Académique Documents
Professionnel Documents
Culture Documents
Abstract : The purpose of the study is for the represent five service quality dimensions known as
improvement of the retail service quality of grocery ‘tangibles’, ‘reliability’, ‘responsiveness’,
stores in five major cities in Tamilnadu, India. It ‘assurance’ and ‘empathy’.
covers four retail formats. The grocery stores are The SERVQUAL scale has been tested and
normally open around 8.30 am in the morning. adapted in a large number of researches conducted in
Sometimes when the respondent needs to purchase various service settings, cultural contexts and
during the opening hours, they need to wait as it geographic locations and it is suggested that they
takes some time to setup. First, this study aims to require customization to the specific service sector in
calculate service quality score of the individual which they are applied [4]. The researcher has
respondent and based on the SERVQUAL customized for Grocery retailing. After finding the
dimensions, Then study aims at predicting the SERVQUAL, the researcher aims at predicting the
customer’s willingness to wait, to buy grocery with customer willingness to wait, to buy grocery with the
the contribution of demographics and SERVQUAL contribution of demographics and servqual score.
score using binary logistic regression. There are 7
predictors. The important factor that contributes to 2. Literature Review
the prediction is identified along with the group in
specific. A Few suggestions are given in the It is a universal fact that customer satisfaction is
conclusion to build customer relationship and the base for repeated purchase. For example, when
ultimately improve the service quality. customers perceive a good product or service, each
one of them will typically share their experience few
people in their circle. It is a fact that nearly one-half
1. Introduction
of Indian’s business is built upon this informal,
Measuring service quality is an interesting task “word-of-mouth” communication. Improvement in
because of its distinctive characteristics: customer retention by even a few percentage points
Intangibility, heterogeneity, inseparability and can increase profits by 25 percent or more [5]. The
perishability (Buttle, 1996) [1]. Service quality is University of Michigan found that for every
about the concepts of perceptions and expectations of percentage increase in customer satisfaction, there is
service (Parasuraman et al., 1985, 1988) [2] . an average increase of 2.37% of return on investment
Customers’ perceptions of service are the result of a [6]. Most people praise the businesses that service
comparison of the expectations of services with the them the way they like to be serviced; they will even
actual-service experience. When perceptions exceed pay more for that service.
expectations the service has resulted in consumer
surplus; it is adequate, when it just equals with the
expectations; the service will be said as bad, poor or The message is obvious finally that the
insufficient, when it does not meet their expectations customer’s willingness to wait is an indication that
and has led to customer frustration (Gitomer, 1998) the customers are satisfied and customers not willing
[3]. to wait may not be loyal to the store at present and
Based on this perspective, Parasuraman et al. chances are there that they might also change.
developed a scale for measuring service quality, Customer satisfaction is an asset that should be
which is well known as SERVQUAL. This scale monitored and managed just like any physical asset.
operationalises service quality by calculating the This is true for both service-oriented and product-
difference between Expectations and Experiences of oriented organizations [7].
the respondents in five major places in Tamilnadu,
evaluating both in relation to the 22 items that
5. Research Objectives
In accordance with the study, the research
encompasses the following objective:
multiplied by the weight (converted to 1), to get the Table 2: Model Summary
weighted average score in the construct. Then, -2Log Cox & Snell Nagelkerke
calculated the average of “Weighted Average of the Step likelihood R Square R Square
Construct” to arrive the SERVQUAL score of - 1 633.246a .113 .156
0.2877. This score says that the expectation of the a. Estimation terminated at iteration number 4 because
service quality is more than their experience while parameter estimates changed by less than .001.
shopping grocery. The SERVQUAL score is not
highly negative (-0.2877, the maximum negative The Table2 provides us with the R square
score is -7) which means that the retail service statistics and there are two measures, Cox & Snell
quality is close to normal. [9] and Nagelkerke. Both these uses a little different
formula, but both are equally valid. In this case, Cox
Table SA2: Analysis of SERVQUAL Score - Individual Respondent & Snell is 0.113, and Nagelkerke is 0.156. These
Sl.No Servqual score of a No. of Respondent numbers indicate scope for improvement in fit over
. Respondent Respondent s (%)
the baseline model. The value shows that there is
1 -1 to -0.75 32 5.90%
little scope to contribute to the improvement of the
2 -0.75 to -0.5 109 20.11% fit over the baseline model.
3 -0.5 to -0.25 142 26.20%
Table 3: Hosmer and Lemeshow Test
4 -0.25 to 0 177 32.66% Step Chi-square df Sig.
5 0 to 0.25 75 13.84% 1 5.755 8 .675
6 0.25 to 0.5 7 1.29%
In Table3, A chi-square statistic is computed by
Grand Total 542 100.00% comparing the observed frequencies with those
expected under the linear model. The significance
The above table SA2 shows the servqual score of value is 0.675. A nonsignificant chi-square indicates
the respondent. 5.90% of the Respondents scored that the data fit the model the best. This test is also to
between -1 to 0.75, 20.11% of the respondent scored know the goodness of fit and since the significance
between -0.75 to -0.5, 26.20% of the respondent value is 0.675, it fits the Best.
scored between -0.5 to -0.25, 32.66% of the The below table4 helps us to predict the
respondent scored between -0.25 to 0, 13.84% of the covariates that aids the prediction of the respondents
respondent scored between 0 to 0.25, 1.29% of the those who have the intention to wait to shop grocery
respondents scored between 0.25 to 0.5. when the shop is yet to open. The equation to find
the probability value to predict is as below. The “B”
7.2. Prediction - Who Will Wait to Shop values refer to the log-odds of being willingness to
Grocery wait, to buy grocery during the opening hour. We
can insert these into the binary logistic regression
The below is the summary and discussion of the equation as below:
results of binary logistic regression.
