The adoption of self-service check-in kiosks among commercial airline passengers

Abstract

This study examines passenger adoption of self-service check-in kiosks based on the technology acceptance model (TAM) and its extension. This study investigated the relationships between three independent variables (technology self-efficacy, need for human interaction, and perceived enjoyment) and two TAM's cognitive variables (perceived usefulness and ease of use). The analysis of these relationships is novel in aviation self-service technology research. Structural equation modeling (SEM) was employed to analyze 346 valid responses. The analysis reveals that enjoyment enhances the perception of self-service check-in kiosks as useful and easy to use, resulting in greater intention to use them. Moreover, passengers with higher confidence in their technological self-efficacy are more likely to perceive check-in kiosks as helpful. However, passengers may perceive kiosks as less useful when they require more human interaction. The results of this study contribute to the current knowledge on aviation self-service adoption. Additionally, managerial implications are offered to improve the future of airlines' passenger services.

1. Introduction

The aviation sector is experiencing rapid growth and is considered as one of the fastest-expanding service industries globally, with over 4 billion passengers in 2019 [1]. The large volume of passenger traffic passing through airports presents a major issue for operators in dealing with congestion and slow service times. Although air travel dramatically decreased during the COVID-19 pandemic, the number of passengers is expected to recover to pre-pandemic levels by 2024 [2]. Therefore, adequate service delivery to passengers is a key management concern. Accordingly, airlines are increasing efforts to improve their facilities and provide better travel experiences for passengers, especially through recent technological advances [3].

Airlines now offer more efficient service models using self-service technologies (SSTs). The International Air Transport Association introduced the Fast Travel Program in 2007 to enhance airport and airline performance. This program includes improvements in check-ins, baggage handling, document verification, booking, boarding, and baggage recovery [4]. Moreover, airlines' information technology investment in passenger services increased in 2021 and was among the top three priorities [5]. This change enables airlines to handle more passengers in the same area by allowing them to perform some tasks independently. SSTs are automated systems that enable users to access and utilize services independently. Early SST concepts have been implemented across various industries, including automatic teller machines, online banking, and supermarket self-checkout machines [6]. Airlines have adopted and integrated different types of SSTs, including check-in channels, such as websites, mobile applications, and kiosks. According to the annual report of Thai Airasia, 50.16 % of passengers adopted self-check-in services, and the most adopted check-in channels were the airline's website (42.68 %), kiosks (33.96 %), and mobile applications (23.36 %) [7]. This suggests that self-check-in options are increasingly popular, with kiosks being a significant, though not the most utilized, channel for self-service check-in. These technologies provide cost savings and operational efficiency and strengthen customer loyalty and satisfaction [[8], [9], [10], [11]].

While it was emphasized that companies can gain a competitive edge through quicker and easily accessible services [12], businesses must also weigh the risks and hidden expenses associated with SST implementation [13]. It was found that SST failures in performing simple tasks can diminish the perceived value of the service and increase staffing costs when customers frequently require assistance from service personnel [14]. Airlines and airports that provide self-service check-in kiosk channels face some challenges. For example, switching from traditional check-in counters to self-service check-in kiosks may pose difficulties for individuals who are not familiar with the technology, consequently leading to decreased personal interaction [15]. The interface and process of using kiosks vary among the airlines passengers choose, leading to potential confusion that should be standardized for ease and enjoyment [16]. Moreover, most airline kiosks are primarily used for check-in processes. Passengers with checked baggage must join another line, either at self-bag drop kiosks or airline check-in counters, which makes self-service check-in kiosks less convenient for some passengers [17]. This attracted researchers to investigate passengers’ adoption of airline self-service check-in kiosks.

Numerous studies have explored aviation SST in general, with specific attention paid to various technologies [[18], [19], [20], [21], [22], [23]]. Some of these studies focused on specific technologies. Several of these studies utilized the technology acceptance model (TAM) and consistently found that usefulness and ease of use significantly affect passenger attitudes toward technology, consequently shaping usage intention [20,[24], [25], [26]]. Moreover, perceived behavioral control [21,27] and subjective norms [20,28] can influence the inclination to use the technology. The use of check-in kiosks is driven by service process [16]. Additionally, age, education, purpose of travel, waiting time, and airline type affect passengers’ decisions to choose their specific check-in channel [15].

