, “Website”, “Staff skills”, “Ticket counter”, “Sex” and “Age” represent the regressors. At the bottom of the table, “/cut1”, “/cut2”, “/cut3” and “/cut4” are the cut-points that depend on the specificity of the ordinal logistic regression (OLR) model. An OLR model, in fact, can also be interpreted in terms of a latent variable. Specifically, suppose that the manifest response Y_i results from grouping an underlying continuous variable Y*_i using cut-points θ₁ < θ₂ <…< θ_J−1, so that Y_i takes the value 1 if Y*_i is below θ₁, the value 2 if Y*_i is between θ₁ and θ₂, and so on, taking the value J if Y*_i is above θ_J−1. The threshold parameters of 5.88, 8.11, 12.74 and 17.06 tell us that there are five possible values for Y: Y_i = 1 if Y*_i ≤ 5.88; Y_i = 2 if 5.88 ≤ Y*_i ≤ 8.11; Y_i = 3 if 8.11 ≤ Y*_i ≤ 12.74; Y_i = 4 if 12.74 ≤ Y*_i ≤ 17.06 and Y_i = 5 if Y*_i ≥ 17.06.

The other columns show, respectively, the coefficients, the standard errors, the z-test statistic values, the p-values and the confidence intervals, with a 95% level of confidence associated with independent variables and threshold parameters. A one-unit increase in the overall perception of airport services’ reliability and punctuality is associated with a 0.42 increase in the expected value of “Overall satisfaction” on the log odds scale, given that all of the other variables in the model are held constant. A one-unit increase in the perception of terminal tidiness is associated with a 0.73 increase in the expected value of “Overall satisfaction” on the log odds scale, given that all of the other variables in the model are held constant. A one-unit increase in the perception of air conditioning and heating efficiency is associated with a 1.00 increase in the expected value of “Overall satisfaction” on the log odds scale, given that all of the other variables in the model are held constant. A one-unit increase in the perception of recharge points for mobile devices in public areas is associated with a 0.42 increase in the expected value of “Overall satisfaction” on the log odds scale, given that all of the other variables in the model are held constant. A one-unit increase in the perception of availability, quality and prices of bars and restaurants is associated with a 0.60 increase in the expected value of “Overall satisfaction” on the log odds scale, given that all of the other variables in the model are held constant. A one-unit increase in the perception of the airport website is associated with a 0.46 increase in the expected value of “Overall satisfaction” on the log odds scale, given that all of the other variables in the model are held constant. A one-unit increase in the perception of staff skills is associated with a 0.73 increase in the expected value of “Overall satisfaction” on the log odds scale, given that all of the other variables in the model are held constant. A one-unit increase in the perception of ticket counter services is associated with a 0.37 increase in the expected value of “Overall satisfaction” on the log odds scale, given that all of the other variables in the model are held constant. “Sex = 1” (male) is associated with a 0.48 decrease in the expected value of “Overall satisfaction” on the log odds scale, given that all of the other variables in the model are held constant. A one-unit class increase in the variable “Age” is associated with a 0.22 increase in the expected value of “Overall satisfaction” on the log odds scale, given that all of the other variables in the model are held constant. The stepwise selection method allowed us to obtain all p-values < 0.05 and consequently, all regressors significantly different from zero at the 5% significance level. The passengers’ overall satisfaction appeared to be influenced by the perception of air conditioning and heating efficiency/control of environmental conditions within the airport, the skills and usefulness of the staff, the tidiness of the terminal, the availability, quality and prices of bars and restaurant services in the terminal, the quality of the information offered on the airport website, the availability of recharge points for mobile devices in public areas, airport services’ reliability and punctuality, and the ticket counter and information services offered; while among the socio-economic factors, “Sex” and “Age” were useful to explain differences between passengers’ overall satisfaction, unlike “Trip purpose” and “Flight frequency”, which were not found to be relevant in discriminating between different passengers’ expressed satisfaction. In line with the study of Eboli and Mazzulla [21], the services related to the helpfulness of personal safety and security, toilets inside the terminal, internal and external sign-posting, and the availability of city center–airport surface links were not useful to explain the passengers’ overall satisfaction.

