Sunteți pe pagina 1din 11

African Journal of Business Management Vol.6 (46), pp. 11426-11436, 21 November 2012 Available online at http://www.academicjournals.org/AJBM DOI: 10.5897/AJBM12.

445 ISSN 1993-8233 2012 Academic Journals

Full Length Research Paper

Customer perception towards banking sector: Structural equation modeling approach


R. Renganathan, S. Balachandran and K. Govindarajan
School of Management, SASTRA University, Thanjavur, India.
Accepted 27 August, 2012

In order to survive and excel in the competitive scenario, organizations have to understand the customers requirements. Service quality can be measured with the help of the two important instruments, namely SERVQUAL and SERVPERF. BANKSERV model can be used to measure the service quality of the banking sector. Customers perception with regard to various services rendered by the banking industry was taken for this study. Data were collected from 300 customers of the bank located in different parts of Tamilnadu and Pondicherry, India. In order to evaluate the association between the variables used in the model, structural equation modeling (SEM) was used for data analysis. The findings of the research showed that, absolute fit indices fits the sample data and reveals that the proposed model has the acceptable fit, by way of satisfying the recommended values. Key words: Banking industry, service quality, customers perception, structural equation modeling (SEM), confirmatory factor analysis.

INTRODUCTION Cronin and Taylor (1994) studied the service quality in banks, pest control, dry cleaning, and fast food services with the help of SERVPERF which clarified the difference in the global measure of service quality. In banking sector, service quality is related to bank loyalty through satisfaction (Bloemer et al., 1998). In the globalized environment, banking industry in India is witnessing keen competition in acquiring new and retaining existing customers. Nationalized banks in India have to compete with private banks, and multinational banks. The BANKSERV model was developed in Australia by Avkiran (1994) to measure service quality in retail banking as perceived by customers. Service quality is measured based upon the expectation and perception of the customers in BANKSERV model. The objective of this study was to explore the impact of individual aspects of banking operation on various types of customers perceptions of service quality. In this study, data were collected based upon the bank customers perception (SERVPERF) about the four- factor pertaining to BANKSERV model (Avkiran, 1994:15) with some modifications (refer appendix) in order to suit Indian environment. Data were collected from 300 customers of nationalized banks such as State Bank of India, Indian Overseas Bank, Bank of India, Indian Bank, and Canara Bank from several places like Thanjavur, Trichy, Kumbakonam, Dindugal of Tamilnadu and Pondicherry, India. Banking customers opinion about banking service quality were elicited from the perspective of banks staff conduct, credibility, communication and customer accessibility. Statistical methods like exploratory factor analysis and confirmatory factor analysis were used. To develop a suitable model, structural equation modeling (SEM) technique was used for analyzing the data. All the measures used in structural equation modeling fit the recommended values indicating a good fit for the collected data.

Significance of the study Nationalized banks in India have to compete with other private banks and also with foreign banks, since banking

*Corresponding author. E-mail: renganathanss@yahoo.com.

Renganathan et al.

11427

industry is privatized and globalized in India. In order to sustain and excel in the competitive environment nationalized banks in India have to satisfy and delight their existing customers and formulate suitable strategies to acquire new customers. Hence, this type of research gains importance.

LITERATURE REVIEW Most research in the area of service quality has been based upon the model developed by Parasuraman et al. (1985, 1988), which incorporates a comparison of customer expectations and perceptions of service performance. The long term survival and success of any organization depends upon the high quality products and services providing value for money, for the consumer, which will lead to consumer satisfaction and loyalty (Zeithaml et al., 1990; Robledo, 2001). The positive consequences of companies achieving high levels of customer satisfaction and service quality are well documented (Rust and Zahorik, 1993; Zeithaml et al., 1990). Service quality is a critical issue in the service industry and of particular importance for financial service providers who characteristically offer products that are homogeneous in nature (Stafford et al., 1998). Service business success has been associated with the ability to deliver superior service (Gale, 1990). If we want to manage something, it should be measured first. Without measurement, managers cannot be sure of whether service quality gaps exist or not (Christopher et al., 2006) and, of course, measurement is needed to determine whether goals for improvement are being met after changes have been implemented (Christopher et al., 2006). In general, it is difficult to measure and quantify service quality. The main purpose of measuring service quality is to ensure whether service is provided as per the expectations of the customers. Zeithaml et al. (1990) suggest that the criteria used by customers in molding their expectations and perceptions fit in five dimensions of service quality: tangibles, reliability, responsiveness, assurance and empathy. According to Parasuraman et al. (1988), service quality has become a great competitor and the most powerful competitive weapon, which many leading service organizations possess. The SERVQUAL model has come in for criticism from, for example, Johns (1996) argues that it may be too cumbersome for general use, but provides a useful service tool, which can point the way forward for more rigorous quality monitoring. Cronin and Taylor (1992) were the first to offer a theoretical justification for discarding the expectations portion of SERVQUAL in favor of just the performance measures included (that is, what they termed SERVPERF). The term performanceonly measures has thus come to refer to service quality