Despite abundant research on aviation SSTs, the roles of perceived enjoyment (PENJ), need for human interaction (NHI), and technology self-efficacy (TSE) have received less attention in aviation. Only a few studies have discussed these issues [9,[29], [30], [31]]. While the primary goal of customers is to swiftly and effectively solve their problems with minimal inconvenience, they may also find pleasure in completing tasks [32]. The adoption of SSTs is influenced by the level of enjoyment experienced during their use [9,33]. Effective human interactions between service providers and customers are necessary to provide high-quality services. It was suggested that when offering SSTs, a service provider should consider the significance of human interaction, as it can impact air traveler adoption behavior [26]. However, this element was missing when using SSTs [34]. When performing any task, it is essential for people to have confidence in their abilities to succeed. Individuals who are confident in their competence to fulfill a specific task are more inclined to undertake it than those who are uncertain about their abilities [35]. Therefore, confidence in using technology can increase SST adoption [36].

Additionally, the impact of these three variables on TAM's cognitive variables has seldom been explored in aviation SSTs. A lack of understanding of these factors can result in passenger dissatisfaction, stemming from negative kiosk experiences. The underutilization of kiosks leads to operational inefficiency from an airline's perspective. Therefore, the objectives of this study are as follows:

• 1.Examine the influence of cognitive variables on the intention to use.
• 2.Investigate the influence of the NHI, PENJ, and TSE on cognitive variables.

By examining these connections, this study expands the current TAM to better understand the factors influencing passengers’ intentions to use self-service check-in kiosks. Moreover, this study provides valuable insights for kiosk providers to improve their services, increase passenger use, and enhance the overall travel experience.

1.1. Technology acceptance model (TAM)

Extensive research has been conducted on SST adoption in the aviation industry. Numerous studies have focused on individual technologies such as online ticketing, online check-ins, chatbots, and self-service check-in kiosks. Additionally, numerous studies have applied behavioral intention theories, such as the TAM [37] and extended TAM (TAM2) [38]. Alternative theories have emerged to address technology adoption, including the unified theory of acceptance and use of technology (UTAUT) [39] and technology readiness (TR) [40]. However, these concepts were not widely applied to aviation SST [41].

In 1986, Fred Davis proposed the TAM to explain how people adopt different technologies. The original model suggests that real technology usage is determined by the user's intention, which is shaped by attitudes toward technology usage, based on two cognitive variables: perceived usefulness (PU) and perceived ease of use (PEOU). Both these factors are influenced by external stimuli. PU positively impacts the intention to use technology (INTU). This intention is also affected by PEOU. Several researchers have extensively applied and tested this model and have found it robust [42]. Subsequently, an extended version of the theory, TAM2, was introduced to improve its explanatory power [38]. TAM2 incorporated social influence (subjective norms and voluntariness) and cognitive instrumental factors (demonstrability and job relevancy) into its framework.

Research has validated the connections posited by the TAM. Passenger attitudes toward SSTs are positively affected by PEOU and PU, which consequently influence the passenger's intention. Moreover, PU is influenced by PEOU [20,24,25]. Furthermore, the association between the intention to use and subjective norms was affirmed in the context of airline e-ticketing [20,43].

TAM incorporates additional variables to provide a more robust explanation. The overall TR of passengers positively impacts their attitudes, which could influence their intention to use check-in kiosks [22,44]. Crowded airports and the clarity of the passenger role increased the use of self-service check-in kiosks [45]. Moreover, the study investigated passenger trust in an airline e-ticketing system and found a significant impact on the intention to use it [20,46]. However, heightened passenger concerns about their privacy could reduce their intention to use these service [28,47]. It was discovered that when passengers were forced to use SSTs, their attitude and intention to use were adversely impacted, which could result in carrier-switching intentions [48]. Studies have also found differences in passengers’ demographic characteristics. While business passengers were found to value self-service check-ins more than leisure passengers [11], it was argued that business travelers placed more importance on conventional check-in procedures because they usually have memberships or business class privileges that allow them to spend less time checking in at a special counter [15]. It was also observed that young to middle-aged passengers tend to use SST channels more than older passengers [11,49]. Furthermore, the desire to contact human staff was found to have a negative impact on the use of check-in kiosks [24,26,45]. Additionally, research has verified that PENJ has a favorable impact on attitude, intention, and actual use of SSTs [9,30,31]. PENJ was also found to moderate the level of impact that the intention to use SSTs has on actual use [16].

