• Tidak ada hasil yang ditemukan

3.1 Chapter Outline

In this segment, the chapter describes and provides overviews about the materials and methods applied to solve the research problem. Different related elements that are discussed include research design to point out how the researcher methodically gathered data, population to indicate the characteristics of the respondents, sample and sampling technique tactics to define how the researcher selected the participants, data collection procedures to indicate the instrument used and attributes of items to help collect data and finally data analysis approaches to describe how the researcher captured and analyzed descriptive and inferential statistics.

Overall, the research framework for this study is shown in figure 3 comprising a sequence of steps followed; collecting primary data extracted from the survey questionnaires, pre-processing data, analyzing data, and presenting results to prove the correlation of customer service modernization and customer satisfaction.

37

Figure 0-3 Research framework for this dissertation

3.2 Research Design

The researcher relied upon qualitative and quantitative research design to conduct this study. A research design is a plan outlining how the study's primary questions and issues are addressed.

The research used a descriptive survey (see appendix I). The respondents were made aware of the intended purpose of this study, especially the important of collecting primary data that could be utilized for analysis and informing conclusions pertaining to Shell petrol stations customer service in Oman. A descriptive research design indicates the methodically approach to gathering primary data with to characterize a certain incidence, circumstances, or population. Specifically, a research design aids in addressing the identified study problem's "when," "where," and "how,"

as opposed to its "why." Important descriptive studies include those conducted using attitude and opinion surveys as well as surveys of organizational practices. The researcher is supposed to identify and describe the variation in these and other phenomena to minimize error in analysis (Atmowardoyo 2018).

38

3.3 Population of the Study

The current study targeted the population of customers who fuel at Shell petrol stations in Oman.

Population refers to groups of people or things that are known to share important features. Every member of a group or category has a unifying quality with the others. The term "target population" refers to the intended group. It is common practice in descriptive studies to first characterize the study population, and then to report on the findings of a study conducted on a subset of that group. Populations for research may be segmented differently and distinctively, including by region, age, sex, and other qualities and factors including profession, religion, and ethnicity (Dannels 2018). For confidentiality, the researcher signed confidentiality undertaking (see appendix II).

3.4 Sample and Sampling Technique

A sample is a representative selection from the larger population from which judgments are made. The researcher has to make sure all the subgroups included in the study are properly accounted for, and the sample needs to be chosen such that it appropriately reflects the population as a whole (Dannels 2018). Yamane (1967) developed a formula to help calculate the appropriate sample size (at a 95% confidence level and a 5% level of significance), see Equation 3.1.

𝑛 = 𝑁

1+𝑁(𝑒2) ………. . Equation

3.1

Where n depicts size of the sample whereas N designates the size of population. With this formula, the sample size for the current investigation was determined as follows:

39 n = N / (1 + N(e^2))

where: n = the required sample size N = the size of the target population e = the desired level of precision (expressed as a proportion)

Assuming a desired level of precision of 5% (0.05), a 95% confidence level, and a target population size of 383, we can plug these values into the formula:

n = 383 / (1 + 383(0.05^2)) n = 383 / (1 + 0.9575) n = (383 / 1.9575) n = 195.6

Rounding up to the nearest whole number, the appropriate sample size is 196. Therefore, the required sample size is 196, assuming a desired level of precision of 5%, a 95% confidence level, and a target population size of 383.

The calculated sample size implies that 196 people who often fill up their cars at Shell stations in Oman constituted the study's real sample. Simple random sampling was used to select the participants. If they want to generalize about the whole population, researchers use a method called simple random sampling. When selecting a sample from a larger population, the simple random sampling technique assures the notion that every participant in the population of study has an equal chance or coincidence of being picked. Simplicity and unbiasedness are two of its most appealing features. Thus, simple random sampling may also be used for inferential statistical analyses on collected data (Puy et al. 2018).

3.5 Data Collection Instrument

The best data collecting strategies make it easy and convenient for respondents to supply their information. This study used online semi-structured questionnaires to collect primary data. This

40

researcher utilized Google forms as a tool to create, distribute, and collect responses by developing a questionnaire on the platform (see appendix I). When compared to other research procedures, like the interview, the questionnaire provides more structure, which may increase the efficiency and accuracy of the investigation. Respondents may complete the survey whenever and whenever they have internet access, which is a huge perk of an online survey. A higher response rate may be achieved by giving respondents sufficient time to finish the survey and the ability to save and resume work at a later time (Attal et al. 2018).

3.5.1 Collected Data Validity

A valid research instrument aids the researcher in achieving the intended research objectives and provides verifiable results. Content validity characterizes that questions that have been framed for a study are classified within the domain of a given variable or construct. In contrast, face validity entails judging whether the questions assist in attaining the research objectives (Yusoff 2019). 3.5.2 Collected Data Reliability

The reliability of research instrument is the capacity of this tool to consistently provide the same data outcomes when used several times. Cronbach's alpha measures the internal consistency of the questionnaire. The researcher utilized this measurement used to evaluate the reliability of the constructed questionnaire that was applicable for data collection related to the concepts of the explored research problem. In most cases, a coefficient of 0.7 or above is considered satisfactory to reveal the reliability that a questionnaire of a study adopts (Quintão et al. 2020).