measures that are based only consumers perceptions of the performance of a service provider, as opposed to the difference (or gap) between the consumers performance perceptions and their performance expectations (Mass et al., 2010). SERVQUAL also has some criticisms on both operational and theoretical issue (Buttle, 1996). Asubonteng et al. (1996) did 18 critical reviews on empirical studies of service quality and conclude that SERVQUAL instrument is industry specific. When SERVQUAL applied for retail banking, problems were identified with regard to its dimensionality and the value of expectation scores (Lam, 1995). In order to suit the Australian banking industry, Avkiran (1994) propounded BANKSERV instrument based on SERVQUAL model. Avkiran (1994) evaded the probable psychometric difficulties connected with SERVQUAL (Avkiran, 1999:62). Negatively worded questionnaire items of SERVQUAL instrument were also avoided (Babakus and Boller, 1992). BANKPERFs major advantage over BANKSERV is that there is strong theoretical support for a performance-only measure of service quality (Cronin and Taylor, 1992, 1994). Structural equation modeling (SEM) is a statistical modeling technique that combines factor analysis and multivariate multiple regressions (Hair et al., 2006). Structural equation provides estimation of multiple and interrelated dependence relationship and the capacity to stand for unobserved concepts in these association and explanation for measurement error in the estimation process (Hair et al., 1998). The primary aim of SEM is to explain the model of a sequence of inter-related dependence associations simultaneously among a set of dormant (unobserved) constructs, each measured by one or more manifest (observed) variables (Yu-Kai, 2009). SEM is a multivariate technique which combines confirmatory factor analysis modeling from psychometric theory and structural equations modeling (Yu-Kai, 2009). In order to recognize a right model for the sample data, fit indices have no single statistical test of significance (Schumaker and Lomax, 1996). There are number of goodness of fit (GOF) indices with which to make comparisons, thus fit should be evaluated from the standpoint of numerous fit statistics (Campbell et al., 1995:6). The overall fit measures, the goodness-offit statistic (GFI), adjusted goodness-of-fit statistic (AGFI), root mean squared residual (RMR), and the normed fit index (NFI) (Bentler and Bonett, 1980), are all useful measures in assessing the quality of the hypothesized measurement model. Absolute fit indices determine how well a priori model fits the sample data (McDonald and Ho, 2002). Confirmatory factor analysis (CFA) is one of the most commonly employed tools to test the construct validity of developed instrument (Hair et al., 2006). This technique provides a more precise interprettation of dimensionality than the exploratory factor analysis (EFA) technique (Diana, 2006). CFA can be used as an interpretation of model fit indices

11428

Afr. J. Bus. Manage.

(Schumaker and Lomax, 1996).

Objectives of the study The objectives of this study are as follows: 1. To study the demographic profile of the banking customers. 2. To identify the determinants of service quality which are very vital to banking industry. 3. To examine the usefulness of the BANKSERV (measuring only performance) model for assessing customer perceptions of service quality in banking industry. 4. To evaluate whether all the measures fit the recommended value, indicating a good fit of the structural model for the collected data.

customers involved in this study. As per the Table 1, out of 300 bank customers, 55.3% were male and 44.7% were female. With regard to the level of education, 30% of bank customers had post graduate degree, 34.3% of the bank customers had under graduate degree, and 7.7% of the bank customers had diploma qualification. With reference to monthly income of the customers, 23.3% bank customers income were below 10,000, 21.7% of the bank customers income were in the range of 11,000 to 20,000, 15.3% of the banking customers income were in the range of 21,000 to 30,000 and 8.3% of the banking customers monthly income were in the range of 31,000 to 40,000. The collected data reveals that, in terms of occupation, 25.7% of the banking customers were private employees, 19.7% of the banking customers were Government employees, 18% of the banking customers were self employed and 16.3% of the banking customers were home makers.

METHODOLOGY Data collection and sample Avkirans (1994) 17-item BANKSERV instrument was adopted with some modifications (Appendix 1) for data collection. In this research, a perceptions-only measure (similar to SERVPERF) was used to collect data from the bank customers. A five-point scale (5 indicating strongly agree and 1 indicating strongly disagree) was used in preference to a seven-point scale to increase the sensitivity of the measure. In this study, Banking customers perceptions were measured with a self administered questionnaire. This study uses both primary and secondary data. Secondary data was collected from websites and various journals. Primary data was collected based upon the bank customers perception (SERVPERF) about the four factors pertaining to BANKSERV model (Avkiran, 1994:15) with some modifications. Data were collected from 300 bank customers of State Bank of India, Indian Overseas Bank, Bank of India, Indian Bank, Canara Bank from several places like Thanjavur, Trichy, Kumbakonam, Dindugal of Tamilnadu and Pondicherry, India. Banking customers opinion about banking service quality were elicited from the perspective of banks staff conduct, credibility, communication and customer accessibility. Out of the total 400 questionnaires distributed, researcher was able to collect only 300 questionnaires fully completed in all aspects which amount to 75% of response rate.