1.2. Hypothesis development

PEOU refers to the extent to which an individual perceives that utilizing a specific system would involve minimal effort [37]. This variable measures the ease or difficulty of technology adoption for the user. The ease of using technology significantly influences the likelihood of adoption. By contrast, technologies that are difficult to use are more likely to be unpopular. Although this well-known relationship has been discussed in several studies, consensus on the outcome has not been reached. Some studies have indicated a positive correlation between them [50,51], while others have found the opposite [21,28]. Given this inconsistency, it is valuable to reexamine the correlation between PEOU and technology adoption in the context of airline check-in kiosks. Based on this proposition, the authors establish Hypothesis 1 as follows:

Hypothesis 1

PEOU positively affects passengers' INTU.

The extent to which an individual believes that utilizing a specific technology can enhance their work performance is referred to as PU [37]. It measures the level of conviction that using a particular system can provide better results than not using one. If a person believes that technology is useful, they will likely be willing to use it. Usefulness can manifest in faster processes, greater convenience, or improved outcomes. Consequently, an increase in PU leads to increased intention to use technology [21,46,52]. However, some studies have not found a significant relationship between these two variables [26,51]. Additionally, increasing the ease of using technology may lead to greater user exploration and benefits. A positive relationship has been observed between the PEOU and PU [38]. The belief that self-service check-in kiosks are easy to use influences usefulness perception [24,26], leading to Hypotheses 2 and 3:

Hypothesis 2

PU positively affects passengers' INTU.

Hypothesis 3

PEOU positively affects passengers' PU.

In the extended TAM, external variables such as perceived behavioral control and subjective norms are associated with the TAM's cognitive variables [38]. Researchers who adopted the TAM in their field of study discovered several new factors that affect PEOU, PU, and behavioral intention [53]. As part of this study, the authors aim to investigate the impact of three independent variables (NHI, PENJ, and TSE) on cognitive variables in the TAM (Fig. 1). These relationships have not been studied previously from the perspective of aviation SSTs.

Fig. 1.

Theoretical framework adapted from TAM.

Traditionally, providing services has required direct encounters between the user and company's service employee. These technological changes foster service enhancements by integrating SSTs. Despite the long waiting time associated with the traditional service channel, a customer may still opt to receive assistance from human staff. This implies the need to interact with humans, or, in other words, a person's desire to retain interactions with frontline service employees [54,55]. Even in the digital era, many customers value the personal touch provided by frontline service employees [56]. Additionally, customers may choose to receive services from staff to avoid using SSTs [57]. Earlier studies indicate that the need for service from airline employees negatively affects the use of check-in kiosks [24,26,45]. In contrast, airline passengers who want to avoid interacting with staff are more likely to use kiosks [58]. Passengers who prefer face-to-face interaction with staff may find check-in kiosks challenging to use. Moreover, some customers perceive the value of human services as part of their service experience [59]. Therefore, some customers may find SSTs less attractive. It was found that the NHI could reduce the perception that SSTs are useful, and potentially lead to avoidance [60]. Given the established importance of human interaction in service contexts, it is crucial to examine how the need for such interactions influences passengers' perceptions of self-service check-in kiosks. However, the research addressing this issue in the airline industry is limited. Therefore, Hypotheses 4 and 5 are proposed as follows:

Hypothesis 4

NHI negatively affects passengers' PEOU.

Hypothesis 5

NHI negatively affects passengers' PU.

Previous research has highlighted the significance of enjoyment in explaining the acceptance of new technologies. An individual may perform a task expecting it to be enjoyable rather than appreciating the results [33]. PENJ refers to the belief that an experience is pleasurable in itself, without considering any potential performance impacts resulting from using the system [61]. In the context of SSTs, enjoyment is deemed important and valued by customers [55]. For example, one study has shown that the level of enjoyment that a person experiences while using can impact their willingness to continue using Internet Protocol Television [62] and e-shopping re-patronage [63]. Previous studies have revealed the favorable impacts of PENJ on passenger attitude [31], satisfaction, behavioral intention [30], and actual use [9]. When the use of technology creates joy, the difficulty associated with it may be underestimated [64,65], influencing the attitude that the technology is simple to use [66]. Individuals with this belief are more likely to value the advantages of SSTs [54]. In other words, increased enjoyment shapes a person's perception that using technology is less complicated, and thus, more worthwhile. Several studies on mobile payment have identified the significant influence of PENJ on PEOU and PU [67,68]. Similar findings were confirmed in the context of student e-portfolio [69]. However, most studies on the PENJ of technology acceptance have focused on other domains, such as mobile payments, e-shopping, and e-portfolios. This study aims to investigate PENJ in the context of airline kiosks. Therefore, Hypotheses 6 and 7 are proposed as follows:

Hypothesis 6

PENJ positively affects passengers' PEOU.