41

3.6 Data Analysis Techniques

3.6.1 Regression Analysis

Prior to being analysed, coding and cleaning of data is necessary. Descriptive statistics captured include frequency, percentage, mean, and standard deviation. For inferential analysis, multiple linear regression analysis also used to analyse collected data. Tables, charts and graphs are used for data presentation. For analysis, the following multiple regression model was used:

𝑌 = 𝛽0+ 𝛽1𝑋1+ 𝛽2𝑋2+ 𝛽3𝑋3+ 𝛽4𝑋4+ 𝜀 Where; Y Customer satisfaction in Shell petrol stations in Oman

β0 represents the y-intercept

β1 denotes coefficient of service visibility β2 depicts coefficient of service reliability β3 designates service effectiveness and efficiency β4 representscoefficient of service customization

X1 represents the independent variable of service visibility X2 represents the independent variable of service reliability

X3 represents the independent variable of service effectiveness and efficiency X4 represents the independent variable of service customization

ε represents the error term

Multiple linear regression analysis conducted in SPSS version 27.0 and SPSS AMOS version 14.0 used for just finding factor loadings in validity sections.

42 3.6.2 Hypotheses Testing

The research hypotheses are tested using the multilinear regression model, generating the summary model, analysis of variance (ANOVA), and regression coefficients in SPSS to evaluate the statistical significance of the independent variables and their impact on the dependent variable - customer satisfaction. The summary model gives an overview of the model, including the coefficient of determination (R²), the F-test for overall significance, and the degrees of freedom for the model and error terms. The calculated R-squared value indicates the extent to which the independent variables (service visibility, service reliability, service effectivity and efficiency, and modernized customer service) jointly explain variations in the independent variable (customer satisfaction). The regression coefficients provide the estimated values of the independent variables, their t-values, p-values, and standard errors, which are useful for testing the null hypotheses. The H0 is rejected if the calculated p-value is less than 0.05 (p < 0.05). The ANOVA table indicates the source of variation, degree of freedom (df), F- ratio, mean squares, error terms, and p-value, which help in determining the overall significance of the regression model. Therefore, the hypotheses for this study are tested using regression analysis.

3.7 Theoretical Framework

3.7.1 SERVQUAL Model

This study used SERVQUAL theoretical framework to assist in explaining customer behaviour based on different service quality dimensions. Maghsoodi et al. (2019) clarify that SERVQUAL instrument is a significant instrument to assess service quality. For its distinctive dimensions, they range from tangibles (facility appearance, service provider appearance), reliability (assured

43

accurate service delivery), responsiveness (willingness to deliver service in a rapid manner), and assurance (trust, motivation, confidence) to empathy (individualized delivery, adjusted attention). These aspects are clear indication that service delivery and service quality are closely associated with customer satisfaction. The definition of service quality concept from the outlook of customer satisfaction is described as the realization of perceived quality that extends beyond customer’s expectations, aspirations, and needs (Fida et al. 2020). These explanations underline the applicability of the model in the energy service sector. The rational of its choosing for the current dissertation is affirmed regarding how the model facilitates discussing the extent to which service delivery in fuel stations depict customer satisfaction and, more so, the impact of service reliability, service visibility, and service effectiveness and efficiency.

3.7.2 SEVQUAL in the Fuel Sector

The utilization of SERVQUAL model in oil businesses service industry cannot be exaggerated because the framework is revealed to facilitate measuring and determining the importance of service quality. In a study related to delivery of services in fuel oil channels in Turkey, service quality is revealed as an important factor that affect customer’s choice. The dimensions that enabled the researcher to draw this conclusion are access, extra non-fuel services, assurance, empathy, reliability, and tangibles (Kaynar 2020). A peculiar characteristic of Kaynar’s (2020) study is how it distinguished the model’s implementation in the fuel sector. The author created a questionnaire and examined 22 unique items related to servqual scale adapted from Turkish fuel business outlets (see appendix III). Similarly, Purohit and Jain (2022) reviewed the application of various service quality dimensions associated with SERVQUAL framework in petrol-retailing in Indian context. Result from literature findings revealed that petrol retailing

44

underlying service quality dimensions are context-specific and that customers weigh these aspects differently, which made the authors conclude that petrol retailing service are principally identified from customers’ perspective (Purohit & Jain 2022). However, in their previous study conducted in 2020, Purohit and Jain (2022) had identified that petrol retailing outlets had thus far not managed to enhance customer expectations. Research in this area had a literature gap to fill. Therefore, this current work is positioned to evaluate how service quality and service delivery are impacted by customer satisfaction.

3.8 Chapter Summary

Adequate understanding of the approaches to collecting data from the customers that use Shell petrol station provided the context to solving the research problem. Determining the methods and materials to utilize in data collection is crucial; it paves way to accomplish the outlined aim of the study and set the groundwork for getting and analysis diverse responses employed to reject or accept the hypotheses. In the next chapter of data processing and results presentation, the data collected utilizing the study methodology are described to pave way for their interpretation.

45