Construct reliability and BANKSERV dimensions

validity

analysis

for

Data analysis Collected data were analyzed with the help of software package SPSS and analysis of moment structure (AMOS) 16. Statistical techniques like descriptive analysis, reliability analysis, exploratory factor analysis, confirmatory factor analysis were used to evaluate the service quality. Structural equation modeling (SEM) was used for data analysis.

While applying Likert- types scales in research, it is necessary to calculate the Cronbachs alpha coefficient for reliability and consistency (Joseph et al., 2003). Table 2 shows the component and total reliabilities of SERVQUAL scores. The findings show that Cronbachs alpha for all dimensions except for customer accessibility are above 0.70 (George and Mallery, 2003: 231) which indicates a high level of internal consistency for the scale. Moreover, as per the Table 2, over all Cronbachs alpha value for the BANKSERV dimensions is 0.904. The Cronbachs alpha values for the perceptions subscales are 0.835, 0.749, 0.852 and 0.643 for staff conduct, credibility, communication and customer accessibility. Table 3 shows the corrected item-total correlations; that is, the scores for an item and the summated scores of the rest of the items comprising a subscale (for example, the subscale measuring the credibility dimension of service quality) were correlated. Of the individual perception items, all the items had correlation with the total scores that was higher than the 0.35 cut-off value suggested by Saxe and Weitz (1982). The item-total correlations for the perceptions scale are ranging from 0.374 to 0.705 (Table 3). Table 3 also contains item means and standard deviations.

Exploratory factor analysis Factor analysis is used to identify a smaller number of factors underlying larger number of observed variables. Table 4 shows Kaiser-Meyer-Olkin (KMO) and Bartlett's Test. The KMO ranges from 0 to 1, with higher values indicating greater suitability. Ideally, this value is to be greater than 0.7. According to Kaiser, a KMO measure of

RESULTS AND DISCUSSION Profile of the respondents Table 1 shows the demographic profile of the banking

Renganathan et al.

11429

Table 1. Demographic profile of the bank customers (n=300)

S/N

Characteristics

Categories 18- 25 years 26-33 years 34-41 years 42-49 years 50-57 years 58 years and above Male Female SSLC//HSC Under graduation Post graduation Diploma others Below 10,000 11,000- 20,000 21-000- 30,000 31,000- 40,000 41,000- 50,000 51,000 & above Not applicable Self employed Government employee Private employee Student Home makers others

Age

Number of respondents 53 63 67 55 34 28 166 134 75 103 90 23 9 70 65 46 25 16 14 64 54 59 77 40 49 21

Percentage (%) 17.7 21.0 22.3 18.3 11.3 9.3 55.30 44.70 25.0 34.3 30 7.7 3.0 23.3 21.7 15.3 8.3 5.3 4.7 21.3 18.0 19.7 25.7 13.3 16.3 7.0

Gender

Educational qualification

Monthly income (* 10Symbol of Indian country money-Rupees)

Occupation

Table 2. Result of reliability analysis for BANKSERV dimensions.

Dimensions Staff conduct Credibility Communication Customer accessibility Overall reliability analysis for BANKSERV dimensions

Number of attributes 7 3 5 2 Cronbach's alpha No. of Items

Cronbachs alpha 0.835 0.749 0.852 0.643 0.904 17

0.9 to 1.0 is marvelous, 0.8 to 0.9 meritorious, 0.7 to 0.8 middling, 0.6 to 0.7 mediocre and 0.5 to 0.6 miserable (Marcus et al., 2006). Table 4 shows, with regard to bank customers perception, Kaiser-Meyer-Olkin measure of sampling adequacy (MSA) is 0.833 and Bartletts test of 2 sphericity is significant [Chi-square x (55) = 1025.8,

p<0.001]. The correlations between the variable and factor, with possible values range from -1 to +1 are shown as the rotated factor loadings in Table 5. For a good factor solution, a particular variable should load high on one factor and low on all other factors in the rotated factor

11430

Afr. J. Bus. Manage.

Table 3. Mean, Std. deviation, corrected item-total correlation.

S/N Q1 Q2 Q3 Q4 Q5 Q6 Q7 Q8 Q9 Q10 Q11 Q12 Q13 Q14 Q15 Q16 Q17


Source: Primary data.