Hypothesis 7

PENJ positively affects passengers' PU.

To perform a general task, people must believe that they have sufficient skills to accomplish it. Individuals' perceptions of their ability to perform a task, considering their skills and resources, define the concept of self-efficacy [70]. Previous research has found that self-efficacy influences attitudes toward using technology by perceiving technology use as effortless and beneficial [71]. Moreover, behavioral intention is driven by self-efficacy in various settings [[72], [73], [74]]. The concept of self-efficacy was applied in the context of technology, referring to an individual's confidence in effectively using technology [75]. In contrast, people with low TSE believe that their ability to use technology is limited [36]. Self-service check-in kiosks usually require passengers to perform tasks independently, which can either encourage or demotivate their confidence in using the technology. Additionally, several studies have supported the positive links between TSE and PU [[76], [77], [78]] as well as PEOU [36,77,79]. However, research on aviation SST adoption has rarely explored the concept of TSE. Therefore, Hypotheses 8 and 9 are proposed as follows:

Hypothesis 8

TSE positively affects passengers' PEOU.

Hypothesis 9

TSE positively affects passengers' PU.

2. Methods

This study employed a quantitative survey research methodology using online questionnaires to collect data. The researchers began by designing the questionnaire, and a pilot test was conducted to ensure the reliability and validity of the measurement items. The questionnaires were distributed via Google Forms. Finally, covariance-based structural equation modeling (SEM) was used to analyze the data.

2.1. Questionnaire design

The questionnaire consisted of two sections: (1) demographics of the respondents and (2) passenger perceptions of using kiosks. The constructs and measurements were developed following the TAM and previous literature. The adapted theoretical framework consisted of 26 measurement items from six constructs. The core TAM constructs were retrieved from Davis et al. [80] and Demoulin and Djelassi [81], which were previously adopted to investigate the adoption of other aviation SSTs [28,46] and the adoption of self-service check-in kiosks [51]. The measurements for PENJ and NHI were adopted from Ku and Chen [16] and Taufik and Hanafiah [51], respectively. Both these studies focused on self-service check-in kiosks. The measurements for TSE were adopted from Wangpipatwong et al. [36], with minor wording adjustments to fit the context of this study. The authors conducted a reliability analysis of the initial constructs with a sample size of 100. The constructs' Cronbach's alpha ranged from 0.82 to 0.93, which was deemed acceptable [82]. Therefore, the questionnaire was considered reliable for data collection. The respondents were required to answer, “How often do you use the self-check-in counter?” as the screening question. If they chose “never,” they were not required to answer the remaining questions and were excluded from the analysis. This study implemented 7-point Likert scale questionnaire. The respondents who had experience using self-check-in kiosks were required to rate their level of agreement with each item on a scale, where “1” represents “strongly disagree,” “4” represents “neutral,” and “7” represents “strongly agree.”

2.2. Data collection

Online questionnaires hosted on Google Forms were distributed through social media channels to air travelers who had experienced flying in Thailand and who potentially used self-check-in kiosks. An analysis conducted by Kepios showed that approximately 62 % of the general population utilizes social media [83]. Therefore, it is reasonable to assume that many travelers use it. The survey distribution period was from January to March 2023, and the sample size of 346 exceeded the recommended threshold of ten times the number of measurement items [84]. Initially, 447 responses were obtained. However, the authors excluded respondents who had never used kiosks, had incomplete responses, or were outliers. The range of skewness values was between −1.205 and −0.484, while the range of kurtosis was from −0.497 to 1.015. This indicates that the data normality was acceptable, as the skewness and kurtosis values were within the bounds of +2 and −2 [85]. Respondents’ characteristics are presented in Table 1.

Table 1.

Characteristics of respondents (N = 346).

