Javascript is required
1.
E. Wąsowicz-Zaborek, “Content analysis of hotel reviews as a quality management tool: Preliminary verification of the SERVQUAL scale,” Turyzm/Tourism, vol. 33, no. 1, pp. 29–40, 2023. [Google Scholar] [Crossref]
2.
A. Parasuraman, V. A. Zeithaml, and L. L. Berry, “A conceptual model of service quality and its implications for future research,” J. Mark., vol. 49, no. 4, pp. 41–50, 1985. [Google Scholar] [Crossref]
3.
N. Ayvaz-Çavdaroğlu, S. Iyanna, and M. Foster, “Smart service quality in hospitality–A quantitative assessment using MCDM and clustering methods,” Int. J. Hosp. Manag., vol. 123, p. 103931, 2024. [Google Scholar] [Crossref]
4.
R.-C. Jou and Y.-J. Day, “Application of revised importance–performance analysis to investigate critical service quality of hotel online booking,” Sustainability, vol. 13, no. 4, p. 2043, 2021. [Google Scholar] [Crossref]
5.
J. Wojciechowska-Solis, M. Śmiglak Krajewska, and D. Viti, “Identification of the quality gaps in the services offered in accommodation facilities in rural areas: The case of the Lublin region,” Sustainability, vol. 14, no. 24, p. 16873, 2022. [Google Scholar] [Crossref]
6.
M. Kozłowski and J. Korzeniewski, “Studying the effectiveness of investments in hotel services using customer sentiment analysis,” Qual. Quant., vol. 60, no. 3, pp. 8013–8041, 2026. [Google Scholar] [Crossref]
7.
A. Parasuraman, V. A. Zeithaml, and L. L. Berry, “SERVQUAL: A multiple-item scale for measuring consumer perceptions of service quality,” J. Retail., vol. 64, no. 1, pp. 12–40, 1988. [Google Scholar]
8.
Y. Adhikari, M. P. Wagley, and N. Dahal, “Developing the contextualized SERVQUAL instrument for measuring the service quality of Nepali resorts: An application of the modified Delphi technique,” SAGE Open, vol. 13, no. 3, p. 21582440231195952, 2023. [Google Scholar] [Crossref]
9.
A. Baquero, “Is customer satisfaction achieved only with good hotel facilities? A moderated mediation model,” Adm. Sci., vol. 13, no. 4, p. 108, 2023. [Google Scholar] [Crossref]
10.
M. A. I. Gazi, M. A. Islam, A. A. Masud, A. R. B. S. Senathirajah, S. Biswas, and R. A. Shuvro, “The moderating impacts of COVID-19 fear on hotel service quality and tourist satisfaction: Evidence from a developing country,” Cogent Soc. Sci., vol. 10, no. 1, p. 2331079, 2024. [Google Scholar] [Crossref]
11.
M. H. Asgeirsson, T. Gudlaugsson, and G. T. Jóhannesson, “The relationships between service quality, reputation, and performance in hospitality,” Tour. Hosp., vol. 5, no. 3, pp. 736–752, 2024. [Google Scholar] [Crossref]
12.
A. Mohsin, H. Rodrigues, and A. Brochado, “Shine bright like a star: Hotel performance and guests’ expectations based on star ratings,” Int. J. Hosp. Manag., vol. 83, pp. 103–114, 2019. [Google Scholar] [Crossref]
13.
D. Buhalis and R. Law, “Progress in information technology and tourism management: 20 years on and 10 years after the Internet—The state of eTourism research,” Tour. Manag., vol. 29, no. 4, pp. 609–623, 2008. [Google Scholar] [Crossref]
14.
A. F. Alnaim, A. E. E. Sobaih, and I. A. Elshaer, “Measuring the mediating roles of E-Trust and E-Satisfaction in the relationship between E-Service Quality and E-Loyalty: A structural modeling approach,” Mathematics, vol. 10, no. 13, p. 2328, 2022. [Google Scholar] [Crossref]
15.
N. Chotisarn and T. Phuthong, “Impact of artificial intelligence-enabled service attributes on customer satisfaction and loyalty in chain hotels: Evidence from coastal tourism destinations in western Thailand,” Soc. Sci. Humanit. Open, vol. 11, p. 101306, 2025. [Google Scholar] [Crossref]
16.
L. Chen, C. F. Goh, Y. M. Low, O. K. Tan, and K. Y. Lim, “From service to intention: A Technology Acceptance Model perspective on e-service quality in online hotel booking,” SAGE Open, vol. 15, no. 4, p. 21582440251397834, 2025. [Google Scholar] [Crossref]
17.
O. S. Olawuyi and C. Kleynhans, “A bibliometric analysis of service quality in the hospitality industry (2014–2024),” Adm. Sci., vol. 15, no. 6, p. 215, 2025. [Google Scholar] [Crossref]
18.
V. Perdomo-Verdecia, M. Sacristán-Díaz, and P. Garrido-Vega, “RbQUAL: Integrating SERVQUAL and BERT to analyse service quality dimensions in hotel online reviews,” J. Hosp. Tour. Insights, vol. 9, no. 11, pp. 218–237, 2026. [Google Scholar] [Crossref]
19.
J. A. Martilla and J. C. James, “Importance-performance analysis,” J. Mark., vol. 41, no. 1, pp. 77–79, 1977. [Google Scholar] [Crossref]
20.
P. R. Lingadkar and K. G. Sankaranarayanan, “Hospitality and tourism internship programme: Expectation versus reality,” Afr. J. Hosp. Tour. Leis., vol. 12, no. 3, pp. 911–926, 2023. [Google Scholar] [Crossref]
21.
I. K. W. Lai and M. Hitchcock, “Importance–performance analysis in tourism: A framework for researchers,” Tour. Manag., vol. 48, pp. 242–267, 2015. [Google Scholar] [Crossref]
22.
B. B. Boley and E. Jordan, “Leveraging IPA gap scores to predict intent to travel,” J. Hosp. Tour. Manag., vol. 57, pp. 97–101, 2023. [Google Scholar] [Crossref]
23.
V. Perdomo-Verdecia, P. Garrido-Vega, and M. Sacristán-Díaz, “An fsQCA analysis of service quality for hotel customer satisfaction,” Int. J. Hosp. Manag., vol. 122, p. 103793, 2024. [Google Scholar] [Crossref]
24.
K. Kalnaovakul and P. Promsivapallop, “Hotel service quality dimensions and attributes: An analysis of online hotel customer reviews,” Tour. Hosp. Res., vol. 23, no. 3, pp. 420–440, 2023. [Google Scholar] [Crossref]
25.
J. Charan and T. Biswas, “How to calculate sample size for different study designs in medical research?,” Indian J. Psychol. Med., vol. 35, no. 2, pp. 121–126, 2013. [Google Scholar] [Crossref]
26.
A. M. Gadermann, M. Guhn, and B. D. Zumbo, “Estimating ordinal reliability for Likert-type and ordinal item response data: A conceptual, empirical, and practical guide,” Pract. Assess. Res. Eval., vol. 17, no. 3, pp. 1–13, 2012. [Google Scholar] [Crossref]
27.
B. D. Zumbo, A. M. Gadermann, and C. Zeisser, “Ordinal versions of coefficients alpha and theta for Likert rating scales,” J. Mod. Appl. Stat. Methods, vol. 6, no. 1, pp. 21–29, 2007. [Google Scholar] [Crossref]
28.
Y. F. Zakariya, “Cronbach’s alpha in mathematics education research: Its appropriateness, overuse, and alternatives in estimating scale reliability,” Front. Psychol., vol. 13, p. 1074430, 2022. [Google Scholar] [Crossref]
29.
K. S. Taber, “The use of Cronbach’s alpha when developing and reporting research instruments in science education,” Res. Sci. Educ., vol. 48, pp. 1273–1296, 2018. [Google Scholar] [Crossref]
30.
Y. Yang, L. Wang, and J. Yan, “Perceived service quality mediates the effect of physical servicescape on tourists’ intention to stay in rural homestays,” Sci. Rep., vol. 15, p. 29506, 2025. [Google Scholar] [Crossref]
Search
Research article