Mean 4.25 4.10 4.01 3.79 4.06 4.25 3.82 3.81 4.16 4.27 3.77 4.07 3.58 3.70 4.07 3.77 3.39

Std. Deviation 0.661 0.752 0.846 0.816 0.822 0.874 1.002 0.911 0.828 0.787 0.965 0.786 0.987 1.017 0.878 1.142 1.141

Corrected item-total correlation 0.534 0.605 0.666 0.466 0.645 0.577 0.544 0.677 0.571 0.587 0.676 0.558 0.705 0.556 0.584 0.485 0.374

Table 4. Kaiser-Meyer-Olkin (KMO) and Bartlett's test.

Kaiser-Meyer-Olkin measure of sampling adequacy Approx. Chi-square Bartlett's test of sphericity Degrees of freedom Significance

0.833 1025.886 55 0.000

Table 5. Rotated component matrixa.

S/N Q7 Q4 Q6 Q10 Q9 Q1 Q2 Q17 Q16 Q15

Banking Service quality items Staff of this branch is ready to apologize for a mistake Expression of genuine concern if there is a mistake in my account Staffs of this branch are neat I feel secured in my dealings with the branch staff Branch staff is able to rectify a mistake Branch staffs are willing to help me Service of the branch staff is very prompt Branch staffs are providing service requirements even after banking hours This branch has sufficient number of counters to meet service requirements Branch staff telling me about ATM and net banking facilities

1 0.793 0.723 0.699 0.603 0.600

Component 2

0.855 0.805 0.818 0.677 0.651

Extraction method: Principal component analysis. Rotation method: Varimax with Kaiser Normalizations. a. Rotation converged in 4 iterations.

matrix (Ajai and Sanjaya, 2006). As per the Table 5, it can be inferred that out of 17 of banking service quality items, 10 items are having more than 0.50 factor loadings. These 10 items were taken for further analysis.

Structural equation assessment

modeling

(SEM):

Model

fit

Structural equation modeling was used to analyze the

Renganathan et al.

11431

Table 6. Model fit indices.

Fit Indices Chi-square 2 Chi-square/degree of freedom (x /d.f.) Comparative Fit index (CFI) Goodness of Fit Index (GFI) Adjusted Goodness of Fit Index (AGFI) Normated Fit Index ( NFI) Incremental Fit Index (IFI) Tucker Lewis Index (TLI) Root mean square error of approximation (RMSEA) Parsimony goodness-of-fit index (PGFI)

Results 72.829 (0.000) DF- 29 2.51 0.947 0.955 0.915 0.917 0.948 0.918 0.071 0.50

Suggested values P-value >0.05 5.00 ( Hair et al., 1998) >0.90 (Hu and Bentler, 1999) >0.90 ( Hair et al. 2006) > 0.90 (Daire et al., 2008) 0.90 (Hu and Bentler, 1999) Approaches 1 0.90 ( Hair et al., 1998) < 0.08 ( Hair et al., 2006) Within 0.5 (Mulaik et al., 1989)

suitability of the model based upon the collected samples. As recommended by Anderson and Gerbing (1988), measurement model to test the reliability and validity of the survey instrument was analyzed first, and by using AMOS version 16 the structural model was analyzed. The structural equation model (SEM) is most useful when assessing the causal relationship between variables as well as verifying the compatibility of the model used (Peter, 2011). Structural equation modeling evaluates whether the data fit a theoretical model. In order to evaluate the model, emphasis was given to Chi-square/degrees of 2 freedom (x /df), CFI, GFI, AGFI, TLI, IFI, RMSEA and PGFI (Table 6). As per the result, Chi square statistics with p = 0.000 does not show a good fit of the model. Nevertheless according to Schumaker and Lomax (1996), a sample size of over 200 (300 in this research), could affect Chi-Square statistics to indicate a significant probability level (p=0.00). Consequently, this model is considered for further interpretation in the goodness of fit measures. Common model-fit measures like chi2 square/degree of freedom (x /df), the comparative fit index (CFI), root mean square error of approximation (RMSEA), the normed fit index (NFI), incremental fit index (IFI), and the Tucker Lewis index (TLI) were used to estimate the measurement model fit. Table 6 shows the estimates of the model fit indices from AMOS structural modeling. According to Gerbing and Anderson (1992), the criteria for an acceptable model are as follows: RMSEA of 0.08 or lower; CFI of 0.90 or higher; and NFI of 0.90 or higher. The fit between the data and the proposed measurement model can be tested with a chi-square goodness-to-fit (GFI) test where the probability is greater than or equal to 0.9 indicates a good fit (Hu and Bentler, 1999). The GFI of this study was 0.955 more than the recommended value of 0.90 the other measures fitted satisfactorily; AGFI=0.915, CFI=0.947, TLI=0.918, IFI=0.948 and 2 NFI=0.917 with x /df < 3 at 2.51 and RMSEA=0.07 (Bagozzi and Yi, 1988) indicate a good absolute fit of the model. Goodness of fit indices support the model fit and

these emphasized indices indicate the acceptability of this structural model. For the purpose of testing the model fit null hypothesis and alternative hypothesis are framed.