Characteristics	Sub-categories	Frequency	Percentage
Age	18–20 years	31	9.0
	21–35 years	281	81.2
	36–50 years	28	8.1
	≥50 years	6	1.7
Gender	Male	115	33.2
	Female	231	66.8
Education	High school and below	24	6.9
	College diploma	11	3.2
	Bachelor's degree	272	78.6
	Post Graduate degree	39	11.3
Monthly Income	≤9,000 THB	74	21.4
	9,001–15,000 THB	96	27.7
	15,001–25,000 THB	67	19.4
	25,001–35,000 THB	53	15.3
	35,001–45,000 THB	23	6.6
	45,001–55,000 THB	4	1.2
	55,001–65,000 THB	12	3.5
	≥65,001 THB	17	4.9
Frequency of flying	Once or twice a year	200	57.8
	3–4 times a year	77	22.3
	More than 4 times a year	69	19.9

Note:1 USD = 36 THB (approx.).

2.3. Data analysis

First, respondents’ demographics were examined using frequency and descriptive statistics. Next, following a two-stage method [86], separate analyses were performed: the measurement model was first, followed by the structural model. The authors implemented a confirmatory factor analysis (CFA) to evaluate how well the data fit the measurement model. Convergent and discriminant validity were addressed at this point, along with possible common method bias (CMB). Finally, using IBM SPSS AMOS 29, the authors tested the structural model and hypotheses using SEM.

3. Results

3.1. Assessment of the measurement model

The CFA results illustrate the goodness-of-fit indices of the framework (Table 2). The chi-square (X²) statistic was 741.81, with 284 degrees of freedom (df) (p < 0.001). The X²/df ratio was 2.52, which was less than 3 [87]. The root mean square error of approximation was below 0.08 (RMSEA = 0.066), and the standardized root mean square residual was less than 0.05 (SRMR = 0.036), suggesting an adequacy of fitness [88]. The comparative fit index (CFI = 0.948), incremental fit index (IFI = 0.948), normed fit index (NFI = 0.916), and Tucker-Lewis index (TLI = 0.94) were all above the suggested cutoff point of 0.9 [89]. Cronbach's Alpha and composite reliability fell between 0.88 and 0.94, indicating acceptable reliability of the constructs [82,90]. The average variance extracted (AVE) was evaluated and found to exceed a 0.5 threshold, affirming the convergent validity [91].

Table 2.

Measurement model assessment result from CFA.

Constructs/Items	Item-Construct Loading
	Standardized	t-statistics
Perceived Usefulness (PU) (α = 0.94, C.R. = 0.94, AVE = 0.74)
Using self-check-in kiosks enhances my convenience on check-in.	0.86	a
Using self-check-in kiosks improves my check-in process.	0.88	22.1
Using self-check-in kiosks enables me to accomplish check-in more quickly.	0.88	22.1
I find self-check-in kiosks useful in the check-in process.	0.86	21.5
I can save time by using the self-check-in kiosks.	0.83	19.9
Perceived Ease of Use (PEOU) (α = 0.93, C.R. = 0.94, AVE = 0.79)
Interacting with self-check-in kiosks does not require a lot of my mental effort.	0.81	a
I find the self-check-in kiosks to be easy to use.	0.93	21.8
The self-check-in kiosk instructions are clear and understandable.	0.94	21.9
I find it easy to check in using self-check-in kiosks.	0.86	19.3
Perceived Enjoyment (PENJ) (α = 0.91, C.R. = 0.91, AVE = 0.73)
I feel good being able to use self-check-in kiosks.	0.75	a
I was very happy when I used self-check-in kiosks.	0.90	17.7
I thought that the content of self-check-in kiosks was very interesting.	0.84	16.2
I enjoyed using self-check-in kiosks.	0.91	17.9
Need for Human Interaction (NHI) (α = 0.88, C.R. = 0.88, AVE = 0.65)
My self-check-in kiosk experience will be much better with help from a real person.	0.68	a
Having human contact in providing services makes the process enjoyable for the customer.	0.87	14.1
Personalize attention from the service employee is important to me.	0.88	14.2
I like interacting with a real person who provides the service.	0.78	12.9
Intention to Use (INTU) (α = 0.92, C.R. = 0.92, AVE = 0.70)
I will continue to use self-check-in kiosks in the future.	0.79	a
I intend to use self-check-in kiosks, which enable me to accomplish my check-in process more quickly.	0.82	17.1
I will use self-check-in kiosks rather than the check-in counter to complete my check-in process.	0.79	16.3
Given that I have access to the self-check-in kiosks, I predict that I would use them.	0.90	19.3
Assuming I have access to the self-check-in kiosks, I intend to use them.	0.89	18.9
Technology Self-efficacy (TSE) (α = 0.93, C.R. = 0.93, AVE = 0.76)
I feel confident using communication devices that I have never used before.	0.88	a
I feel confident using software that I have never used before.	0.89	23.2
I feel confident troubleshooting technological problems.	0.88	22.7
I feel confident working on technology although there was no one assisting.	0.85	21.5

Note; α = Cronbach's Alpha; C.R. = Composite Reliability; AVE = Average Variance Extracted.