Prioritizing Hotel Service Improvements Using SERVQUAL and Sensitivity Analysis of Importance–Performance Classifications

Sema Kayapınar Kaya1,
Sıla Hayva1,
Çağlar Karamaşa2*
1
Department of Industrial Engineering, İzmir Demokrasi University, 35140 İzmir, Türkiye
2
Department of Business Administration, Anadolu University, 26470 Eskişehir, Türkiye
Journal of Operational and Strategic Analytics
|
Volume 4, Issue 4, 2026
|
Pages 254-267
Received: 09-17-2026,
Revised: 11-02-2026,
Accepted: 11-17-2026,
Available online: 11-23-2026
View Full Article|Download PDF

Abstract:

Hotel managers need defensible priorities when service failures compete for limited operational resources. This study examines whether hotel service-improvement priorities remain consistent across individual attributes, reporting domains, and alternative importance–performance analysis (IPA) specifications. A voluntary survey at a three-star city hotel in İzmir, Türkiye, conducted from November 2025 to March 2026, yielded 411 questionnaires covering 22 paired expectation and perception items. An assessment using an adapted service quality scale (SERVQUAL) was combined with paired comparisons, effect sizes, a conservative multiplicity check, and sensitivity analysis of domain composition and IPA crosshairs. Seventeen items differed at the nominal 5% level, and at least 12 met the Bonferroni threshold for 22 comparisons. Common-area cleanliness (-1.826), room cleanliness (-1.367), equipment functionality (-1.023), and reservation accuracy (-0.810) showed the largest negative gaps. Tangibles and reliability remained improvement priorities under all four specifications examined. Excluding a broadly worded security-measures item changed the digital-services gap from -0.175 to approximately -0.315 and moved the domain into the improvement-priority quadrant. Domain expectation means ranged from 4.895 to 4.954, which restricted discrimination along the importance axis. The findings show how item-level evidence and specification checks help interpret priorities derived from near-ceiling expectations in this hotel. The analysis provides a practical basis for selecting housekeeping, maintenance, reservation, and data-handling processes for operational review, while distinguishing persistent priorities from classifications that depend on measurement choices.

Keywords: Hotel service operations, Service improvement prioritization, SERVQUAL, Importance–performance analysis, Sensitivity analysis, Operational decision support

1. Introduction

Hotel managers must decide which service failures to address when housekeeping, maintenance, reception, and digital transactions compete for staff time and managerial attention. The lowest-rated attribute provides a useful starting point, but its score alone does not establish the order in which corrective work should proceed. Managers also need to know whether an apparent priority persists when individual ratings are combined into broader service domains and when the reference points used for comparison change. Analysis of hotel reviews has shown that guests evaluate a broad set of physical, interpersonal, and process attributes rather than a single undifferentiated service outcome [1]. A decision framework must retain this detail while producing a manageable set of processes for investigation.

The service quality scale (SERVQUAL) assesses service quality by comparing expected and perceived service [2]. This paired structure helps locate shortfalls, although translating those shortfalls into operational priorities requires choices about measurement and aggregation. Subsequent hospitality research has incorporated smart service quality and examined how conventional and technology-enabled services differ across customer groups [3]. Work on online hotel booking has also linked website functionality, information, customer relations, and privacy to service-improvement priorities [4]. These studies establish the relevance of digital encounters to hotel operations. They also raise a practical question: when favorable ratings for one attribute offset unfavorable ratings for another, does the resulting domain score still identify the process that needs attention?

Expectation–performance gaps identify perceived deficiencies, but they do not determine the cost, feasibility, or likely return of a corrective action. A rural-accommodation study found shortfalls across the conventional SERVQUAL dimensions [5], whereas research combining investment information with customer sentiment addressed the separate question of whether expenditure was associated with better evaluations [6]. For managers, the distinction separates diagnosis from resource allocation. Survey evidence can identify cleaning, equipment, reservation, or transaction processes that warrant review; selecting an investment then requires information about causes, operating constraints, and costs. Before that stage, a more immediate requirement is to establish whether the diagnostic priorities themselves depend on the way the survey results are summarized.

This study examines service-improvement priorities at a three-star city hotel in İzmir using an adapted SERVQUAL instrument [7]. The questionnaire was informed by contextualized instrument development in resort services [8] and covered physical conditions, service execution, staff interaction, personalized attention, and digital services. The analysis addresses three connected questions. Which attributes show the largest expectation–performance shortfalls? Do domain-level summaries preserve the operational information contained in those attributes? How sensitive are the resulting importance–performance analysis (IPA) priorities to near-ceiling expectations and to the inclusion of a general security item within digital services? The sequence moves from identifying specific deficiencies to examining whether aggregation and classification preserve their relevance for local managerial decisions.

The contribution is a case-based assessment of the consistency of service-improvement priorities, rather than a new service-quality scale or an estimate of investment returns. The analysis compares item-level shortfalls with domain-level classifications, distinguishes nominal statistical findings from conservative familywise evidence, and tests alternative IPA reference points and a narrower definition of digital services. This combination identifies which priorities persist across the specifications examined and where a domain label conceals a specific shortfall. The findings are then translated into processes to review and indicators to monitor. This links the survey to operational decision support while keeping the recommendations within the evidence provided by a single-hotel, cross-sectional assessment.

2. Literature and Analytical Positioning

2.1 Service Quality Across Physical and Interpersonal Processes

Hotel service quality depends on the conditions of the stay and on how employees manage requests, problems, and expectations. Evidence linking hotel facilities to satisfaction also identifies roles for personnel and business organization [9]. A physical deficiency may therefore require coordination across functions: equipment failure involves maintenance and room inspection, while a reservation discrepancy can involve the booking interface, front desk, and housekeeping.

SERVQUAL measures tangibles, reliability, responsiveness, assurance, and empathy through paired expectation and perception items [7]. Adapting the wording to a hotel setting can improve relevance, but it can also alter the meaning of a dimension. The modified Delphi approach used for Nepali resorts illustrates the role of contextual assessment in instrument development [8]. Here, the adapted dimensions serve as operational reporting domains; their labels do not establish that the original five-factor structure has been reproduced.