Hypothesis Null hypothesis (H0): The hypothesized model has a good fit. Alternate hypothesis (H1): The hypothesized model does not have a good fit. As per the Table 5, it is clear that values of all the items are above the suggested value of 0.5 (Hair et al., 2006). According to Bollen (1989a), the higher the probability associated with Chi-square, the closer the fit between the hypothesized model and the perfect fit. The test of our null hypothesis H0, that BANKPERF is a three-factor structure as shown in Figure 1, yielded a chi-square value of 72.829 with 29 degrees of freedom and a probability of less than 0.0001(p < 0.0001). It is suggesting that the fit of the data to the hypothesized model is not entirely adequate. As per the result, Chi square statistics with p = 0.000 does not show a good fit of the model. Nevertheless, according to Schumaker and Lomax (1996), a sample size of over 200 (300 in this study), could affect Chi-square statistics to indicate a significant probability level (p=0.00). Consequently, this model is considered for further interpretation in the goodness of fit measures. According to Barbara (2009), both the sensitivity of the Likelihood ratio test to sample size and its basis on the chi-square distribution, which assumes that the population (that is, H0 is correct), have led to problems of fit are now widely known. According to Jreskog and Srbom (1993), chi-square statistic equals (N-1) Fmin,, (sample size-1, multiplied by the minimum fit function) this value tends to be substantial when the model does not hold and when sample size is large. Barbara (2009) stated that, researchers have addressed the chi-square

11432

Afr. J. Bus. Manage.

0.45

e5
0.41

Apology (Q7) Concern (Q4) Neatness (Q6) Security (Q10) Mistake (Q9)
0.21 1.13 1.00 1.00 1.52 0.24 1.02

e4
0.45

e3
-0.09 0.38

Staffs conduct and Credibility

e2
0.10 -0.12 0.44

e1

0.22

e7 e6
0.15

1 1

Help (Q1)
Prompt (Q2)

0.41 0.73 1.00

0.19

Staffs conduct

0.19 0.87 1 1 1

e10
0.87

After banking Hours (Q17) Number of Counters (Q16) ATM& (Q15) Net banking

0.33 1.14 1.15 1.00

e9
0.44

e8

Customers accessibility & Communication c

Figure 1. BANKPERF structural model -confirmatory factor analysis.

Table 7. Regression weights: (group number 1 - default model).

S/N Q9 Q10 Q6 Q4 Q7 Q2 Q1 Q15 Q16 Q17

<--<--<--<--<--<--<--<--<--<---

Factor Factor 1 Factor 1 Factor 1 Factor 1 Factor 1 Factor 2 Factor 2 Factor 3 Factor 3 Factor 3

Estimate 1.000 1.003 1.133 1.021 1.516 1.000 .733 1.000 1.146 1.143

S.E. 0.104 0.135 0.125 0.172 0.083 0.166 0.166

C.R. 9.655 8.409 8.204 8.837 8.851 6.895 6.886

P *** *** *** *** *** *** ***

Label par_1 par_2 par_3 par_4 par_6 par_10 par_11

limitations by developing goodness-of-fit indices that take a more practical approach to the evaluation process. Hair et al. (1998) suggested the value for the fit statistic minimum discrepancy/degrees of freedom (CMIN/DF), otherwise chi-square/ degrees of freedom as 5. As per the Table 6, the value for the chi-square/degrees of freedom is 2.51 which is less than the accepted cut off value of 5.

Significance tests of individual parameters Table 7 shows the unstandardized coefficients and associated test statistics. The amount of change in the dependent or mediating variable for each one unit change in the variable predicting it is symbolized by the unstandardized regression coefficient. The Table 7 shows the unstandardized estimate, its standard error

Renganathan et al.

11433

Table 8. Standardized regression weights: (group number 1 - default model).