^aItems constrained for identification purposes.

Table 3 presents the results of the discriminant validity test. The correlations were lower than the square root of the AVE, meeting the Fornell-Larcker criterion [91]. Furthermore, the Heterotrait-Monotrait (HTMT) criterion, a modern approach to examining the discriminant validity problem [92], was used in this study. All HTMT ratios were below 0.85, as suggested by Kline [87]. Hence, this study confirms the fit and reliability of the measurement model.

Table 3.

Result of discriminant validity tests.

Constructs	Means	SD	PU	PEOU	PENJ	NHI	INTU	TSE
PU	6.01	1.00	0.86	0.81	0.74	0.25	0.74	0.60
PEOU	5.74	1.13	0.81	0.89	0.77	0.31	0.75	0.67
PENJ	5.60	1.12	0.73	0.76	0.85	0.40	0.72	0.58
NHI	5.55	1.12	0.24	0.30	0.38	0.81	0.34	0.14
INTU	5.73	1.02	0.74	0.75	0.70	0.35	0.84	0.60
TSE	5.53	1.17	0.60	0.68	0.58	0.15	0.60	0.87

Note: Values arranged diagonally were square roots of the AVE; Downside of the triangle was Fornell-Larcker criterion, with all correlations significant at 0.01 level except TSE and NHI at 0.05 level; Upside of the triangle was Heterotrait-Monotrait (HTMT) ratio.

It is important to address CMB while conducting research on human behavior, as it could lead to measurement errors and, consequently, biased results. CMB could arise from respondents' efforts to maintain consistency in their answers or meet social expectations, rather than answering based on genuine feelings [93]. Although Harman's single-factor method has been widely used in service research [23,94,95], its usefulness and appropriateness have been criticized, and other statistical remedies have been suggested instead [93,96]. Hence, the authors opted to use the common latent factor (CLF) method to examine the variations between the model that contains the CLF and one that does not. The CLF method has been adopted by many recent studies [[97], [98], [99]]. The test illustrated a significant difference between the model with CLF (X² = 692.31, and df = 283) and the model without CLF (X² = 714.81, and df = 284), as the change in 1 degree of freedom resulted in more than a 3.84 change in the X². Therefore, the CLF was included in the structural model and hypotheses testing stage to control for potential CMB [100].

3.2. Structural model and hypothesis testing

The authors opted for IBM SPSS AMOS 29 Graphics software to test the hypotheses. As shown in Table 4, the goodness-of-fit indices illustrate an adequate fit between the structural model and the data, thus warranting further hypothesis testing. The squared multiple correlations (R²) illustrated how well the predictor variables explained the target variables. Overall, the R² values for PU, PEOU, and INTU were 55.6 %, 51.7 %, and 40.4 %, respectively.

Table 4.

Structural model fit statistics and hypothesis testing.

Hypotheses	Path	β	Std. Error	t-value	p	Decision
H1	PEOU → INTU	0.446	0.067	5.0	a	Accepted
H2	PU → INTU	0.237	0.071	2.7	a	Accepted
H3	PEOU → PU	0.552	0.077	6.6	a	Accepted
H4	NHI → PEOU	0.006	0.072	0.1	0.916	Rejected
H5	NHI → PU	−0.199	0.084	−2.7	a	Accepted
H6	PENJ → PEOU	0.507	0.099	7.2	a	Accepted
H7	PENJ → PU	0.224	0.093	3.1	a	Accepted
H8	TSE → PEOU	0.346	0.049	5.6	a	Accepted
H9	TSE → PU	−0.006	0.051	−0.1	0.933	Rejected
Squared Multiple Correlation (R2):
PU		0.556
PEOU		0.517
INTU		0.404
Structural Model Fit Statistics:
X2 = 702.71, df = 286, p < 0.001, X2/df = 2.46, RMSEA = 0.065, SRMR = 0.038
CFI = 0.949, IFI = 0.95, NFI = 0.918, TLI = 0.942

Note.

