No conclusive statements can be made on the main hypothesis by studying the scatter diagram in figure 5.5 or from the descriptive statistics as shown in table 5.6. Hence, the data are further analysed by using inferential statistics. Results from section 5.4 are summarised in table 6.3.
Table 6.3 Summary of testing results
Service quality is dependent on service dealerships for passenger vehicles Not Accepted Service quality is dependent on service dealerships for passenger and commercial vehicles Not Accepted Service quality is dependent upon the vehicle type (passenger vehicles and commercial
vehicles) Not Accepted
The arithmetic means of the scores between the passenger vehicle and commercial vehicle
customers are different no significant difference Service quality is dependent to the type of inventory system Not Accepted
Customers view the levels of service quality provided by the dealerships under the two
different inventory systems are different no significant difference Customers view dealerships as having different level of service quality in the tangible
dimension no significant difference Customers view dealerships as having different level of service quality in the reliability
dimension no significant difference Customers view dealerships as having different level of service quality in the
responsiveness dimension no significant difference Customers view dealerships as having different level of service quality in the assurance
dimension no significant difference Customers view dealerships as having different level of service quality in the empathy
dimension significant difference
There are two predominant factors that may affect the test results of the main hypothesis; namely, the nature of the vehicle the dealerships service and the type of inventory system the dealership uses. During the interrogation of the main hypothesis, different tests were carried out by dividing the dealerships into two groups based on the above factors. Statistical comparison of the SERVQUAL scores does not indicate any significant difference in the service quality provided by dealerships under the new inventory system, which means that the advantage of increased availability of stocks has no significant influence on customers’ view on the level of service quality. In
addition, no significant differences in the level of service quality are notified by customers in passenger or commercial vehicle dealerships. Results from the Chi- square tests indicate that service quality appears to be independent of the dealerships. There are other factors that may affect the studies presented in table 2.2, but the influences by these factors are either not highlighted or investigated, thus making comparison to the current study impossible.
Once these extraneous factors are investigated, the main hypothesis is further analysed in the five dimensions that construct the overall service quality.
Various t-tests show no significant differences along the dimensions of service quality with the exception of the empathy dimension. The differences in the empathy dimension are significant but the importance of this dimension is not heavily weighted. Therefore, such a difference is not reflected on the overall service quality score when tested.
When comparing the dimensional scores in this research to the studies in table 2.2, none of the studies score positively in any dimensions. There is a difference in the current findings; in general, customers view dealerships involved in this study performed better then expected in the tangible dimension. This indicates customers’ perceptions exceed their expectations in such dimension. Although dealerships generally score positively in the tangible dimension, its effects on the overall SERVQUAL scores are negligible due to its low importance weighting.
In addition to the positive scores for the tangible dimension, table 5.5 indicates 35 out of the 90 respondents provide positive scores for the dealerships. This phenomenon disagrees with the research findings of Brown et al. (1993: pp.127-139); which claims “the expected or desired level of service is almost always higher than the perceived level of actual service”. Nonetheless, the average SERVQUAL scores for all the dealerships are negative, similar results are reported in various studies in a different industry.
From the data collected and the findings from various inferential statistics, within the accuracy of this experiment, customers’ views on service quality are the same. This is not influenced by the vehicle type and the inventory method. The only significant difference found is in the empathy dimension, although no explanation is offered for this difference. This study does not disprove the main hypothesis of this research, suggesting that customers view selected service dealerships as having equal levels of service quality.
The next section discusses the relationship between perception scores and SERVQUAL scores.
6.4
Perception Scores and SERVQUAL Scores
Comparison between perception and SERVQUAL scores is presented in section 5.5. The degree of association between the two variables is denoted byR2. The value of
2
R may vary from zero (no correlation) to one (perfect correlation). The correlation value is shown asR2 =0.44. If the R2 value had been 1, it means that all the points fell on a straight line and there is a direct correlation between the two variables. This is clearly not the case as the R2 value is much less than 1. With R2 values much less than 1, these indicate the variables are not completely dependent on each other. Perception scores alone only associated to 44 per cent of the total variation in the SERVQUAL scores. A reason the correlation is low is that customers have different expectations. It seems that perception scores alone provide a partial but not a full measure of service quality, especially as managers may prefer to know their customers’ expectations as well. A superior approach may be to report perceptions, expectations, and SERVQUAL scores.
This chapter concludes with examining the applicability of the PZB approach in the context of this research.