S/N Q9 Q10 Q6 Q4 Q7 Q2 Q1 Q15 Q16 Q17

<--<--<--<--<--<--<--<--<--<---

Factor Factor 1 Factor 1 Factor 1 Factor 1 Factor 1 Factor 2 Factor 2 Factor 3 Factor 3 Factor 3

Estimate 0.593 0.626 0.637 0.614 0.743 0.854 0.712 0.654 0.577 0.576

(GFI) formed by Jreskog and Sorbom, (1993) is able to calculate the proportion of variance (Tabachnick and Fidell, 2007). Model can be evaluated with the help of Normed fit index by means of comparing the Chi-square value of the model with Chi-square of the null model (Bentler and Bonnet, 1980). CFI is important in all SEM programs because its measure is least affected by sample size (Fan et al., 1999). According to McDonald and Ho (2002), CFI, GFI, NFI and the NNFI are the most frequently used fit indices in structural equation modeling. CONCLUSION AND IMPLICATIONS The aim of this research was to carry out an empirical analysis of the factors determining the bank customers perception (SERVPERF) about the four factors-banks staff conduct, credibility, communication and customer accessibility pertaining to BANKSERV model, using a structural equation modeling. This study affirms and develops an instrument of service quality in the context of banking industry, and examines the relationship among banking service quality, banks staff conduct, credibility, communication and customer accessibility. The proposed model (BANKSERV scale was adopted with some modifications) is then calibrated using the data collected from customers of banking sector in Tamilnadu and Pondicherry of India. Four significant determinants of banking service quality identified are: staff conduct, credibility, communication and customer accessibility. The findings show that Cronbachs alpha for all dimensions except for customer accessibility are above 0.70 which indicates a high level of internal consistency for the BANKPERF scale (BANKSERV model with performance only measure). Moreover, over all Cronbachs alpha value for the BANKSERV dimension is 0.904 which is above the cut off value of 0.7. Based on the confirmatory factor analysis, it can be concluded that, the BANKPERF scale (BANKSERV model with performance only measure) used in this study adequately fit into the collected data. It could be very well concluded that the hypothesized three-factor model fits the sample data. Based on the viability and statistical significance of important parameter estimates; the considerably good fit of the model (CFI, GFI, AGFI, NFI, IFI, TLI, RMSEA), it can be concluded that the three-factor model shown in Figure 1 represents an adequate description of BANKPERF structure for the banking industry goodness of fit indices support the model fit and these emphasized indices indicate the acceptability of this structural model. Definitely, this study will be useful for the bankers to ascertain the importance given by the customers for the various important factors pertaining to banking service quality. In the present competitive environment, to retain the existing customers and to enhance customer base, banks have to conduct survey at least once in three

(abbreviated S.E.), and the estimate divided by the standard error (abbreviated C.R. for Critical Ratio). Under the column P, the probability value associated with the null hypothesis that the test is zero is exhibited.

Level of significance for regression weight The probability of getting a critical ratio as large as 9.655 in absolute value is less than 0.001. In other words, the regression weight for Factor 1 in the prediction of Q10 is significantly different from zero at the 0.001 level (twotailed). The probability of getting a critical ratio as large as 8.409 in absolute value is less than 0.001. In other words, the regression weight for Factor 1 in the prediction of Q6 is significantly different from zero at the 0.001 level (two-tailed). For large samples under suitable assumptions, these statements are approximately correct.

Scalar estimates (group number 1 - default model) Maximum likelihood estimates Table 8 shows the standardized estimates for the fitted model. Relative contributions of each predictor variable to each outcome variable can be evaluated by standardized estimates. Figure 1 shows the BANKSERV structural model. Out of 17 banking service quality items, 10 items were taken for confirmatory factor analysis. As per Figure 1, it is clear that customers attach more values to staff conduct compares to other service quality items. Confirmatory factor analysis is furthermore known as measurement model. The root mean square error of approximation enlightens us how the model, with unknown parameter estimates would fit the population covariance matrix (Byrne, 1998). According to Kline (2005), CFI, RMSEA can be utilized along with ChiSquare test to calculate the measurement model fit. As an alternative to Chi-square test, goodness-of-fit statistic

11434

Afr. J. Bus. Manage.

months with bank customers to elicit their opinion pertaining to various factors leading to service quality and satisfaction.

