A Research Report based on an analysis of Service Quality of Sports Centre

Table of Content

1. A brief evaluation of the method and data collected

2. Analysis

3. Reflection and Conclusion

4. Bibliography

A Research Report based on an analysis of Service Quality of Sports Centre

It is an immensely complex exercise to measure the quality of a service. There are differences between product and service. In case of product, there are precise specifications such as length, depth, width, weight, color etc. On the other hand, a service can have several intangible or qualitative features. In addition, there are some expectations of the consumer regarding the service, which can differ significantly based on a range of factors such as experience, personal requirements and what other people may have said to them.

To determine the service quality Parasuraman, Zeithaml, and Berry (1994) have developed a methodology known as ‘SERVQUAL’ which is a service quality questionnaire survey methodology. SERVQUAL model suggested five dimensions of service quality. They are reliability, responsiveness, tangibles, empathy, and assurance. Experiences or perceptions and expectation are measured in SERVQUAL model and SERVQUAL model has been well used in retail surroundings (Berry, 1986; Brown et al., 1993; Carmen, 1990; Finn and Lamb, 1991; Dabholkar et al., 1996) including banking, fast food, dry cleaning (Cronin and Taylor, 1992) and health care (Babakus and Mangold, 1989). In diversified businesses as the technologically advanced industrial market(Pitt et al., 1992) and B2B (business to business) services (Brensinger and Lambert,1990) include the application of the SERVQUAL.

However, this model raises some doubts as to whether expectations are being assessed at all. The gap measured by the SERVQUAL model between expectation and perception is uncertain as to whether this is an indication of initial expectation or how the service is continuously reevaluated whilst it is being executed, or experienced. Therefore, SERVPERF model is proposed by Cronin and Taylor (1992), and this model eliminates expectation from the SERVQUAL, and it measures the service quality based on service perception. In this paper, the SERVPERF model is used to measure the service quality of the sports centre.

A brief evaluation of the method and data collected:

The results in this report are based on the survey conducted by a pair of undergraduate students, and the focus of this research is the sports centre which is a part of a campus in university of Lincoln.

A probability sample of students at the University was taken in the sampling method. Moreover, the sampling frame was composed of the e-mail addresses of all students. Simple random sampling was used. The sample size is 200, which is appropriate to present the population. Over sampling was determined based on the anticipation of non-response. An initial sample of 300 was conducted and students were emailed directing them to a web link to the survey hosted by surveymonkey.com. After 2 weeks, with various reminders the total response was 202 useable replies. There can be a chance of non-response bias. If all the questions have not been answered or returned, Questionnaire non-response may occur (Hussy and Hussy, 1997). However, the response rate of 67% was reasonable. Men accounted for 62.4% of the sample and female accounted for 37.1% of the sample. Respondents who use the sport centre twice a week is 23.3% of the sample, respondents who use the sport centre once a week or less than a week is 19.3% of the sample, and respondents who never use the sport centre is 22.8% of the sample. Member of the sports centre accounted for 51.6% of the sample and customers who pay each time accounted for 48.4% of the sample. Those who study sports accounted for 33.7% of the sample and business student, media student, and other course student accounted for sequentially 20.3%, 11.9% and 34.2% of the sample. Therefore, we concluded that the sample could be deemed representative of the population from which it was drawn.

A standardized questionnaire is used to collect the data. The questionnaire contained three sets of Questions. The first set consisted of 21 separate questions regarding five dimensions of service quality. Sub questions 1 through 6 calculated tangible elements of the service. Sub Questions 20 and 21 calculated the empathy dimension, whilst 16 through 19 calculated the assurance dimension. Questions 7, 10, 11, 12 focused on reliability and questions 8, 9, 13, 14, 15 calculated responsiveness. The questionnaire follows a same format to previous research in this field. Moreover, measurement validity was maintained; five point likert scales is used for rating the service performance in the first set questions of the questionnaire. The second set of questions in the questionnaire used to weight the importance attached to each of the five dimensions of service quality. The third set of questions was used to get the information of respondent characteristics, which are independent variables. This set of questions allowed us to determine the demographics of the sample and facilitate analysis of variance later. In addition, the questions of questionnaire appear to be measuring the construct, and it was pre-tested for face validity with possible users of sports centre. Moreover, no problems were found with face validity with potential users of the sports centre and the questionnaire was well thought-out to be reliable. The data collected from this survey was then entered into the Statistical Package for the Social Sciences (SPSS). The next section will analyze the data and discuss the key findings.

Analysis:

1.)

Initially, we want to find out that how the respondents rated the five dimensions of service. Therefore, we conducted the descriptive analysis to find that how the respondents rated the each dimension of service using SPSS. The Output of SPSS is given below in the Table 1.

Table 1: Descriptive Statistics of the importance of each dimension of service quality

illustration not visible in this excerpt

It can be realized from the table 1 that the mean of the tangibles dimension is the highest (25.34), which means more respondents give priority to the appearance of the sports centre, physical facilities such as staff, gym equipment, the sports hall, and changing rooms of the sports centre than the other dimensions. Therefore, the sports centre should give more priority to the appearance of the sports centre, physical facilities such as staff, gym equipment, the sports hall, changing rooms than other dimension of service suggested by the SERVPERF model. The mean of the confidence dimension is the lowest (17.26), which means that fewer respondents give priority to the knowledge and manners of the sports centre's staff, and their ability to supply and maintain confidence and trust within its customer than other dimensions. Therefore, we can say that the sports centre should provide less priority to the knowledge and manners of the sports centre's staff, and their ability to supply and maintain confidence and trust within its customer than the other dimensions of service. In case of reliability and responsiveness dimension, the mean is almost same which means that almost same numbers of respondents are attached to the ability of the sports centre and its staff to implement the service promised reliably and accurately, and the sports centre's willingness to help users and provide a prompt and friendly service. Therefore, the sports centre need to give the same priority to these two dimensions.