β = Standardized Beta Coefficient.

^ap < 0.01.

To begin with TAM's main constructs, the effect of PEOU on INTU (β = 0.446, t-value = 5, and p=<0.01) and PU (β = 0.552, t-value = 6.6, and p < 0.01) were both positive and significant, thereby proving H1 and H3. The positive effect of PU on INTU was also significant (β = 0.237, t-value = 2.7, and p < 0.01), thus confirming H2.

While the effect of NHI on PEOU was not significant (β = 0.006, t-value = 0.1, and p = 0.916), its adverse effect on PU was significant (β = −0.199, t-value = −2.7, and p=<0.01). Hence, H4 was rejected, and H5 was supported. Furthermore, PENJ had a positive and significant impact on PEOU (β = 0.507, t-value = 7.2, and p < 0.01) and PU (β = 0.224, t-value = 3.1, and p < 0.01), thus proving H6 and H7. Finally, the TSE had a significant positive impact on PEOU (β = 0.346, t-value = 5.6, and p < 0.01), but its impact on PU was not significant (β = −0.006, t-value = −0.1, and p < 0.933), therefore confirming H8 and rejecting H9. The hypothesis testing is summarized and illustrated in Fig. 2.

Fig. 2.

The summarized hypotheses testing.

4. Discussion

The study's results offer significant theoretical insights into the factors influencing technology acceptance in the aviation industry, particularly in the context of SSTs such as check-in kiosks. Our findings reinforce and extend the established TAM by highlighting the crucial role of PENJ in shaping perceptions of ease of use and usefulness, which subsequently impact intention to use. Additionally, our results reveal a complex interplay between NHI and PU. This suggests that in the aviation realm, the desire for human assistance may be more closely tied to perceptions of value and functionality than to ease of use. Finally, while confirming the influence of the TSE on PEOU, our study challenges previous findings regarding its impact on PU. This indicates that the expectations of outcomes from technology use may be independent of individual technological confidence in this context.

This study confirms the relationship between NHI and PU (H5). This interesting finding suggests that passengers who have a greater NHI perceive kiosks as less useful. One possible reason for this outcome is that passengers who have a greater desire for human interaction may believe that service staff are more helpful than SSTs, such as kiosks. Another possible reason could be that when passengers have a stronger desire for human interaction, they tend to avoid using kiosks [60], which may hinder their perception of kiosks’ usefulness. Passengers may presume that human staff can respond better to their requests than kiosks. However, the findings of this study do not suggest a noteworthy correlation between the desire for human interaction and easy-to-use perception of passengers (H4). The desire for human interaction and perception of ease of use may address different passenger needs. While ease of use focuses on the functional aspects of the kiosk (e.g., clear instructions and intuitive navigation), the desire for human interaction may emerge from the need for emotional reassurance, personalized assistance, or complex problem-solving that kiosks might not be able to provide.

These findings suggest a correlation between perception of enjoyment and perception of ease of use (H6). Passengers who find using kiosks enjoyable may perceive kiosks as easy to use. This may have resulted from cognitive dissonance, a psychological concept that refers to the discomfort experienced when holding conflicting beliefs [101]. In other words, passengers who believe that kiosks offer an enjoyable experience may tend to adjust their perceptions of the kiosks' ease of use to align with their positive perceptions of enjoyment. Additionally, the results suggest a significant correlation between PENJ and PU (H7). This relationship may be influenced by motivated reasoning in which individuals are driven to form and maintain beliefs that align with their desired outcomes or emotions [102]. In this context, passengers who enjoy using kiosks may be motivated to view them as useful because this belief reinforces their positive experiences. This could lead to a subconscious emphasis on the kiosks’ benefits while minimizing any perceived drawbacks.

This study also showed an association between higher TSE and higher PEOU (H8). This result indicates that passengers who are confident in their technological skills are more likely to approach new technology with confidence and are thus less likely to perceive difficulty in using kiosks [36]. For example, users can easily follow the guidelines displayed on the screen. However, PU is not significantly influenced by TSE (H9), which contradicts the results of previous studies [76,78]. A possible explanation could be that regardless of high or low TSE, passengers expect the same outcome from using kiosks, leading them to perceive the same levels of usefulness. This outcome prompts a reevaluation of the role of TSE in the TAM and highlights the need for further research to understand the specific factors that drive PU in the aviation industry.