LIMITATIONS AND FUTURE RESEARCH This study has got its own limitations. Since this study was confined only to nationalized banks in India, the findings of the study may not be fully applicable to private and foreign banks in India. Further, the study focuses only banking industry; hence generalization may not be applicable to other industries. Without any discrimination opinions expressed by customers of various branches of nationalized banks spread over Tamilnadu and Pondicherry, India, were considered as such. The determinants of the service quality may be investigated across multi-national settings. Further research can be carried out in the banking industry, especially in the globalized environment from the perspective of different determinants of service quality which will facilitate to enhance the quality of service and customer satisfaction.
REFERENCES Ajai GS, Sanjaya GS (2006). Statistical methods for practice and research, A guide to data analysis using SPSS, Response books, New Delhi. p. 143. Anderson JG, Gerbing DW (1988). Structural Equation Modeling in Practice; a review two step approach. Psychol. Bull. 103:411- 423. Asubonteng P, McCleary KJ, Swan JE (1996). SERVQUAL revisited: a critical review of service quality. J. Serv. Mark. 6(6):62-81. Avkiran NK (1994). Developing an Instrument to Measure Customer Service Quality in Branch Banking. Int. J. Bank Mark. 12(6):10-18. Avkiran NK (1999). Quality Customer Service Demands Human Contact. Int. J. Bank Mark. 17(2):61-71. Babakus E, Boller GW (1992). An Empirical Assessment of the SERVQUAL Scale. J. Bus. Res. 24(3):253-268. Bagozzi RP, Yi Y (1988). On the evaluation of structural equation models. J. Acad. Mark. Sci. 16(1):74-94. Barbara MB (2009). Structural Equation Modeling with AMOS. Basic concepts, applications, and programming. 2nd Edn, Routledge, Taylor & Francis Group, New York. pp. 76-84. Bentler PM, Bonnet DC (1980). Significance Tests and Goodness of Fit in the Analysis of Covariance Structures, Psychol. Bull. 88(3):588606. Bloemer J, De Ruyter K, Peters P (1998). Investigating Drivers of Bank Loyalty: Complex Relationship between Image, Service Quality and Satisfaction. Int. J. Bank Mark. 16(7):276-286. Bollen KA (1989a).Structural equations with latent variables, New York: Wiley. Buttle F (1996). Relationship marketing. In: Buttle, F. (Ed.). Relationship Marketing: Theory and Practice. Paul Chapman Publishing, London pp. 116. Byrne BM (1998). Structural Equation Modeling with LISREL, PRELIS and SIMPLIS: Basic concepts, applications and programming. Mahwah, New Jersey: Lawrence Erlbaum Associates. Campbell TC, Gillaspy JA, Thompson B (1995). The factor structure of the Bem Sex- Role Inventory (BSRI); A Confirmatory Factor Analysis. Paper presented at the annual meeting of the Southwest Educational Research Association, Dallas,6. Christopher L, Jochen W, Jayanta C (2006). Marketing of services, people, technology, strategy, Pearson education, 393. Cronin J, Taylor A (1992). Measuring Service Quality: A Reexamination and Extension. J. Mark. 6:55-68.

Cronin JJ, Taylor SA (1994). SERVPERF versus SERVQUAL: Reconciling performance-based and perceptions-minus-expectations measurement of service quality. J. Mark. 58(1):125-131. Daire H, Joseph C, Michael RM (2008). Structural Equation Modeling: Guidelines for Determining Model Fit. Electron. J. Bus. Res. Methods 6(1):53-60. Diana DS (2006). Exploratory or Confirmatory Factor Analysis Paper No. 200-31, SUGI 31 Proceedings, San Francisco, California March 26-29. Fan X, Thompson B, Wang L (1999). Effects of Sample Size, Estimation Methods, and Model Specification on Structural Equation Modeling Fit Indexes. Struct. Equ. Model. 6(1):56-83. Gale TB (1990). The role of marketing in total quality management. In: QUIS-2 Quality in Services Conference Proceedings, University of St. Johns. Minnesota, 5 November 1990. George D, Mallery P (2003). SPSS for Windows step by step: A simple guide and reference. 11.0 update, 4th edn., Boston: Allyn & Bacon, 231. Gerbing DW, Anderson JC (1992). Monte Carlo evaluations of goodness of fit indices for structural equation models. Sociol. Methods Res. 21(2):132160. Hair JF, Anderson RE, Tatham RL, Black WC (1998), Multivariate Data Analysis, Prentice- Hall, Upper Saddle River, New Jersey. In: Marcin Pont and Lisa McQuilken (2002). Testing the Fit of the BANKSERV Model to BANKPERF Data. ANZMAG conference proceedings.865. Hair JF, Anderson RE, Tantham RL (2006). Multivariate Data Analysis.10th edn., Prentice Hall: New Jersey. In: Malek ALMajali, Nik Kamariah Nik Mat (2011). Modeling the antecedents of internet banking service adoption (IBSA) in Jordan: A Structural Equation Modeling (SEM) approach. Journal of Internet Banking and Commerce. 16(1):8-13. Hair JF, Anderson RE, Tatham RL, Black WC, Babin BJ (2006). Multivariate Data Analysis, 6th edn., Pearson Education, New Delhi. 734-735 Hu LT, Bentler PM (1999). Cutoff Criteria for Fit Indexes in Covariance Structure Analysis: Conventional Criteria versus New Alternatives, Struct. Equ. Model. 6(1):1-55. Johns N (1996). The developing role of quality in the hospitality industry, Service Quality in Hospitality Organization, editors Olsen, D. Teare, R. Gummesson. E, Cassell, New York. pp. 9 26. Jreskog K, Srbom D (1993). LISREL 7: Users Reference Guide. Chicago, IL: Scientific Software International Inc. In: Barbara M Byrne, Structural equation modeling with AMOS, Routledge,Taylor Francis. 2. 76-77 Joseph A Gliem, Rosemary R Gliem (2003). The Midwest research-topractice conference in Adult, continuing, and community education, The Ohio State University, Columbus, OH, October 8-10. Kline RB (2005). Principle and practice of Structural equation modeling. 2nd Edn. In: Hooper D, Coughlan J,Mullen M R (2008).Structural Equation Modelling: Guidelines for Determining Model Fit. The Electronic Journal of Business Research Methods. 6 (1): 53 - 60 Lam SS (1995). Assessing the Validity of SERVQUAL: An Empirical Analysis in Hong Kong, Asia Pac. J. Qual. Manage. 4(4):33-40. Mass HAPM, Noor AA, Alias R (2010). Validating SERVPERF Model in Government Agencies. J. Hum. Resourc. Adult Learn. 6(1):5 McDonald RP, Ho MHR (2002). Principles and Practice in Reporting Statistical Equation Analyses. Psychol. Methods 7(1):64-82. Mulaik SA, James LR, Van Alstine J, Bennet N, Lind S, Stilwell CD (1989). Evaluation of Goodness-of-Fit Indices for Structural Equation Models. Psychol. Bull. 105(3):430-445. Parasuraman A, Zeithaml VA. Berry LL (1985). A conceptual model of service quality and its implications for future research, Journal of Marketing. 49 (Fall), 41-50 Parasuraman A, Zeithaml VA. Berry LL (1988). A multiple item scale for measuring consumer perceptions of service quality, Journal of retailing, 64(1): 12-40. Peter T (2011). Adoption of Mobile money technology: Structural Equation Modeling Approach. Eur. J. Bus. Manage. 3(7):2011. Robledo MA (2001). Measuring and managing service quality: integrating customer expectations. Manage. Serv. Qual. 11:22-31. Rust RT, Zahorik AJ (1993). Customer satisfaction, customer retention

