Retail Service Quality Assessment Using Interval-Valued Pythagorean Fuzzy Approach ^†

Abstract

In this study, a service quality (SQ) quantitative assessment method for retail industry operations was established to transform the perceptions and expectations of consumers into measurement framework requirements in the Indian market. The benefit of the recently created framework using interval-valued Pythagorean fuzzy to analyze the SQ is to handle imprecise human assessments, which is lacking in traditional SQ assessment methods. Therefore, we proposed a two-stage SQ assessment method by applying an extended novel Fuzzy methodology. We conducted a systematic literature review to identify the service quality and its products in the framework of the retail industry. In addition, we identified the gaps and proposed the measurement framework through consumer expectations and perceptions gaps. The present research findings confirmed that reliability and tangibility are important SQ dimensions reflecting the customer’s opinions. The findings help retail businesses make collective decisions on high-priority areas and effectively allocate resources to meet the needs of their customers.

1. Introduction

The retail industry is the cornerstone of the economic structure today. Throughout the pandemic, this industry has provided a wide range of services to people. Retail supermarkets and large department stores provide a wide range of products and services, offering consumers flexibility with a rapid growth rate. For instance, India’s food and retail industry ranks fifth globally, with retail sales accounting for 60% of total sales according to the Ministry of Food Processing Industries. It makes up the majority of India’s retail industry and has generated more than USD 10 billion by 2022. In India, the total number of retail supermarkets is increasing significantly. The rapid increase suggests that consumers are more likely to shop at retail supermarkets than before. Retail supermarkets are facing intense competition due to new competitors. Retailers are now driven to expand their offerings beyond what they traditionally have provided.

Najib and Sosianika [1] defined service quality (SQ) as the perceived evaluation using an assessment procedure in which consumers compare the services with expectations. The expected and perceived quality of service are important to meet the customer’s expectations, which are influenced by their experiences with the service, while their perception of the service affects the perceived quality [2]. Supermarkets are one of the fastest-growing retail industries. An excellent SQ is related to high customer loyalty and the service elements they deliver. Despite this, today’s consumers want to be rewarded, making their attitudes and expectations of businesses open to change.

Scholars have evaluated and proposed various SQ models and methods for measuring SQ. In this research, a multiple-perspective technique was utilized to assess customer satisfaction and the impact of customer perceptions in the retail industry. Consumers are important stakeholders because they allow firms to create revenue. Consumers influence the trustworthy assessment of a retail supermarket. As a result, customer satisfaction is measured using the SQ provided to consumers. Zhang et al. [3] suggested that increasing the SQ is critical to maintaining a competitive position in India. To maintain an active competitive edge in the competing market, supermarkets must follow global trends and adopt advanced technology while concentrating on customers’ interests and desires by offering better retail services. Researchers have examined the SQ dimensions in developed countries, but the perceptions of the SQ of retail consumers in India remain ignored [4]. Since the earlier studies were mainly conducted in countries with the more established retail industry, applying the same techniques in India without modification is inappropriate. Therefore, understanding the retail industry in India and appropriate resource distribution is critical for optimal assessment.

This study aims to illustrate the notion of assessing retail supermarkets’ consumers’ overall impression of SQ and identifying the SQ dimensions of retail supermarkets from the consumers’ perspective. In this study, for an analysis of the retail supermarket chain, we determined the SQ dimensions of supermarkets, a correlation between the influencing factors of the SQ, and resource-allocating methods to each SQ dimension using the interval-valued Pythagorean fuzzy (IVPF).

2. Literature Review

Retail supermarkets are increasing in India, being propelled by the public’s cooperation with retailers. The present market structure is shaped and governed by marketing concepts and consumers [1]. Retail merchants want minimal challenges in India. Since this hinders organized retail expansion, more retail supermarkets have difficulty with technological and infrastructure improvements. To build and sustain retail businesses, a value must be appreciated by their consumers. However, it is difficult to balance the supply chain network and the use of information technology. The demand can be met with structured stores and developed markets. Therefore, the retail supermarket business must prioritize cutting-edge technology and global trends to meet customer needs and expectations by providing superior services to maintain a competitive edge [5].

India is an emerging market with the potential for rapid economic growth and development. However, the Indian retail supermarkets and the Indians’ view of SQ have been paid very little attention to. As India is one of the top emerging marketplaces, it is crucial to determine the SQ influencing retail supermarkets and consumer loyalty.