These findings confirm those of prior research [21,24,26,30,52,54,103,104], demonstrating a strong link between PEOU and passengers’ intention to use kiosks (H1). This aligns with both the theory of planned behavior (TPB) and the TAM, which emphasize the role of perceived behavioral control and PEOU in shaping behavioral intentions. In the context of kiosks, a user-friendly interface, clear instructions, and minimal steps can enhance the PEOU, ultimately leading to increased adoption.

Furthermore, the study reveals a significant positive relationship between PU and kiosk usage intentions (H2). This finding is consistent with the UTAUT, which posits that performance expectancy (i.e., PU) is a key determinant of behavioral intention. By emphasizing the benefits of kiosks, such as time savings, convenience, and greater control over the check-in process, airports can tap into passengers’ desire for efficiency and enhance their intention to use these self-service options. Interestingly, this study found a strong correlation between PEOU and PU (H3), which echoes the findings of Lu et al. [26]. This finding aligns with both the TAM and UTAUT, which acknowledge the interplay between these two constructs in shaping technology acceptance.

This study has several managerial implications. Although check-in kiosks are available at many airports and airlines, passengers require human assistance. Traditional check-in counters are required for this purpose. However, the non-significant correlation between the NHI and PEU suggests that deploying an abundance of staff solely to assist passengers at kiosks may not be necessary. Nevertheless, it is still important to have a few staff members available to address any issues or complex inquiries that may arise to ensure a smooth and positive passenger experience. Thus, there should be a gradual transition from traditional check-in counters to kiosks. Kiosks need to be useful, easy to use, and genuinely deliver these attributes. Clear and easy-to-understand instructions should be provided to enable passengers to follow the process stepwise. Additionally, airports and airlines should actively communicate with passengers to ensure that the use of kiosks is not challenging and offers several benefits. This can be achieved through promotional materials, such as video clips and brochures, highlighting the kiosk's user-friendliness and advantages. These materials can be integrated into in-flight announcements or boarding passes. Moreover, the experience of using the kiosks should be enjoyable. Therefore, the physical design of kiosks should be engaging, whereas the user interface should be playful and vivid. Airlines can customize the visual design of kiosks by incorporating the local culture or the airline's brand through colorful artwork. Simultaneously, the user interface can be designed to allow passengers to personalize their check-in experience by providing animated instructions featuring a passenger avatar, adding fun and individualization.

This study provides valuable insights into the factors that influence Thai passengers' adoption of self-service check-in kiosks. However, it is important to acknowledge the limitations of this study. The exclusive focus on Thai passengers raises questions regarding the generalizability of the findings to travelers from diverse cultural backgrounds. Future research should investigate the influence of cultural factors on technology acceptance by including participants from other countries. Additionally, by focusing on passengers with prior experience using self-service check-in kiosks, this study may not fully capture the challenges faced by technologically inexperienced users. Future studies should explicitly include participants who are less familiar with such technology to identify potential barriers to its adoption. Furthermore, as this study provides a cross-sectional analysis of the intention to use, longitudinal research designs would allow for a more comprehensive understanding of the dynamic relationship between intention to use and actual use over time. Finally, while the TAM offers valuable insights, other frameworks, such as the UTAUT and TPB, could provide alternative perspectives on technology adoption. Future research could apply these frameworks and potentially consider integrating them, such as the emotional TAM, to gain a more comprehensive understanding of the factors influencing passengers’ acceptance of self-service check-in kiosks.

Compliance with ethical standards

• •This study was reviewed and approved by Mae Fah Luang University Ethics Committee on Human Research, with approval number EC 23017-12.
• •All participants in the online questionnaire were informed that their participation was entirely voluntary. They were given the option to decline participation at any point. Before proceeding to the main questions, participants were asked to confirm their consent to participate in the survey.

Data availability

Data will be made available on request.

CRediT authorship contribution statement

Phutawan Ho Wongyai: Writing – original draft, Project administration, Methodology, Investigation, Formal analysis, Conceptualization. Kamonpat Suwannawong: Data curation, Conceptualization. Panisa Wannakul: Data curation, Conceptualization. Teeris Thepchalerm: Writing – original draft, Validation, Project administration, Investigation, Formal analysis, Conceptualization. Tosporn Arreras: Visualization, Methodology, Formal analysis.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Appendix A. Supplementary data

The following is/are the supplementary data to this article: