Page 107 - Contributed Paper Session (CPS) - Volume 8

CPS2205 Abdul Aziz A. Rahaman et al.
final model gives better predictions than if you just guessed based on the
marginal probabilities for the outcome categories.
It was also observed that the p-value (0.640) is greater than the
significance level (0.05). This means that we fail to reject the null hypothesis
that the fitted model is consistent with the observed data. Thus we conclude
that the data and the model predictions are similar at 95% confidence level
which implies a good model. The Pseudo R-square (Nagelkerke=76.8%)
indicates that the predictor variables explains most of the proportions of
variation between customer satisfaction (response). There is however about
23.2% of the variability which is unaccounted for, which may be due to
research related errors. Additionally, in the SEM framework, Service Quality
accounted for about 83.9%, 82.8% and 80.5% of the variability recorded in
Assurance, Tangibility and Reliability respectively. Meanwhile, Loyalty
explained about 86.2%, 83% and 80.1% for the variance in Trustworthiness,
Commitment and Corporate Image respectively. Customer satisfaction
recorded 90.6% for Service Charge. This implies that Customer satisfaction
accounts for majority of the variation in the bank’s Service Charge.

4. Discussion
This study has established that there is a link between service quality and
customer retention at Universal banks’ in Ghana. This study finds service
quality impacts on customer satisfaction and customer retention at Universal
banks’ in Ghana. This result is consistent with finding of other scholars
(Ndubisi, 2007 & Titko and Lace 2010). Usually, customer satisfaction is the
important predictor of customer retention, but this study establishes service
quality has great impact on customer retention simultaneously with customer
satisfaction. Again, the empirical results show customer satisfaction has the
mediating role between service quality and customer retention. It implies that
quality has a direct impact on customer satisfaction and indirect impact on
customer retention through satisfaction, which is at variant to other studies
(Zineldin, 2006). Meanwhile, previous researches (Jamal and Nasr, 2003 and
Parasuraman et al., 1993) found that there is no important relationship
between customer satisfaction and tangible aspects of service quality, in
contrast, this study noted that tangible significantly influence customer
satisfaction. The study here asserted that satisfaction is strongly influenced by
service quality dimension on responsiveness which contradicts a previous
study Banergee (2012). There are overwhelming arguments that it is more
expensive to win new customers than to keep existing ones (Hormozi and
Giles, 2004).
In conclusion, the ordinal logit model fit the data set adequately and
provides a better model fit indices compared to the SEM, where some of its
model fit indices are off the threshold. Also, the parameter estimates and fit

96 | I S I W S C 2 0 1 9