The standard deviation for confidence dimension of service quality is 7.133 which is the lowest standard deviation, and it indicates that the data points remarkably close to mean comparing to the other dimensions. It means that a majority of the respondents gauge the importance of confidence dimension very closer to the mean 17.26. On the other hand, the standard deviation for tangibles dimension of service quality is 11.399 which is the highest standard deviation among the standard deviations of five dimensions, and it indicates that the data are extended over a large range of values comparing to the other dimensions. Moreover, it means that a majority of the respondents gauge the importance of tangibles dimension not closer to the mean 25.34 comparing to confidence dimension. Therefore, the standard deviation of the empathy, responsiveness, Reliability are sequentially 10.10, 8.07, 7.42 which actually represent that how the majority of the respondents are closer to the mean of these dimensions.

The skewness (1.785) of the tangibles dimension is greater than 0 which represents the right skewed distribution and the majority of the values are found on the left of the mean, with extremely high values to the right. In addition, the kurtosis is 7.506 which represent Leptokurtic distribution, more sharp than a normal distribution, with values found around the mean and thicker tails. This means high possibility for extremely high values. Moreover, the skewness (1.258) of reliability dimension is greater than 0 which represents the right skewed distribution and the majority of values are found on the left of the mean, with extremely high values to the right. In addition, the kurtosis (4.898) is greater than 3 which represent Leptokurtic distribution, more sharp than a normal distribution, with values found around the mean and thicker tails. This means high possibility for extremely high values. In addition, the skewness (1.401) of responsiveness dimension is greater than 0 which represents the right skewed distribution and most values are found on the left of the mean, with extremely high values to the right. In addition, the kurtosis (4.033) is greater than 3 which represent Leptokurtic distribution, more sharp than a normal distribution, with values found around the mean and thicker tails. This means high possibility for extremely high values. Additionally, the skewness (.547) of confidence dimension is greater than 0 which represents the right skewed distribution and most values are found on the left of the mean, with extremely high values to the right. In addition, the kurtosis (.808) of confidence dimension is less than 3 which represent Platykurtic distribution, more flat than a normal distribution with a wider peak. The possibility for extremely high values is less than for a normal distribution, and the values are widely extended around the mean. Furthermore, the skewness (0.445) of empathy dimension is greater than 0 which represents the right skewed distribution and most values are found on the left of the mean, with extremely high values to the right. In addition, the kurtosis (0.599) of empathy dimension is less than 3 which represent Platykurtic distribution, more flat than a normal distribution with a wider peak. The possibility for extreme values is less than for a normal distribution, and the values are widely extended around the mean.

2.)

In the previous section, we have found that respondents were not attached in the same aspects to the each dimension. Their attachment to each dimension was different. In this section, we want to find out that whether the respondents vary in the importance attached to each of the five dimensions according to their personal characteristics. In the following, we conducted five different One-way Anova test based on their characteristics like gender, usage, year, member and course. We will include the Hypothesis testing once to demonstrate the procedure. We will report statistically significant results quoting the statistics rather than the procedure for the rest of the one-way Anova tests. The results of the SPSS of the five one-way Anova tests are presented and analyzed below:

a) An Analysis of variability between Gender and respondents importance attached to each dimension:

A one-way Anova test is conducted to analyze the variability between Gender and respondent’s importance attached to each dimension by using SPSS. We find that the significance value for tangibles, reliability, responsiveness, confidence, and empathy are sequentially 0.853, 0.321, 0.552, 0.934, and 0.133. For this test, the key result is to look at the significance of the F-test, and there are no values less than or equal to .05. Therefore, it can be said that there are no statistically significance differences between gender and respondents importance attached to each dimension. No output for this test is given as all values are statistically insignificant. For this analysis, the following hypotheses are taken into consideration. They are given below:

Hypothesis for tangible Dimension:

Null Hypothesis (H0): Respondents importance attached to tangibles dimension does not differ according to the gender

Alternative Hypothesis (H1): Respondents importance attached to tangibles dimension differs according to the gender

Hypothesis for reliability Dimension:

Null Hypothesis (H0): Respondents importance attached to reliability dimension does not differ according to the gender

Alternative Hypothesis (H1): Respondents importance attached to reliability dimension differs according to the gender

Hypothesis for responsiveness Dimension:

Null Hypothesis (H0): Respondents importance attached to responsiveness dimension does not differ according to the gender.

Alternative Hypothesis (H1): Respondents importance attached to responsiveness dimension differs according to the gender.

Hypothesis for confidence Dimension:

Null Hypothesis (H0): Respondents importance attached to confidence dimension does not differ according to the gender.

Alternative Hypothesis (H1): Respondents importance attached to confidence dimension differs according to the gender.

Hypothesis for empathy Dimension:

Null Hypothesis (H0): Respondents importance attached to empathy dimension does not differ according to the gender.

Alternative Hypothesis (H1): Respondents importance attached to empathy dimension differs according to the gender.

b) An Analysis of variability between usage and respondents importance attached to each dimension:

A one-way Anova test is conducted to analyze the variability between Usage and respondent’s importance attached to each dimension by using SPSS. We will not include the Hypothesis testing for this One-way Anova test and we will report statistically significant results quoting the statistics rather than the procedure. For this test, the key result is to look at the significance of the F-test, and there are no values less than or equal to .05. Therefore, it can be said that empathy dimension has the significance value equal to .05. Therefore, there is statistically significance differences between usage and respondents importance attached to empathy dimensions. The significance values of the F-test for other dimensions are greater than .05 so there are statistically insignificant.

[...]