The relationship between quality and satisfaction also depends on the service context. Research on hotel infrastructure and employee-related attributes has examined satisfaction under pandemic-related conditions [10], while a study of Icelandic hotels distinguished tangible and performed quality and related both to satisfaction and loyalty [11]. Differences across star categories further caution against transferring priorities between hotel segments without local evidence [12]. These studies justify a case-specific diagnosis, although a single-hotel survey cannot establish the priorities of the wider accommodation sector.

2.2 Digital Service Quality Within the Guest Journey

Information technology has reshaped tourism distribution and service delivery over several decades [13]. In hotels, the relevant processes include access to booking channels, reservation information, payment, and handling of personal data. E-service research has linked service quality to trust, satisfaction, and loyalty [14], while research on AI-enabled hotel services has examined relationships among service attributes, satisfaction, and loyalty [15]. Such findings support examining digital encounters, but they do not imply that a new technology is necessarily the most urgent improvement.

Recent online-booking research integrates efficiency, system availability, information quality, and privacy with technology acceptance and managerial priority assessment [16]. A hospitality bibliometric analysis likewise documents the growing role of AI and text-based methods [17]. The Review-based Quality (RbQUAL) uses Bidirectional Encoder Representations from Transformers (BERT)-based classification to interpret hotel reviews through service-quality dimensions [18]. These approaches extend the available data and methods, whereas the present survey provides paired evaluations of specified attributes. It therefore complements review-based monitoring without claiming comparable scale or predictive capability.

Digital service measurement depends on the meaning of the attributes being combined. Personal-data protection and transaction security concern information handling directly. A general security question may also capture physical safety or surveillance. To assess the effect of this ambiguity, the analysis compares the four-item digital domain with a narrower three-item definition that excludes general security.

2.3 From Gap Diagnosis to Defensible Priorities

The original IPA matrix compares importance and performance using four quadrants [19]. Its appeal is practical: managers can distinguish relatively weak, important attributes from well-performing ones. However, importance measurement and crosshair placement are analytical choices. An expectation-based application in hospitality and tourism internships provides precedent for using expectations as an importance proxy [20], but its student context does not validate equivalence between expectations and hotel guests’ importance judgments.

Methodological work on IPA emphasizes reliability, validity, and the interpretation of importance [21]. Gap scores have also been used to explain travel intentions [22], while configurational analysis examines combinations of service conditions associated with satisfaction [23]. These approaches address different decision questions. The present analysis focuses on identifying local service shortfalls and testing whether their priority classifications depend on aggregation choices.

Hotel review research has identified context-specific service dimensions and differences across customer groups [24]. This reinforces the need to inspect the attributes within a domain before applying a general label. Table 1 locates the present analysis among selected approaches. Its distinctive emphasis is the agreement, or disagreement, between item-level shortfalls and dimension-level priority classifications.

Table 1. Selected approaches to hotel service-quality analysis

Study

Evidence and Method

Relevance to the Present Decision Problem

[3]

300 participants; extended SERVQUAL, best–worst method, and clustering

Distinguishes smart service quality from conventional domains and compares preferences across groups.

[4]

453 foreign tourists; revised IPA of online hotel booking

Examines digital attributes and illustrates how the analytical specification influences improvement priorities.

[5]

699 rural-accommodation users; SERVQUAL gap assessment

Provides comparative evidence on the context dependence of service shortfalls.

[11]

1,487 guests at nine hotels; quality and performance analysis

Connects physical and performed service quality with satisfaction and loyalty.

[16]

355 online-booking users; structural modeling and importance–performance mapping

Links digital attributes with technology acceptance and managerial priorities.

[18]

701 reviews from two hotels; BERT-based classification

Uses unstructured guest evidence for service-quality diagnosis.

[24]

25,267 reviews of 56 hotels; machine-learning analysis

Identifies service dimensions that vary with customer group and setting.

[23]

Five-star hotel case; fuzzy-set qualitative comparative analysis

Examines service configurations associated with satisfaction rather than assuming independent additive effects.

Note: SERVQUAL = service quality scale; IPA = importance--performance analysis; BERT = Bidirectional Encoder Representations from Transformers.

3. Methods

3.1 Setting, Participants, and Survey Administration

The study was conducted at a three-star city hotel in the Alsancak area of Konak, İzmir, Türkiye. The property serves domestic and international guests and uses digital channels for reservations, information, and transactions. A quantitative cross-sectional design was used, with the individual guest as the unit of analysis.

Data were collected between November 2025 and March 2026 through an online questionnaire on forms.app. Guests accessed the questionnaire using a Quick Response (QR) code at the hotel and participated voluntarily. The questionnaire included corresponding expectation and perception sections covering the same 22 attributes. Convenience sampling was used; participation was not based on a random draw from the hotel’s guest population.

After incorrectly recorded responses were excluded during screening, 411 questionnaires were retained. Item comparisons used complete expectation–perception pairs, with valid paired samples ranging from 391 to 411. The effective sample size therefore varied by item. Domain summaries were calculated from item means as defined in Section 3.3; the limitations of the respondent-level correlation analysis are addressed in Section 5.5.

The sample-size formula for estimating a proportion was considered as a planning reference [25]. With $Z$ = 1.96, $p$ = 0.50, and an absolute error of 0.05, it gives 384.16, or 385 when rounded upward. This calculation neither makes a convenience sample representative nor establishes power for paired mean differences. The interpretation of this study rests on the observed sample and effect estimates, rather than a claimed population margin of error.

3.2 Instrument and Operational Domains

The instrument adapted the SERVQUAL expectation–perception structure [7], with hotel wording informed by the contextualized resort instrument [8]. Digital items drew on e-service research [4], [14], [16]. Each of the 22 attributes was evaluated twice on a five-point Likert scale, producing 44 ratings per fully completed questionnaire. Higher scores indicated greater expectations or more favorable perceived performance.

Five items were assigned to tangibles: physical appearance, room cleanliness, common-area cleanliness, quietness and sleep comfort, and food and beverage hygiene. Reliability comprised equipment functionality, room preparation according to the reservation, provision of basic services, and reservation-information accuracy. Responsiveness covered meeting requests, check-in and check-out speed, complaint resolution, and proactive identification of problems.

Assurance comprised employee professionalism, location, price–performance balance, and guest-density management. This hotel-specific grouping extends beyond the conventional SERVQUAL definition and is interpreted accordingly. Personalized attention was measured by one item corresponding to the empathy concept. It is treated as an observed indicator, with no separate internal-consistency coefficient.

The baseline digital-services domain comprised access, adequacy of security measures, personal-data protection, and transaction security. The label of Item 15, adequacy of security measures, does not identify a specifically digital context. The four-item domain is therefore retained as a baseline and compared with a three-item definition excluding Item 15. The comparison tests sensitivity to domain composition; it cannot determine how individual guests understood the broad item. Table 2 gives the baseline item assignments.

Table 2. Baseline item assignments and expectation--performance gaps
ItemService AttributeReporting Domain$\boldsymbol{E}$$\boldsymbol{P}$$\boldsymbol{P-E}$
1Modern and well-maintained physical appearanceTangibles4.9464.985+0.039
2Room cleanliness standardsTangibles4.9953.628-1.367
3Proper functioning of room equipmentReliability4.9473.924-1.023
4Cleanliness and orderliness of common areasTangibles4.9463.120-1.826
5Preparation of the room in accordance with the reservationReliability4.9904.983-0.007
6Meeting guest requestsResponsiveness4.9494.942-0.007
7Speed of check-in and check-out proceduresResponsiveness4.9494.985+0.036
8Prompt resolution of complaintsResponsiveness4.9464.993+0.046
9Professional behavior of employeesAssurance4.9104.949+0.039
10Personalized attention to guestsPersonalized attention4.8954.939+0.044
11Hotel location meeting expectationsAssurance4.9034.993+0.090
12Price-performance balanceAssurance4.9464.944-0.002
13Smooth provision of basic servicesReliability4.9294.995+0.066
14Accessibility of digital servicesDigital services4.9034.939+0.036
15Adequacy of security measuresDigital services4.7434.988+0.245
16Protection of personal dataDigital services4.9484.388-0.560
17Data security in digital transactionsDigital services4.9904.569-0.421
18Accurate management of reservation informationReliability4.9484.138-0.810
19Quietness and sleep comfortTangibles4.8874.805-0.083
20Management of guest densityAssurance4.9504.987+0.038
21Employees identifying problems in advanceResponsiveness4.9004.885-0.015
22Food and beverage hygieneTangibles4.9934.973-0.020
Note: $E$ and $P$ are the means for complete expectation--perception pairs; paired $n$ ranges from 391 to 411. Differences of 0.001 between a displayed gap and the subtraction of rounded means reflect rounding. Personalized attention is a single observed item. Item 15 belongs to the baseline digital domain and is excluded in the sensitivity calculation in Section 4.6.
3.3 Gap Scores and Aggregation

Let ${E}_{ij}$ and ${P}_{ij}$ denote the expectation and perception ratings of respondent $j$ for item $i$. Each item mean was calculated using its set $V_i$ of complete expectation–perception pairs, with $n_i=\left|V_i\right|$. The item gap was defined as:

$G_i=\overline{P}_i-\overline{E}_i=\frac{1}{n_i} \sum_{j \in V_i}\left(P_{i j}-E_{i j}\right)$
(1)

A negative gap indicates perceived delivery below expectations; a positive gap indicates delivery above expectations. The magnitude remains in units of the five-point response scale. A statistically detectable positive gap is not automatically a practically important advantage.

For domain $d$ with item set $A_d$ and $m_d$ items, the reporting means and gap were calculated as:

$P_d=\frac{1}{m_d} \sum_{i \in A_d} \overline{P}_i ; I_d=\frac{1}{m_d} \sum_{i \in A_d} \overline{E}_i ; G_d=P_d-I_d$
(2)

$I_d$ is an expectation-based importance proxy, not a separately elicited importance score. Domain means give equal weight to their constituent items. Because item-specific paired samples differ, these means are not necessarily identical to averaging respondent-level complete-case domain scores. The two quantities should not be interchanged when reproducing the results. Item gaps are retained throughout interpretation to prevent positive and negative attributes from canceling without explanation.

3.4 Reliability and Exploratory Associations

Internal consistency was assessed using ordinal alpha based on polychoric correlations. This coefficient concerns the latent responses underlying ordinal categories [26], [27]. Alpha must be interpreted in relation to the instrument’s structure and assumptions, rather than as a standalone test of validity [28], [29]. The overall expectation and perception coefficients were 0.986 and 0.815, respectively. They summarize the 22-item scales and do not establish the validity of six distinct domains.

Skewness and kurtosis described the response distributions. Pearson correlations summarized associations among respondent-level perception-domain scores, with $N$ = 411 specified for the correlation analysis. Spearman coefficients provided an ordinal comparison. These associations describe how ratings vary together and are not used to infer causal relationships between hotel processes.

3.5 Paired Inference and Multiplicity

Two-sided paired $t$-tests compared expectation and perception ratings under $H_0$: $E\left(P_{i j}-E_{i j}\right)$ = 0, with nominal significance assessed at 0.05. Figure 1 presents 95% confidence intervals for the paired mean gaps. Cohen’s $d_z$ expresses each gap relative to the standard deviation of the paired differences. Interpretation emphasizes the gap magnitude and confidence interval alongside significance, because concentrated responses and multiple comparisons can make a significance label an incomplete account of a service difference.

Familywise evidence was assessed conservatively from the $p$-value categories accompanying the item results. For 22 comparisons, the Bonferroni threshold is:

$\alpha_{\text {item }}=\frac{0.05}{22}=0.002273$
(3)

An item with $p <$ 0.001 meets this threshold. Categories of $p <$ 0.01 or $p <$0.05 do not establish an adjusted decision without the exact $p$-value. The calculation therefore identifies a lower bound on the number of familywise-significant items; it does not recover exact adjusted $p$-values.

The 15 unique correlations among six domains formed a separate exploratory family. Approximate two-sided $p$-values were calculated from the rounded Pearson coefficients and $N$ = 411 using the $t$ transformation with 409 degrees of freedom, and compared with 0.05/15 = 0.003333. These summary-based calculations assess the reported associations and do not constitute respondent-level re-estimation.

3.6 Baseline Importance–Performance Analysis and Classification Limits

Baseline IPA used the expectation mean $I_d$ as the vertical coordinate and the perception mean $P_d$ as the horizontal coordinate. Following the four-quadrant logic [19], the crosshairs were defined as the unweighted means of the six reporting-domain means:

$I_c=\frac{1}{6} \sum_{d=1}^6 I_d ; P_c=\frac{1}{6} \sum_{d=1}^6 P_d$
(4)

Domains were classified as “Concentrate here” when $I_d \geq I_c$ and $P_d

Mean crosshairs were used because the scale midpoint would place every domain in the high-expectation, high-performance quadrant. This is a relative comparison among six domains. It is not an absolute threshold for acceptable service. Moreover, the narrow spread of expectation means makes the vertical classification sensitive to small changes. Exact equality and proximity to a crosshair are therefore interpreted as boundary conditions rather than strong differences in managerial importance [21].

Figure 1. Paired expectation–performance gaps by service attribute
Note: Error bars show 95% confidence intervals for paired mean differences. Stars denote unadjusted tests: *$p <$ 0.05, **$p <$ 0.01, ***$p <$ 0.001. Bar labels are unstandardized gaps; standardized effects are given in the text.
3.7 Sensitivity Analysis and Operational Interpretation

Sensitivity was examined using the item and domain means. First, the digital-services domain was recalculated from Items 14, 16, and 17, excluding Item 15. The other five domain means were held fixed, and both crosshairs were recalculated. This comparison isolates the effect of a narrower definition of digital services.

Second, two alternative crosshair rules were examined: the median of the six domain means and an overall mean weighted by the baseline numbers of items per domain. The weighted rule gives each of the 22 items equal influence on the overall reference point, whereas baseline IPA gives equal influence to each reporting domain. These are descriptive specification checks, not sampling-confidence estimates.

Third, negative item gaps were ranked separately from domain quadrants. Attributes with substantial shortfalls were retained for managerial review even when the corresponding domain had a low-priority label. Intervention costs and implementation outcomes were outside the survey, so the recommendations concern processes to investigate and monitor.

4. Results

4.1 Sample Profile

Of the 411 respondents, 78.6% were women, 18.7% were men, and 2.7% selected another category or did not specify gender. Guests aged 18–24 accounted for 51.6%, and those aged 25–34 for 34.5%. Cultural travel accounted for 80.3% of responses, 92.2% were first-time guests, and 86.9% were from Türkiye. Appendix A presents the full percentage distributions (Table A1).

These concentrations are relevant to interpretation. The results primarily describe younger, domestic, first-time visitors traveling for cultural purposes. The voluntary sample cannot establish whether their priorities match those of business travelers, older guests, repeat visitors, or the hotel’s full guest population.

4.2 Reliability and Response Concentration

Overall ordinal alpha was 0.986 for expectations and 0.815 for perceptions (Table 3). The higher expectation coefficient must be interpreted alongside the concentration of responses near the scale maximum. These overall coefficients describe the 22-item scales; they do not demonstrate the internal consistency or factorial validity of the six reporting domains.

Table 3. Overall ordinal reliability
ScaleOrdinal AlphaItems
Expectations0.98622
Perceptions0.81522

Expectations were strongly concentrated near the upper limit. Skewness ranged from -14.28 to -1.11 for expectations and from -18.79 to 0.11 for perceptions; the corresponding kurtosis ranges were -0.76 to 203.21 and -1.34 to 365.05. Absolute skewness exceeded 2 for 21 expectation items and 17 perception items. These distributions indicate severe restriction in the variation of many ratings.

Domain expectation means ranged from 4.895 to 4.954, a spread of only 0.059 points. Such limited variation constrains the importance axis of the IPA. The paired-test results therefore need to be read alongside effect magnitudes and the sensitivity of the quadrant assignments; statistical significance does not restore the discrimination lost through near-ceiling responses.

4.3 Item Gaps and Paired Comparisons

At the nominal 0.05 level, expectation–perception differences were significant for 17 of the 22 items: seven gaps were negative and ten were positive. Common-area cleanliness had the largest shortfall (-1.826), followed by room cleanliness (-1.367), equipment functionality (-1.023), and reservation-information accuracy (-0.810). Their $d_z$ values were -1.25, -1.35, -0.86, and -0.58, respectively. These results place housekeeping, maintenance, and reservation handling at the center of the operational diagnosis.

Personal-data protection (-0.560; $d_z$ = -0.52) and transaction security (-0.421; $d_z$ = -0.41) also fell below expectations. These are guests’ assessments of protection, not evidence that a technical breach occurred. Quietness and sleep comfort showed a smaller negative gap (-0.083; $d_z$ = -0.12). Room preparation, meeting requests, price–performance balance, proactive problem identification, and food hygiene did not differ significantly at the nominal level.

Security measures had the largest positive gap (+0.245; $d_z$ = 0.50), although the item’s broad wording limits its interpretation. The other significant positive gaps were +0.036 to +0.090, with $d_z$ values of 0.13–0.25. For these attributes, perceived performance slightly exceeded expectations, although the differences were small. Figure 1 presents the item gaps and their confidence intervals.

4.4 Conservative Evidence After Accounting for Multiplicity

Twelve items had $p <$ 0.001 and therefore meet the Bonferroni threshold of 0.002273. Six were negative gaps—Items 2, 3, 4, 16, 17, and 18—and six were positive—Items 1, 8, 10, 11, 13, and 15. The six principal shortfalls therefore remain supported under this conservative familywise check.

Items 7, 9, 14, and 20 had $p <$ 0.01, while Item 19 had $p <$ 0.05. These thresholds alone do not establish whether the items meet the Bonferroni criterion. The familywise result is therefore expressed as a lower bound of 12 items, including all six principal negative gaps.

4.5 Exploratory Associations Among Reporting Domains

Responsiveness, assurance, and personalized attention were strongly associated: $r$ = 0.814 for responsiveness–assurance, 0.823 for responsiveness–attention, and 0.930 for assurance–attention. The Spearman coefficients showed a similar pattern, ranging from 0.77 to 0.95. In particular, the assurance–attention association suggests substantial overlap in these guest evaluations.

The remaining Pearson coefficients ranged from -0.031 to 0.124. Four weak positive associations met the nominal 0.05 criterion, but none met the threshold of 0.003333 for 15 comparisons. Only the three strong interpersonal associations met that conservative reference criterion (Figure 2). Restricted response variation and the item groupings should be considered when interpreting the weak associations.

Figure 2. Exploratory correlations among perception reporting domains
Note: The reported correlation sample size is $N$ = 411. An asterisk denotes approximate two-sided $p <$ 0.003333, calculated from the displayed three-decimal Pearson coefficients using the Bonferroni threshold for 15 comparisons. Personalized attention is a single item.
4.6 Baseline Priorities and Sensitivity to Analytical Choices

The baseline crosshairs were $I_c \approx$ 4.927 and $P_c \approx$ 4.732 (Table 4). Tangibles and reliability combined higher expectations with lower performance and fell in “Concentrate here” (Table 5). Responsiveness fell in the “Keep up the good work” quadrant. Assurance was close to the expectation crosshair and is best treated as a boundary case. Personalized attention combined lower relative expectations with high performance, while digital services fell below both crosshairs.

Table 4. Deterministic sensitivity of priority classifications
Specification$\boldsymbol{I}$ Crosshair$\boldsymbol{P}$ CrosshairDigital-Services PositionTangibles/Reliability
Baseline equal-domain means4.9274.732Low priorityBoth concentrate here
Baseline median crosshairs4.9324.830Low priorityBoth concentrate here
Baseline item-weighted crosshairs4.9324.684Possible overkillBoth concentrate here
Exclude Item 15; equal-domain means4.9354.717Concentrate hereBoth concentrate here
Note: $I$ denotes importance, proxied by guest expectation ratings, and $P$ denotes performance, measured by guest perception ratings. Calculations use rounded item and domain means. Only the final row changes the digital item set. These comparisons test analytical choices and do not estimate sampling uncertainty or quadrant-membership probabilities.
Table 5. Baseline domain means and relative IPA classifications
DomainExpectationPerceptionGapBaseline Interpretation
Tangibles4.9534.302-0.651Concentrate here
Reliability4.9544.510-0.444Concentrate here
Responsiveness4.9364.951+0.015Keep up the good work
Assurance4.9274.968+0.041Near the boundary between ``Keep up the good work'' and ``Possible overkill''
Personalized attention4.8954.939+0.044Possible overkill; no spending inference
Digital services4.8964.721-0.175Low relative priority; inspect item gaps
Note: IPA = importance--performance analysis.

Within digital services, positive gaps for access and general security offset shortfalls in data protection and transaction security. Excluding Item 15 gave $I_d \approx$ 4.947, $P_d \approx$ 4.632, and $G_d \approx$ -0.315. With recalculated crosshairs of $I_c \approx$ 4.935 and $P_c \approx$ 4.717, digital services moved to “Concentrate here”; tangibles and reliability remained there (Figure 3). The digital gap increased in magnitude by approximately 80%. This change shows how domain composition affects the priority diagnosis.

Figure 3. Baseline IPA and sensitivity to excluding the broad security-measures item
Note: IPA = importance–performance analysis. Blue circles show the baseline domain means. The orange diamond shows digital services computed from Items 14, 16, and 17. Dashed lines are baseline crosshairs; dotted lines are recalculated crosshairs for the narrower definition. The arrow denotes a change in specification, not change over time. The expectation axis is expanded to show its restricted range.

Alternative crosshairs produced a different pattern of stability. The median rule gave $I_c \approx$ 4.932 and $P_c \approx$ 4.830; weighting the six means by their baseline item counts gave $I_c \approx$ 4.932 and $P_c \approx$ 4.684. Tangibles and reliability remained in the improvement-priority quadrant under both rules. Digital services remained below the importance crosshair under both, although its performance classification changed under the item-weighted rule. Assurance moved below the importance crosshair. Tangibles and reliability remained improvement priorities across the specifications examined, whereas the digital-services quadrant depended on both item composition and crosshair placement.

4.7 Agreement Between Item and Domain Evidence

Figure 4 compares expectations and perceptions across the baseline domains. Tangibles and reliability had the largest negative domain gaps, contained the four largest item shortfalls, and remained improvement priorities across all four specifications. Agreement between the item and domain results supports their priority status.

The digital-services result differs. Its baseline gap was moderate and its quadrant indicated low relative priority, yet its data-protection and transaction items had shortfalls that survived the conservative multiplicity check. The narrower digital definition made these concerns more visible. Responsiveness, assurance, and personalized attention had small positive domain gaps, but these averages should be interpreted alongside the constituent item results.

Figure 4. Baseline expectation and perception means across reporting domains
Note: Gaps are perception minus expectation. The axis is restricted to the range of the domain means. Personalized attention is a single observed item.

5. Discussion

5.1 Substantive Findings in the Hotel Context

The largest expectation–performance shortfalls concerned housekeeping, equipment functionality, and reservation handling. Common-area and room cleanliness, equipment functionality, and reservation accuracy combined substantial gaps with evidence that survived the conservative multiplicity check. These results give managers specific processes to examine rather than a general instruction to raise service quality.

The prominence of physical conditions is consistent with research linking hotel facilities and tangible quality to satisfaction [9], [11]. Nevertheless, the ordering is local. Rural-accommodation research reported a different balance of shortfalls across conventional SERVQUAL dimensions [5], and research across hotel star categories found differences in performance relative to expectations [12]. Research on rural homestays also relates physical servicescape, perceived service quality, and staying intention [30], but these relationships do not establish the priorities of a city hotel. The present sample’s young, domestic, first-time profile provides a further reason not to transfer the ranking directly to other properties.

The small positive interpersonal gaps indicate that sampled guests rated request handling and staff interaction as meeting or slightly exceeding their expectations. However, the assurance–attention correlation was high, personalized attention had one item, and the assurance grouping included heterogeneous attributes. These findings are more useful for monitoring local service performance than for distinguishing separate psychological constructs.

5.2 Digital Service Aggregation and Priority Reversals

The digital priority reversal is the main analytical finding. General security received a positive gap and a lower expectation score than the other digital items. Including it therefore raised the domain’s performance and lowered its expectation-based importance. Once it was excluded, the two data-related shortfalls became more prominent in the aggregate result. Relying on the baseline quadrant alone would have understated their operational relevance.

This finding is compatible with the multidimensional view of online service in booking research, where functionality, information, and privacy are not interchangeable [4], [16]. Smart service-quality research also distinguishes the appeal of technology from the importance assigned to it [3]. The comparison shows that favorable ratings of general security can mask lower ratings of personal-data protection and digital transaction security within an aggregated domain.

Perceived weakness must be distinguished from technical failure. Guests may respond to unclear privacy communication, unfamiliar payment procedures, actual service difficulties, or several of these at once. The survey cannot identify the mechanism. The findings support a targeted review of data handling and transaction procedures, but do not establish a security breach or determine the scope of technical changes required. Follow-up assessment should connect the guest ratings with records of the processes concerned.

5.3 Contribution to Operational Decision Support

The analysis connects the size of service shortfalls with the consistency of the priorities assigned to them. Effect sizes and the multiplicity check qualify the paired-test results; IPA summarizes the domain comparisons; sensitivity calculations identify dependence on item definitions and reference rules. Together, they distinguish persistent priorities from classifications requiring further scrutiny.

The findings extend the practical concerns raised in methodological work on IPA [21]. Studies relating gap scores to travel intentions [22] and service configurations to satisfaction [23] test consequences beyond those measured here. The present contribution is narrower: it shows, within one hotel, which process priorities persist and how aggregation can obscure a specific service weakness.

Structured surveys can be complemented by guest reviews. Review-based studies identify context-specific dimensions [24], and RbQUAL automates their classification [18]. Comparing these sources could establish whether cleaning, equipment, reservation, and data concerns recur among guests who did not complete the survey. Such triangulation would provide an independent check on the local diagnosis.

5.4 Managerial Actions and Monitoring

Housekeeping should receive the first operational review because room and common-area cleanliness have the two largest negative gaps. Managers can examine cleaning schedules, completion records, inspection practices, and recurrence of complaints. Room-equipment failures should be reviewed through maintenance records and pre-arrival checks. These actions respond directly to the observed attributes without presuming that more expenditure alone will resolve the problem.

Reservation accuracy requires tracing booking information from the digital channel to the front desk and prepared room. The aim is to identify mismatches in room type, dates, requested arrangements, or communication. The data-related findings call for a separate assessment of how guests receive privacy information and how transactions are completed. Any technical assessment should be conducted by personnel with access to the actual systems; guest ratings alone do not establish system security.

Interpersonal standards should be maintained while corrective actions are directed to the main shortfalls. The “Possible overkill” label for personalized attention does not establish excess staffing or care. Table 6 sets out proposed actions and follow-up measures; their operational effects remain to be evaluated.

Table 6. Proposed operational actions and follow-up indicators
ProcessEvidence From the SurveyProposed ActionFollow-up Measure
HousekeepingCommon-area gap -1.826; room gap -1.367Review schedules, completion records, and inspection consistencyCleaning-related complaints per occupied room-night; inspection pass rate
MaintenanceEquipment gap -1.023Check preventive maintenance and pre-arrival functionality inspectionsEquipment faults per occupied room-night; repeat failures
ReservationsInformation-accuracy gap -0.810Trace booking details across systems and handoversBooking discrepancies per completed stay; resolution time
Data handling and transactionsProtection gap -0.560; transaction gap -0.421Review procedures, privacy information, and transaction supportTransaction failures per attempt; data-handling complaints; verified review findings
Staff interactionSmall positive domain gaps; highly associated ratingsMaintain service standards and monitor effects of other changesRequest-resolution time; complaints about assistance; repeat survey ratings
5.5 Limitations and Further Validation

The study concerns one property and a voluntary convenience sample dominated by younger, domestic, first-time visitors. These features limit generalization. Expectations and perceptions were collected in the same questionnaire, so the gaps may reflect retrospective judgments and shared measurement conditions. They should not be interpreted as changes measured between separate pre-stay and post-stay waves.

The adapted domains also constrain interpretation. Assurance combines heterogeneous attributes, personalized attention has one item, and general security is not explicitly digital. Overall ordinal alpha does not validate these groupings. Near-ceiling expectations and sparse lower response categories restrict discrimination and may affect polychoric estimation. The domains are consequently used for local operational reporting, without claiming a validated six-factor measurement structure.

Inference is limited by the use of significance categories and rounded correlation coefficients in the summary-based checks. Exact adjusted item $p$-values cannot be recovered from the categories. In addition, the respondent-level missing-response rule underlying the reported $N$ = 411 correlations is unspecified, so those associations remain exploratory. The main priority comparisons rely on the transparent aggregation of item means, rather than on the correlation results.

The specification comparisons assess dependence on analytical choices, not sampling uncertainty. Respondent-level resampling, directly elicited importance ratings, and replication across hotel categories would provide stronger evidence for priority stability. Prospective evaluation should link corrective actions to service indicators and costs, since the survey contains no intervention or post-implementation outcomes.

6. Conclusions

The SERVQUAL–IPA assessment identified six principal perceived shortfalls at the studied hotel: common-area cleanliness, room cleanliness, equipment functionality, reservation accuracy, personal-data protection, and transaction security. All six had $p <$ 0.001 and met the conservative Bonferroni criterion for 22 comparisons.

Tangibles and reliability remained improvement priorities across the analytical specifications examined. The digital-services classification changed when general security was excluded, showing that an aggregate quadrant can conceal data-related concerns. Item-level evidence is therefore essential when interpreting the priority matrix.

For this hotel, the findings support focused reviews of housekeeping, maintenance, reservation handling, and guest-facing data processes while maintaining interpersonal service. Repeated measurement can then test whether corrective actions reduce the identified shortfalls and whether the priority ordering changes.

Author Contributions

Conceptualization, S.K.K., S.H., and Ç.K.; methodology, S.K.K., S.H., and Ç.K.; software, S.K.K., S.H., and Ç.K.; validation, S.K.K., S.H., and Ç.K.; formal analysis, S.K.K., S.H., and Ç.K.; investigation, S.K.K., S.H., and Ç.K.; resources, S.K.K., S.H., and Ç.K.; data curation, S.K.K., S.H., and Ç.K.; writing—original draft preparation, S.K.K., S.H., and Ç.K.; writing—review and editing, S.K.K., S.H., and Ç.K.; visualization, S.K.K., S.H., and Ç.K.; supervision, S.K.K., S.H., and Ç.K. All authors have read and agreed to the published version of the manuscript.

Data Availability

The data supporting the findings of this study are available from the corresponding author upon reasonable request.

Conflicts of Interest

The authors declare no conflicts of interest.

Declaration on the Use of Generative AI and AI-assisted Technologies

Claude (Anthropic) assisted with statistical analyses of the raw survey data, figure preparation, and language editing. The authors reviewed and verified the outputs and take full responsibility for the content of the publication.

References
1.
E. Wąsowicz-Zaborek, “Content analysis of hotel reviews as a quality management tool: Preliminary verification of the SERVQUAL scale,” Turyzm/Tourism, vol. 33, no. 1, pp. 29–40, 2023. [Google Scholar] [Crossref]
2.
A. Parasuraman, V. A. Zeithaml, and L. L. Berry, “A conceptual model of service quality and its implications for future research,” J. Mark., vol. 49, no. 4, pp. 41–50, 1985. [Google Scholar] [Crossref]
3.
N. Ayvaz-Çavdaroğlu, S. Iyanna, and M. Foster, “Smart service quality in hospitality–A quantitative assessment using MCDM and clustering methods,” Int. J. Hosp. Manag., vol. 123, p. 103931, 2024. [Google Scholar] [Crossref]
4.
R.-C. Jou and Y.-J. Day, “Application of revised importance–performance analysis to investigate critical service quality of hotel online booking,” Sustainability, vol. 13, no. 4, p. 2043, 2021. [Google Scholar] [Crossref]
5.
J. Wojciechowska-Solis, M. Śmiglak Krajewska, and D. Viti, “Identification of the quality gaps in the services offered in accommodation facilities in rural areas: The case of the Lublin region,” Sustainability, vol. 14, no. 24, p. 16873, 2022. [Google Scholar] [Crossref]
6.
M. Kozłowski and J. Korzeniewski, “Studying the effectiveness of investments in hotel services using customer sentiment analysis,” Qual. Quant., vol. 60, no. 3, pp. 8013–8041, 2026. [Google Scholar] [Crossref]
7.
A. Parasuraman, V. A. Zeithaml, and L. L. Berry, “SERVQUAL: A multiple-item scale for measuring consumer perceptions of service quality,” J. Retail., vol. 64, no. 1, pp. 12–40, 1988. [Google Scholar]
8.
Y. Adhikari, M. P. Wagley, and N. Dahal, “Developing the contextualized SERVQUAL instrument for measuring the service quality of Nepali resorts: An application of the modified Delphi technique,” SAGE Open, vol. 13, no. 3, p. 21582440231195952, 2023. [Google Scholar] [Crossref]
9.
A. Baquero, “Is customer satisfaction achieved only with good hotel facilities? A moderated mediation model,” Adm. Sci., vol. 13, no. 4, p. 108, 2023. [Google Scholar] [Crossref]
10.
M. A. I. Gazi, M. A. Islam, A. A. Masud, A. R. B. S. Senathirajah, S. Biswas, and R. A. Shuvro, “The moderating impacts of COVID-19 fear on hotel service quality and tourist satisfaction: Evidence from a developing country,” Cogent Soc. Sci., vol. 10, no. 1, p. 2331079, 2024. [Google Scholar] [Crossref]
11.
M. H. Asgeirsson, T. Gudlaugsson, and G. T. Jóhannesson, “The relationships between service quality, reputation, and performance in hospitality,” Tour. Hosp., vol. 5, no. 3, pp. 736–752, 2024. [Google Scholar] [Crossref]
12.
A. Mohsin, H. Rodrigues, and A. Brochado, “Shine bright like a star: Hotel performance and guests’ expectations based on star ratings,” Int. J. Hosp. Manag., vol. 83, pp. 103–114, 2019. [Google Scholar] [Crossref]
13.
D. Buhalis and R. Law, “Progress in information technology and tourism management: 20 years on and 10 years after the Internet—The state of eTourism research,” Tour. Manag., vol. 29, no. 4, pp. 609–623, 2008. [Google Scholar] [Crossref]
14.
A. F. Alnaim, A. E. E. Sobaih, and I. A. Elshaer, “Measuring the mediating roles of E-Trust and E-Satisfaction in the relationship between E-Service Quality and E-Loyalty: A structural modeling approach,” Mathematics, vol. 10, no. 13, p. 2328, 2022. [Google Scholar] [Crossref]
15.
N. Chotisarn and T. Phuthong, “Impact of artificial intelligence-enabled service attributes on customer satisfaction and loyalty in chain hotels: Evidence from coastal tourism destinations in western Thailand,” Soc. Sci. Humanit. Open, vol. 11, p. 101306, 2025. [Google Scholar] [Crossref]
16.
L. Chen, C. F. Goh, Y. M. Low, O. K. Tan, and K. Y. Lim, “From service to intention: A Technology Acceptance Model perspective on e-service quality in online hotel booking,” SAGE Open, vol. 15, no. 4, p. 21582440251397834, 2025. [Google Scholar] [Crossref]
17.
O. S. Olawuyi and C. Kleynhans, “A bibliometric analysis of service quality in the hospitality industry (2014–2024),” Adm. Sci., vol. 15, no. 6, p. 215, 2025. [Google Scholar] [Crossref]
18.
V. Perdomo-Verdecia, M. Sacristán-Díaz, and P. Garrido-Vega, “RbQUAL: Integrating SERVQUAL and BERT to analyse service quality dimensions in hotel online reviews,” J. Hosp. Tour. Insights, vol. 9, no. 11, pp. 218–237, 2026. [Google Scholar] [Crossref]
19.
J. A. Martilla and J. C. James, “Importance-performance analysis,” J. Mark., vol. 41, no. 1, pp. 77–79, 1977. [Google Scholar] [Crossref]
20.
P. R. Lingadkar and K. G. Sankaranarayanan, “Hospitality and tourism internship programme: Expectation versus reality,” Afr. J. Hosp. Tour. Leis., vol. 12, no. 3, pp. 911–926, 2023. [Google Scholar] [Crossref]
21.
I. K. W. Lai and M. Hitchcock, “Importance–performance analysis in tourism: A framework for researchers,” Tour. Manag., vol. 48, pp. 242–267, 2015. [Google Scholar] [Crossref]
22.
B. B. Boley and E. Jordan, “Leveraging IPA gap scores to predict intent to travel,” J. Hosp. Tour. Manag., vol. 57, pp. 97–101, 2023. [Google Scholar] [Crossref]
23.
V. Perdomo-Verdecia, P. Garrido-Vega, and M. Sacristán-Díaz, “An fsQCA analysis of service quality for hotel customer satisfaction,” Int. J. Hosp. Manag., vol. 122, p. 103793, 2024. [Google Scholar] [Crossref]
24.
K. Kalnaovakul and P. Promsivapallop, “Hotel service quality dimensions and attributes: An analysis of online hotel customer reviews,” Tour. Hosp. Res., vol. 23, no. 3, pp. 420–440, 2023. [Google Scholar] [Crossref]
25.
J. Charan and T. Biswas, “How to calculate sample size for different study designs in medical research?,” Indian J. Psychol. Med., vol. 35, no. 2, pp. 121–126, 2013. [Google Scholar] [Crossref]
26.
A. M. Gadermann, M. Guhn, and B. D. Zumbo, “Estimating ordinal reliability for Likert-type and ordinal item response data: A conceptual, empirical, and practical guide,” Pract. Assess. Res. Eval., vol. 17, no. 3, pp. 1–13, 2012. [Google Scholar] [Crossref]
27.
B. D. Zumbo, A. M. Gadermann, and C. Zeisser, “Ordinal versions of coefficients alpha and theta for Likert rating scales,” J. Mod. Appl. Stat. Methods, vol. 6, no. 1, pp. 21–29, 2007. [Google Scholar] [Crossref]
28.
Y. F. Zakariya, “Cronbach’s alpha in mathematics education research: Its appropriateness, overuse, and alternatives in estimating scale reliability,” Front. Psychol., vol. 13, p. 1074430, 2022. [Google Scholar] [Crossref]
29.
K. S. Taber, “The use of Cronbach’s alpha when developing and reporting research instruments in science education,” Res. Sci. Educ., vol. 48, pp. 1273–1296, 2018. [Google Scholar] [Crossref]
30.
Y. Yang, L. Wang, and J. Yan, “Perceived service quality mediates the effect of physical servicescape on tourists’ intention to stay in rural homestays,” Sci. Rep., vol. 15, p. 29506, 2025. [Google Scholar] [Crossref]
Appendix

Appendix A. Sample Characteristics

Table A1. Reported participant distributions

Characteristic

Category

Percentage

Gender

Women

78.6

Gender

Men

18.7

Gender

Other/not specified

2.7

Age

18–24

51.6

Age

25–34

34.5

Age

35–44

8.5

Age

45–54

0.5

Age

55–64

2.7

Age

65 or older

2.2

Travel purpose

Cultural

80.3

Travel purpose

Family visit

8.5

Travel purpose

Business

7.5

Travel purpose

Other

3.4

Travel purpose

Education

0.2

Previous stay

First-time guest

92.2

Previous stay

Repeat guest

7.8

Origin

Türkiye

86.9

Origin

Other European countries

5.8

Origin

Americas

4.1

Origin

Asia

2.4

Origin

Africa

0.7

Note: N = 411 retained questionnaires. Percentages may not sum to 100 because of rounding.


Cite this:
APA Style
IEEE Style
BibTex Style
MLA Style
Chicago Style
GB-T-7714-2015
Kaya, S. K., Hayva, S., & Karamaşa, Ç. (2026). Prioritizing Hotel Service Improvements Using SERVQUAL and Sensitivity Analysis of Importance–Performance Classifications. J. Oper. Strateg Anal., 4(4), 254-267. https://doi.org/10.56578/josa040403
S. K. Kaya, S. Hayva, and Ç. Karamaşa, "Prioritizing Hotel Service Improvements Using SERVQUAL and Sensitivity Analysis of Importance–Performance Classifications," J. Oper. Strateg Anal., vol. 4, no. 4, pp. 254-267, 2026. https://doi.org/10.56578/josa040403
@research-article{Kaya2026PrioritizingHS,
title={Prioritizing Hotel Service Improvements Using SERVQUAL and Sensitivity Analysis of Importance–Performance Classifications},
author={Sema KayapıNar Kaya and SıLa Hayva and çAğLar KaramaşA},
journal={Journal of Operational and Strategic Analytics},
year={2026},
page={254-267},
doi={https://doi.org/10.56578/josa040403}
}
Sema KayapıNar Kaya, et al. "Prioritizing Hotel Service Improvements Using SERVQUAL and Sensitivity Analysis of Importance–Performance Classifications." Journal of Operational and Strategic Analytics, v 4, pp 254-267. doi: https://doi.org/10.56578/josa040403
Sema KayapıNar Kaya, SıLa Hayva and çAğLar KaramaşA. "Prioritizing Hotel Service Improvements Using SERVQUAL and Sensitivity Analysis of Importance–Performance Classifications." Journal of Operational and Strategic Analytics, 4, (2026): 254-267. doi: https://doi.org/10.56578/josa040403
KAYA S K, HAYVA S, KARAMAŞA Ç. Prioritizing Hotel Service Improvements Using SERVQUAL and Sensitivity Analysis of Importance–Performance Classifications[J]. Journal of Operational and Strategic Analytics, 2026, 4(4): 254-267. https://doi.org/10.56578/josa040403
cc
©2026 by the author(s). Published by Acadlore Publishing Services Limited, Hong Kong. This article is available for free download and can be reused and cited, provided that the original published version is credited, under the CC BY 4.0 license.