The proportional odds ratios for the ordered logistic regression model are obtained by exponentiating the ordered logit coefficients or by specifying the “or” option. For a one-unit increase in the perception of air conditioning available in the airport, the odds of the fifth category of overall satisfaction versus the combined fourth, third, second and first categories are 2.60 times greater, given that all of the other variables in the model are held constant. For a one-unit increase in the perception of staff skills, the odds of the fifth category of overall satisfaction versus the combined fourth, third, second and first categories are 2.16 times greater, given that all of the other variables in the model are held constant. For a one-unit increase in the perception of terminal tidiness, the odds of the fifth category of overall satisfaction versus the combined fourth, third, second and first categories are 2.05 times greater, given that all of the other variables in the model are held constant. For a one-unit increase in the perception of the prices of bars and restaurants, the odds of the fifth category of overall satisfaction versus the combined fourth, third, second and first categories are 1.82 times greater, given that all of the other variables in the model are held constant. For a one-unit increase in the perception of the airport website, the odds of the fifth category of overall satisfaction versus the combined fourth, third, second and first categories are 1.62 times greater, given that all of the other variables in the model are held constant. For a one-unit increase in the perception of recharge points for mobile devices, the odds of the fifth category of overall satisfaction versus the combined fourth, third, second and first categories are 1.55 times greater, given that all of the other variables in the model are held constant. For a one-unit increase in the perception of the airport service punctuality, the odds of the fifth category of overall satisfaction versus the combined fourth, third, second and first categories are 1.50 times greater, given that all of the other variables in the model are held constant. For a one-unit increase in the perception of ticket counter services, the odds of the fifth category of overall satisfaction versus the combined fourth, third, second and first categories are 1.48 times greater, given that all of the other variables in the model are held constant. The distribution of odds ratios obtained from Stata for the OLR stepwise model is shown in Figure 21. Looking at Figure 21, the points indicate the odds ratio values, and the horizontal lines represent the confidence intervals, with a 95% confidence level associated with each attribute without considering socio-economic factors. We found that the quality of air conditioning available in the airport (marked as “Air conditioning” in Figure 20 and Figure 21) showed the most significant weight in terms of odds ratio on the passenger’s overall satisfaction (2.60), followed by “Staff skills” (2.16), “Terminal tidiness” (2.05), “Bar restaurant” (1.82), “Website” (1.62), “Recharge points” (1.55), “Airport punctuality” (1.50) and “Ticket counter” (1.48). The distribution of the average satisfaction of passengers grouped by sex and age is reported in Figure 22, highlighting how young and male passengers were less satisfied than female passengers over the age of 25, in relation to the indicators shown in Figure 2, Figure 3, Figure 4, Figure 5, Figure 6, Figure 7, Figure 8, Figure 9, Figure 10, Figure 11, Figure 12, Figure 13, Figure 14, Figure 15, Figure 16 and Figure 17, except for “Check-in”, “Air conditioning” and “Information points”.

3.3. Results of the Analysis of the Employees’ Perspective

Of the 26 employees recruited for the study, we found that they perceived the airport security level as the best operational aspect (0.69), followed by the clarity of strategic goals (0.62) and the approach to goals (0.62), the business initiatives (0.62) and staff training (0.62), the motivation on goals (0.50) and the corporate security policy (0.42); whereas the critical areas of improvement for the service operator were “Roles and responsibilities” (0.35) and “Communication of objectives” (0.31). The results of the analysis of the employees’ perspective are reported in

Table 2. Results of the analysis of the employees’ perspective.

Category	Indicator	Employees’ Perception
Corporate mission	Clarity of strategic goals	0.62
	Approach to goals	0.62
	Business initiatives	0.62
Organizational structure	Roles and responsibilities	0.35
	Staff training	0.62
Employee participation	Communication of objectives	0.31
	Motivation on goals	0.50
Safety performance	Corporate security policy	0.42
	Airport security	0.69

Source: Authors’ elaboration.

. Employees’ perspective indicators were calculated by dividing the total number of satisfied employees who answered “Discrete” or “Good” or “Excellent”, by the total number of employees interviewed according to the ENAC (Italian Civil Aviation Transport) methodology on a standard Airport Service Charter. The distribution of the indicators obtained from the analysis of the employees’ perspective is shown in Figure 23.

Regarding “Organizational structure”, 62% of the employees were happy about the training provided by the airport, and only 35% of them were satisfied with the matching between roles and responsibilities. “Corporate mission” indicators revealed that 62% of employees were happy about the clarity and approach of strategic goals and the business initiatives, whereas “Employee participation” indicators suggested that 50% of employees were satisfied with the motivation on goals, and only 31% of them were happy about the communication of objectives. Regarding “Safety performance”, 42% of employees were happy about the corporate security policy, and 69% of them were satisfied with the airport security.

3.4. Results of the Survey with Company Managers

The results of the analysis of the managers’ perspective, when asked about the choice of the main areas for improvement between the main critical operational aspects, resulted in the previous analysis reported in

Table 3. Results of the analysis of the managers’ perspective alternatives.

Managerial Area	Choice
Airside & Operations	Roles and responsibilities
Sales & Marketing	Communication of objectives
Financial Administration	Roles and responsibilities
Safety & Compliance Monitoring	Roles and responsibilities
Health & Safety Environment	Roles and responsibilities
Structural Asset	Communication of objectives

Source: Authors’ elaboration.

. We took into account the two worst operational aspects and included them in a choice set with three alternatives: “Roles and responsibilities”, “Communication of objectives” and “Other”. The experimental task is shown in Appendix A (Figure A11): it consisted of choosing which essential aspect to improve among the worst operational aspects ranked by the employees. The experimental subjects were company managers in the following key areas of operations:

• Airside & Operations Manager
• Sales & Marketing Manager
• Financial Administration Manager
• Safety & Compliance Monitoring Manager
• Health & Safety Environment (HSE) Manager
• Structural Asset Manager.