Renganathan et al.

11435

and market Share. J. Retailing 69(2):193-215. Saxe R, Weitz BA (1982). The SOCO Scale: A Measure of the Customer Orientation of Salespeople. J. Mark. Res. pp. 343-351. Schumaker RE, Lomax RG (1996). A Beginners Guide to Structural Equation Modeling, Lawrence Erlbaum Associates, Mahwah, NJ. Stafford M, Stafford TF, Wells BP (1998). Determinants of Customer Satisfaction in the Auto Casualty Claims Process. J. Serv. Mark. 12(6):426-460. Tabachnick BG, Fidell LS (2007). Using Multivariate Statistics, 5th Edn. New York: Allyn and Bacon. In:Hooper D,Coughlan J, Mullen M R (2008). Structural Equation Modelling: Guidelines for Determining Model Fit. The Electronic J. Bus. Res. Methods 6(1):53-60

Yu-Kai H (2009). The Effect of Airline Service Quality on Passengers Behavioral Intentions Using SERVQUAL Scores, A TAIWAN Case Study. J. Eastern Asia Soc. Transport. Stud. 8:4-5. Zeithaml VA, Parasuraman A, Berry LL (1990). Delivering Quality Service: Balancing Consumer Perceptions and Expectations. The Free Press, New York, NY.

11436

Afr. J. Bus. Manage.

Appendix 1. Questionnaire: Service quality in Banking Industry (Avkiran, 1994: 15) **with some modifications.

S/N

Banking service quality items Dimension 1: Staff conduct Branch staffs are willing to help me Service of the branch staff is very prompt Branch staffs are courteous Expression of genuine concern if there is a mistake in my account Staffs of this branch are facilitators Staffs of this branch are neat Staff of this branch is ready to apologize for a mistake Dimension 2: Credibility Branch staff keeping me informed about matters of concern to me Branch staff is able to rectify a mistake I feel secured in my dealings with the branch staff Dimension 3: Communication Branch staff helping me, to learn how to keep down my banking costs Branch staffs are knowledgeable about banks services and products Branch staffs give quality advice about managing my finances Branch staff telling me about the different types of accounts and investments available Branch staff telling me about ATM and net banking facilities Dimension 4: Customer accessibility ** This branch has sufficient number of counters to meet service requirements Branch staffs are providing service requirements even after banking hours

Strongly agree (5)

Agree (4)

Neither agree nor disagree (3)

Disagree (2)

Strongly disagree (1)

Q1 Q2 Q3 Q4 Q5 Q6 Q7

Q8 Q9 Q10

Q11 Q12 Q13 Q14 Q15

Q16 Q17

S-ar putea să vă placă și