The SQ construct has not been considered in previous studies to explore the difference between perceived service performance and expectations, or perceived performance. SQ dimensions have not been identified yet in detail. The SQ dimensions are multidimensional constructs that are used to assess perceived the SQ. The SQ is vital in the service industry to maintain a leading position in the market and is an indicator of company performance. Based on outstanding service, retailers can compete with structured and larger companies in terms of pricing. Furthermore, emphasizing SQ is crucial in retail supermarkets.

The SQ dimension identification is important for SQ enhancement. There is little empirical research on the elements for quality improvement and metrics to evaluate the SQ in the service industry. Li et al. [6] proposed the most well-known and well-discussed SQ model. Xue [7] proposed the retail service quality scale (RSQS) with a multidimensional architecture at three levels: an overall perception of service, a dimension level, and a sub-dimension level.

We examined the changing needs of Indian consumers when choosing retail stores while foreign large-scale retailers predominate the market. Challenges and issues for retail businesses were also explored in terms of how to maintain customer loyalty in Indian retail supermarkets. SQ dimensions to preserve client loyalty were identified to increase consumer loyalty. A total of twenty-two SQ dimensions were determined through the literature review, as shown in

Table 1. Quality dimensions identified in this research.

Source	Dimension	Questionnaire Items
Li et al. [6], Su and Teng [8]	Tangibility	T1: Modernized equipment.
		T2: The facilities are very appealing.
		T3: Employee clothes are appropriate to be acceptable.
		T4: The atmosphere of the supermarket is clean.
James [9], Su and Teng [8]	Reliability	R1: Satisfy the promise that was made to the consumer.
		R2: When a consumer or client is experiencing difficulty, resolve the issue.
		R3: Retailers provide high-quality service for the first time.
		R4: Meet the pledge by delivering services on time.
		R5: Documents are appropriately maintained in the retail’s database.
Gibson et al. [10], Su and Teng [8]	Responsiveness	RE1: Workers produce information that consumers can readily obtain.
		RE2: Employees accurately serve consumers.
		RE3: Consumer receive regular support from personnel.
		RE4: Workers are available to assist with customer requirements.
Ahn and Chung [11], Su and Teng (2018)	Assurance	A1: Consumer’ confidence is influenced by the employees behavior.
		A2: Consumer have faith in their dealings.
		A3: Consumer should be treated with respect.
		A4: The ability to respond to customer inquiries.
Liang et al. [12], Su and Teng [8]	Empathy	E1: Each consumer received individualized attention.
		E2: The retailer’s hours of operation are convenient for consumers.
		E3: Retailers provide customer service in addition to product sales.
		E4: Retailers are concerned about their consumers needs.
		E5: Retailers understand the specific needs of their consumers.

.

3. Methodology

3.1. Data Collection

A cross-sectional design was employed to gather data from consumers of retail supermarkets. We selected consumers who purchased any goods more than five times as participants in this research. We gathered data from Indian supermarkets. The valid number of responses was 593. To validate the questionnaire survey, three store managers and four professors reviewed the responses.

3.2. Analysis Method

Equation (1) was used to determine the IVPF [13].

$$\tilde{\mathrm{P}}=\left(\left[{\mu }_{\tilde{\mathrm{P}}}^{-}\left(x\right),{\mu }_{\tilde{\mathrm{P}}}^{+}\left(x\right)\right],\left[{\upsilon }_{\tilde{\mathrm{P}}}^{-}\left(x\right),{\upsilon }_{\tilde{\mathrm{P}}}^{+}\left(x\right)\right]\right)|x\in X$$

(1)

The function ${\mu }_{\tilde{\mathrm{P}}}^{-}\left(x\right),{\mu }_{\tilde{\mathrm{P}}}^{+}\left(x\right),{\upsilon }_{\tilde{\mathrm{P}}}^{-}\left(x\right),{\upsilon }_{\tilde{\mathrm{P}}}^{+}\left(x\right):X\to [\mathrm{0,1}]$ is for the lower-degree membership, while ${\upsilon }_{\tilde{\mathrm{P}}}^{+}\left(x\right) \mathrm{a}\mathrm{n}\mathrm{d} {\upsilon }_{\tilde{\mathrm{P}}}^{-}\left(x\right)$ is for the upper-degree membership. The non-membership function is defined as ${\upsilon }_{\tilde{\mathrm{P}}}^{+}\left(x\right) \mathrm{a}\mathrm{n}\mathrm{d} {\upsilon }_{\tilde{\mathrm{P}}}^{-}\left(x\right)$ for the upper and lower degrees. The set $\tilde{\mathrm{P}}$ must exceed the condition $0\le \left({\mu }_{\tilde{\mathrm{P}}}^{+}\left(x\right)+{\upsilon }_{\tilde{\mathrm{P}}}^{+}\left(x\right)\right)\le 1$. The indeterminacy upper and lower measures is calculated by using Equation (2).

$${\pi }_{\tilde{\mathrm{P}}}^{-}\left(x\right)=\sqrt{1-{\mu }_{\tilde{\mathrm{P}}}^{+}\left(x\right)-{\upsilon }_{\tilde{\mathrm{P}}}^{+}\left(x\right)}\mathrm{and} {\pi }_{\tilde{\mathrm{P}}}^{+}\left(x\right)=\sqrt{1-{\mu }_{\tilde{\mathrm{P}}}^{-}\left(x\right)-{\upsilon }_{\tilde{\mathrm{P}}}^{-}\left(x\right)}$$

(2)

3.3. IVPF

The different operations of two IVPF numbers are estimated by using Equations (3)–(6).

$${\tilde{\mathrm{P}}}_1{\oplus \tilde{\mathrm{P}}}_2=\left(\left[\sqrt{{\left({\mu }_1^{-}\right)}^2+{\left({\mu }_2^{-}\right)}^2-{\mu }_1^{-}\times {\mu }_2^{-}},\sqrt{{\left({\mu }_1^{+}\right)}^2+{\left({\mu }_2^{+}\right)}^2-{\mu }_1^{+}\times {\mu }_2^{+}}\right],[{\upsilon }_1^{-}\times {\upsilon }_2^{-},\hspace{1em}{\upsilon }_1^{+}\times {\upsilon }_2^{+}]\right)$$

(3)

$${\tilde{\mathrm{P}}}_1{\otimes \tilde{\mathrm{P}}}_2=\left(\left[{\mu }_1^{-}\times {\mu }_2^{-},\hspace{1em}{\mu }_2^{+}\times {\mu }_2^{+}\right],\left[\sqrt{{\left({\upsilon }_1^{-}\right)}^2+{\left({\upsilon }_2^{-}\right)}^2-{\upsilon }_1^{-}\times {\upsilon }_2^{-}},\sqrt{\sqrt{{\left({\upsilon }_1^{+}\right)}^2+{\left({\upsilon }_2^{+}\right)}^2-{\upsilon }_1^{+}\times {\upsilon }_2^{+}}}\right]\right)$$

(4)

$$\mathrm{k}\tilde{\mathrm{P}}=\left(\left[\sqrt{1-{\left(1-{\left({\mu }^{-}\right)}^2\right)}^{\mathrm{k}}},\sqrt{1-{\left(1-{\left({\mu }^{+}\right)}^2\right)}^{\mathrm{k}}}\right],\left[{\left({\upsilon }^{-}\right)}^{\mathrm{k}},{\left({\upsilon }^{+}\right)}^{\mathrm{k}}\right]\right), \mathrm{where} \mathrm{k} > 0$$

(5)

$${\tilde{\mathrm{P}}}^{\mathrm{k}}=\left(\left[{\left({\mu }^{-}\right)}^{\mathrm{k}},{\left({\mu }^{+}\right)}^{\mathrm{k}}\right],\left[\sqrt{1-{\left(1-{\left({\upsilon }^{-}\right)}^2\right)}^{\mathrm{k}}},\sqrt{1-{\left(1-{\left({\upsilon }^{+}\right)}^2\right)}^{\mathrm{k}}}\right]\right), \mathrm{where} \mathrm{k} > 0$$

(6)

3.4. Defuzzified Average Scores

The defuzzified mean ratings are derived using the anticipated IVPF average scores for every variable (for both the SQ’s expectation and perception). Let $\tilde{\mathrm{M}}\mathrm{e}\mathrm{j}=\left(\left[{\mu }_{\tilde{\mathrm{M}}\mathrm{e}\mathrm{j}}^{-},{\mu }_{\tilde{\mathrm{M}}\mathrm{e}\mathrm{j}}^{+}\right],\left[{\upsilon }_{\tilde{\mathrm{M}}\mathrm{e}\mathrm{j}}^{-},{\upsilon }_{\tilde{\mathrm{M}}\mathrm{e}\mathrm{j}}^{+}\right]\right)$ and $\tilde{\mathrm{M}}\mathrm{p}\mathrm{j}=\left(\left[{\mu }_{\tilde{\mathrm{M}}\mathrm{p}\mathrm{j}}^{-},{\mu }_{\tilde{\mathrm{M}}\mathrm{p}\mathrm{j}}^{+}\right],\left[{\upsilon }_{\tilde{\mathrm{M}}\mathrm{p}\mathrm{j}}^{-},{\upsilon }_{\tilde{\mathrm{M}}\mathrm{p}\mathrm{j}}^{+}\right]\right)$ be the IVPF average of the jth SQ expectation and perception. Then, the defuzzified average/mean expectation $\mathrm{D}\mathrm{e}\mathrm{j}$ and perception $\mathrm{D}\mathrm{p}\mathrm{j}$ of jth SQ item are determined by using Equations (7) and (8) [2,14,15].

$$\mathrm{D}\mathrm{p}\mathrm{j}={\displaystyle \frac{\left(\begin{array}{c}{{\mu }_{\tilde{\mathrm{M}}\mathrm{p}\mathrm{j}}^{-}}^2+{{\mu }_{\tilde{\mathrm{M}}\mathrm{p}\mathrm{j}}^{+}}^2+\left(1-{{\pi }_{\tilde{\mathrm{M}}\mathrm{p}\mathrm{j}}^{-}}^4-{{\upsilon }_{\tilde{\mathrm{M}}\mathrm{p}\mathrm{j}}^{-}}^2\right)+\left(1-{{\pi }_{\tilde{\mathrm{M}}\mathrm{p}\mathrm{j}}^{+}}^4-{{\upsilon }_{\tilde{\mathrm{M}}\mathrm{p}\mathrm{j}}^{+}}^2\right)+\\ {\mu }_{\tilde{\mathrm{M}}\mathrm{p}\mathrm{j}}^{-}{\mu }_{\tilde{\mathrm{M}}\mathrm{p}\mathrm{j}}^{+}+\sqrt[\frac{1}{4}]{\left(1-{{\pi }_{\tilde{\mathrm{M}}\mathrm{p}\mathrm{j}}^{-}}^4-{{\upsilon }_{\tilde{\mathrm{M}}\mathrm{p}\mathrm{j}}^{-}}^2\right)\left(1-{{\pi }_{\tilde{\mathrm{M}}\mathrm{p}\mathrm{j}}^{+}}^4-{{\upsilon }_{\tilde{\mathrm{M}}\mathrm{p}\mathrm{j}}^{+}}^2\right)}\end{array}\right)}{6}}$$

(7)

$$\mathrm{D}\mathrm{e}\mathrm{j}={\displaystyle \frac{\left(\begin{array}{c}{{\mu }_{\tilde{\mathrm{M}}\mathrm{e}\mathrm{j}}^{-}}^2+{{\mu }_{\tilde{\mathrm{M}}\mathrm{e}\mathrm{j}}^{+}}^2+\left(1-{{\pi }_{\tilde{\mathrm{M}}\mathrm{e}\mathrm{j}}^{-}}^4-{{\upsilon }_{\tilde{\mathrm{M}}\mathrm{e}\mathrm{j}}^{-}}^2\right)+\left(1-{{\pi }_{\tilde{\mathrm{M}}\mathrm{e}\mathrm{j}}^{+}}^4-{{\upsilon }_{\tilde{\mathrm{M}}\mathrm{e}\mathrm{j}}^{+}}^2\right)+\\ {\mu }_{\tilde{\mathrm{M}}\mathrm{e}\mathrm{j}}^{-}{\mu }_{\tilde{\mathrm{M}}\mathrm{e}\mathrm{j}}^{+}+\sqrt[\frac{1}{4}]{\left(1-{{\pi }_{\tilde{\mathrm{M}}\mathrm{e}\mathrm{j}}^{-}}^4-{{\upsilon }_{\tilde{\mathrm{M}}\mathrm{e}\mathrm{j}}^{-}}^2\right)\left(1-{{\pi }_{\tilde{\mathrm{M}}\mathrm{e}\mathrm{j}}^{+}}^4-{{\upsilon }_{\tilde{\mathrm{M}}\mathrm{e}\mathrm{j}}^{+}}^2\right)}\end{array}\right)}{6}}$$

(8)

3.5. Consumer Perception and Expectation Scores

Let ${\mathrm{G}}_{\mathrm{j}}$ be the gap score between the expectation and perception on SQ of jth item for all consumers. Then, Equation (9) is used to evaluate the gap score.

$${\mathrm{G}}_{\mathrm{j}}={\mathrm{D}}_{\mathrm{p}\mathrm{j}} -{\mathrm{D}}_{\mathrm{e}\mathrm{j}}, \mathrm{where} \mathrm{j} = 1, 2, 3, ...\mathrm{m}$$

(9)

4. Discussion

Table 2. Consumer expectations and perceptions presented by Pythagorean fuzzy values.

Variables	Expectation Scores	Perception Scores
Tangibility	[0.6973, 0.8081] [0.0561, 0.1176]	[0.71984, 0.84062] [0.06633, 0.14146]
T1	[0.7162, 0.8266] [0.0479, 0.1045]	[0.73568, 0.85635] [0.05808, 0.12914]
T2	[0.6927, 0.8025] [0.0581, 0.1199]	[0.71412, 0.83446] [0.06908, 0.14498]
T3	[0.6921, 0.8042] [0.0581, 0.1219]	[0.70719, 0.82874] [0.0726, 0.15125]
T4	[0.6883, 0.799] [0.0601, 0.1239]	[0.72215, 0.84293] [0.06545, 0.14058]
Reliability	[0.6899, 0.8002] [0.0593, 0.1219]	[0.71852, 0.84051] [0.06743, 0.14476]
R1	[0.7041, 0.8131] [0.0529, 0.1109]	[0.72028, 0.8426] [0.06644, 0.14377]
R2	[0.6877, 0.7985] [0.0603, 0.1241]	[0.71742, 0.83985] [0.06776, 0.14564]
R3	[0.6828, 0.7943] [0.0625, 0.1281]	[0.69894, 0.82071] [0.07876, 0.16159]
R4	[0.6981, 0.8087] [0.0554, 0.1162]	[0.7348, 0.85624] [0.05852, 0.13046]
R5	[0.6769, 0.7863] [0.0653, 0.1302]	[0.72138, 0.84315] [0.06578, 0.14212]
Responsiveness	[0.6859, 0.7961] [0.0616, 0.1256]	[0.7062, 0.8294] [0.0737, 0.15444]
RE1	[0.7062, 0.8168] [0.0521, 0.1113]	[0.71808, 0.84018] [0.06699, 0.14377]
RE2	[0.6758, 0.7874] [0.0663, 0.1344]	[0.69619, 0.81785] [0.07997, 0.16291]
RE3	[0.6745, 0.7847] [0.067, 0.1339]	[0.7073, 0.83094] [0.07282, 0.15389]
RE4	[0.687, 0.7955] [0.0608, 0.1228]	[0.70334, 0.82863] [0.0748, 0.1584]
Assurance	[0.6831, 0.7945] [0.0628, 0.1288]	[0.72721, 0.84986] [0.06303, 0.13882]
A1	[0.6819, 0.7943] [0.063, 0.1298]	[0.70477, 0.82687] [0.07568, 0.15785]
A2	[0.6884, 0.7985] [0.0605, 0.1243]	[0.74602, 0.86834] [0.05324, 0.12342]
A3	[0.6824, 0.7933] [0.0633, 0.1293]	[0.72666, 0.84975] [0.06237, 0.13772]
A4	[0.6797, 0.7919] [0.0643, 0.1319]	[0.73139, 0.85448] [0.06083, 0.13629]
Empathy	[0.6747, 0.7863] [0.0669, 0.135]	[0.71346, 0.83611] [0.07051, 0.15037]
E1	[0.6814, 0.7926] [0.0633, 0.1291]	[0.69806, 0.82005] [0.07898, 0.16192]
E2	[0.6929, 0.8048] [0.0576, 0.1208]	[0.71368, 0.8371] [0.0715, 0.15367]
E3	[0.6668, 0.7777] [0.0708, 0.1402]	[0.71654, 0.83875] [0.0682, 0.14597]
E4	[0.6534, 0.7646] [0.0786, 0.1525]	[0.71247, 0.8349] [0.07062, 0.15004]
E5	[0.6789, 0.7916] [0.0644, 0.1323]	[0.72666, 0.84975] [0.06336, 0.14003]

shows the defuzzied expectation, perceived value, and gap scores of consumers based on the SERVEQUAL scale. The scores of SQ dimensions were A2 (0.2463), T1 (0.2426), R4 (0.2416), A4 (0.2401), and E5 (0.2383), while the high expectations of consumers were observed as T1 (0.2360), RE1 (0.2345), R1 (0.2339), R4 (0.2330), and E2 (0.2323). Further, their expectations for SQ dimensions showed the scores of tangibility (0.6914), reliability (0.6817), responsiveness (0.6766), assurance (0.6740), and empathy (0.6635).

Table 3. Transformed defuzzified value of consumer expectations and perceptions.

Variables	Defuzzified Expectation (E) Values	Defuzzified Perception (P) Values	GAP (P-E)
Tangibility	0.6914	0.6498	−0.0416
T1	0.7150	0.6678	−0.0472
T2	0.6849	0.643	−0.0419
T3	0.6856	0.6359	−0.0497
T4	0.6799	0.6524	−0.0275
Reliability	0.6817	0.6492	−0.0325
R1	0.6988	0.6513	−0.0475
R2	0.6793	0.6481	−0.0312
R3	0.6735	0.627	−0.0465
R4	0.6923	0.6672	−0.0251
R5	0.6647	0.6522	−0.0125
Responsiveness	0.6766	0.636	−0.0406
RE1	0.7025	0.6486	−0.0539
RE2	0.6648	0.6239	−0.0409
RE3	0.6623	0.6374	−0.0249
RE4	0.6769	0.6339	−0.043
Assurance	0.6740	0.6594	−0.0146
A1	0.673	0.6338	−0.0392
A2	0.6798	0.6805	0.0007
A3	0.6728	0.6589	−0.0139
A4	0.6702	0.6643	−0.0059
Empathy	0.6635	0.6439	−0.0196
E1	0.6716	0.6262	−0.0454
E2	0.6865	0.6446	−0.0419
E3	0.6531	0.647	−0.0061
E4	0.6368	0.6425	0.0057
E5	0.6695	0.6590	−0.0105

also shows the IVPF-converted defuzzified values. The difference between SQ expectations and the perceptions of consumers was observed. A negative difference indicated that the retail supermarket did not provide satisfactory service, while a positive difference presented consumer satisfaction. The top five scores were obtained for RES1 (−0.0539), TAN3 (−0.0497), REL1 (−0.0475), TAN1 (−0.0472), and REL3 (−0.0465). Consumers were pleased with EMP4 (0.0077) and ASS2 (0.0217). In addition, the gaps in the SQ dimensions are sorted from highest to lowest: tangibility (−0.0416), responsiveness (−0.0406), reliability (−0.0325), empathy (−0.0196), and assurance (−0.0146). There was a significant difference in tangibility, responsiveness, and reliability compared with other SQ dimensions such as empathy and certainty.

5. Conclusions

We identified important service dimensions for the services of Indian retail supermarkets using SQ. The SQ dimensions were evaluated using IVPF. The key service dimensions were determined using the Fuzzy approach and ranked to develop a framework to examine their importance in terms of difference scores. Tangibility showed a large difference between expectations and perceptions of consumers in the SQ dimensions. Therefore, retail supermarkets need to invest their resources and time in the physical setting and equipment in stores to ensure employee participation in the service-providing process.

Since this research was conducted in retail supermarkets in a developing market, particularly India, it is necessary to conduct further studies to generalize the results and ensure validity. Other industries need to be explored in terms of SQ dimensions. We identified the dimensions through a literature review. Qualitative research on consumers is mandated to generalize the findings of this research for other retail companies. Since individual customers’ views on the SQ of retail supermarkets about customer loyalty were researched, other industries in other countries need to be included in further research.