Of the 6 managers recruited, 50% of them were male and 35–44 years old, 33.3% of them belonged to the fourth age class (45–54 years old) and 16.7% of them were concentrated in the fifth age class (>54 years old), whereas the managers’ average period of service was 17 years. 33.3% of them chose “Communication of objectives” as the essential aspect to be improved, whereas 66.7% of them chose “Roles and responsibilities” as the critical area to be improved, and 0% of them chose the “Other” option.

From the managers’ perspective, the redistribution of the workforce for better matching between roles and responsibilities was the main area for improvement, as the middle managers explained that they found their roles to often be stretched to cover areas they are less comfortable with, and with the results of feeling less focused on the main area of competence. The area of communication of objectives was chosen mainly by the Sales & Marketing Manager and the Structural Asset Manager. However, this area was widely reported as needing improvement by the vast majority of the employees interviewed during the survey, therefore it also deserves to be taken into account.

4. Discussion

This study analyzed the satisfaction of the passengers of an Italian regional airport using an ordinal logistic regression (OLR) model. Furthermore, it compares the passengers’ perception with the employees’ perception, from the point of view of the quality of service and the employees’ perspective, in order to identify possible critical areas of improvement for the service operator. The value of this paper lies in the combination of the passengers’ perspective with the employees’ perception, and the insight derived from triangulating them with performance indicators collected for quality and safety metrics. The field study was also completed via a simple survey, based on the literature [20], aiming to identify which areas of improvement are actually chosen from the operational managers’ perspective, among the possible alternatives resulting from the employees’ survey.

Based on the findings obtained from the study of Eboli and Mazzulla [21], we tested the hypothesis of what variables are to be considered significant or not in relation to passengers’ overall satisfaction, using a modified version of the customers’ satisfaction survey for the employees, to enable a comparison of the two perspectives, and considering extra sections for the employees’ survey on operational areas such as the corporate mission and its communication, organizational structure, employee participation, and safety performance. In line with the study of Eboli and Mazzulla [21], the services related to the helpfulness of personal safety and security, toilets inside the terminal, internal and external sign-posting, and the availability of city center–airport surface links were not useful to explain the passengers’ overall satisfaction. The passengers’ overall satisfaction appeared to be influenced by the perception of air conditioning and heating efficiency, staff skills, terminal tidiness, availability, quality and prices of the bars and restaurants, the airport website, recharge points for mobile devices in public areas, airport services reliability and punctuality, and ticket counter services; while among the socio-economic factors, “Sex” and “Age” were useful to explain the differences between passengers’ overall satisfaction, unlike “Trip purpose” and “Flight frequency”, due to their homogeneous distribution. The findings indicate interesting areas of difference in the perceptions of the passengers and airport employees, both useful in highlighting necessary improvements. Employees were more satisfied compared to passengers, except for “Luggage trolley availability”, wi-fi connectivity inside the terminal, and recharge points for mobile devices in public areas. The judgment most frequently expressed by the passengers recruited was “Discrete”, followed by “Good”—therefore, they were not very satisfied on the whole with the services offered by the airport, but not very dissatisfied either. Young and male passengers were less satisfied than female passengers over the age of 25, in relation to the indicators shown in Figure 2, Figure 3, Figure 4, Figure 5, Figure 6, Figure 7, Figure 8, Figure 9, Figure 10, Figure 11, Figure 12, Figure 13, Figure 14, Figure 15, Figure 16 and Figure 17, except for “Check-in”, “Air conditioning” and “Information points”. On the other hand, the judgment most frequently elicited by the employees interviewed was “Good”, followed by “Discrete” (in line with expectations). Those passengers’ perceptions elicited were in line with the 2020 Airport Service Charter target metrics, except for “Toilet tidiness and services”, “Vending machines”, “Internal sign-posting” and “Public information services”. While employees’ perceptions as elicited were in line with the 2020 Airport Service Charter target metrics, except for “Luggage trolley availability”, “Wi-fi connectivity”, “Recharge points”, “Internal sign-posting” and “Public information services”.

The employees expressed higher satisfaction for the airport security’s level of service, and the critical areas of improvement they identified were around “Roles and responsibilities” and “Communication of objectives”.

From the managers’ perspective, the redistribution of the workforce to better match roles and responsibilities was the main area identified for improvement by the middle managers. They explained that they found their roles were often too stretched, covering areas reaching beyond their area of competence, and resulting in a lack of focus on the main core area of responsibility.

Finally, comparing the perceptions elicited on the quality of service by passengers and employees with the performance on quality indicators, they were in line with the 2020 Airport Service Charter target metrics with only minor deviations identified (e.g. “Toilet tidiness and services”, “Internal sign-posting” and “Public information services”), while it was not possible to compare the perceptions elicited on airport safety by employees with the performance on safety indicators, due to a potential issue of under-reporting that can give a false sense of security; therefore, further analysis on the possible lack of a good reporting culture within the company needs to be explored.