The below is the summary and discussion of the Binary Logistic Regression= α + β1X1 + β2X2 + ….+
results of binary logistic regression. β7X7
= -1.905 + (-0.172) (city)
Table 1: Omnibus Tests of Model Coefficients + 0.534 (gender)
Chi-square df Sig. + 0.584 (education)
Step 1 Step 65.184 16 .000
Block 65.184 16 .000
+ (-1.192) (occupation)
Model 65.184 16 .000 +0.141 (spending)
+ 0.065 (saving)
The Table1 here gives us the ‘omnibus test of + 0.245 (SERVQUAL)
model coefficients’. These values give us an idea of
whether or not the model with our independent Table 4: Variables in the Equation
B S.E. Wald df Sig. Exp(B)
variables fits the data better (i.e. Gives us a better Step City .772 4 .942
prediction of individual scores) than the baseline 1a City (1) -.029 .308 .009 1 .926 .972
model. We can find significance in the final column City (2) -.143 .310 .212 1 .645 .867
on this table 0.000, and this means (significance is City (3) .107 .340 .098 1 .754 1.112
less than .05) the model is significant, which means City (4) -.172 .305 .318 1 .573 .842
Gender(1) .534 .255 4.380 1 .036 1.706
that this model with the 7 predictors is a PERFECT Education 11.338 3 .010
Fit. It shows the goodness of fit [10]. Education(1) 1.044 .326 10.260 1 .001 2.840
Education(2) .699 .276 6.440 1 .011 2.012
Education(3) .584 .286 4.167 1 .041 1.793
Occupation 15.355 5 .009
Occupation(1) -.316 .425 .553 1 .457 .729 a. In predicting the factors that contribute to
Occupation(2) -1.378 .515 7.170 1 .007 .252 the prediction of the intention to wait, from
Occupation(3) -.955 .369 6.696 1 .010 .385
Occupation(4) -.480 .449 1.144 1 .285 .619
the table4 the researcher found that Gender,
Occupation(5) -1.192 .451 6.992 1 .008 .304 Education, Occupation and SERVQUAL
spending on .141 .110 1.621 1 .203 1.151 score are the factor that contributes to the
grocery decision on the intention to wait is correct in
Home savings .065 .132 .245 1 .621 1.067 reality too.
SERVQUAL .245 .094 6.758 1 .009 1.278
Constant -1.905 .608 9.827 1 .002 .149
b. Predicted that among 105 respondents, 88
a.Variable(s) entered on step 1: City in which you have GROCERY respondents are male and 17 respondents
shopping xperience, Gender, Education, Occupation, spending in rocery, are female. So males have the chance to
Home savings, SERVQUALic.
wait more than female. (Appendix A)
c. Predicted that among 105 respondents, 88
Among the 7 predictors in table 4, the variable respondents are graduates and below. This
Gender, Education, Occupation and SERVQUAL shows that the higher educated mostly do
score in table4 alone has significance below 0.05 not wait. So when our customers are of
hence these variables alone contributes to the model. higher education, we should understand that
Place, monthly savings and monthly spending on they do not have time or patience to wait to
Grocery do not contribute to this model at large. shop and look for alternatives. (Appendix
B)
8. Discussion and Conclusion d. Predicted that among 105 respondents, 82
The servqual score of the majority of 32.66% of respondents are Businessman and
the respondent is between -0.25 to 0 shows that the professional, this shows that these
majority of the respondents expects an incremental professions have flexibility in time and have
care to improve the service quality. The negative patience. None of the student waits shows
SERVQUAL score is about 84.87% of respondents. that they have to be timid. (Appendix C)
It clearly shows that the expectation of the e. Predicted that among 105 respondents, 75
respondents during buying grocery is higher and respondents have SERVQUAL score
hence the retailer should put effort. Respondents between -0.25 to +0.25 and none from the
forgot that the increment in service quality is going SERVQUAL score between -0.75 and -1.
to cost more in the products they buy.
The cutoff value is 0.5 to predict whether a The conclusion is that the males, businessman,
respondent will wait or not. The respondent with a Professionals, Graduates & below and those who
probability greater than 0.5 is predicted to have the have SERVQUAL score between -0.25 to 0.25 has
intention to wait to shop grocery. the chance to wait to shop grocery. Hence need to
train the employees to focus on them better [11]. The
Table 5: Binary Logistic Regression Classification emotions of the buyer are to be valued [12] Since
Tablea these people have the patience to wait it should be
Predicted encouraged. The researcher suggests offering a token
at the entry when these people enter immediately
Will you
after opening. This token bears a discount of 4% or
wait to purchase
free grocery of their brand. This might also attract
when they are
the buyers to wait. A healthier approach is to focus
opening the
on consistently providing the dimensions and levels
store/yet to
of quality that customers expect [13].
open in the
Morning? Percentage
Observed No Yes Correct
9. References
Step Will you No 308 47 86.8 [1] F. Buttle, “SERVQUAL: review, critique,
1 wait to purchase Yes 125 62 33.2 research agenda,” European Journal on
when they are Marketing, vol. 30, no. 1, pp. 8–32, 1996.
opening the [2] A. Parasuraman, “Consumer Perceptions of
store/yet to Service Quality,” Spring, vol. 64, no. 1,
open in the 1988.
Morning? [3] J. H. Gitomer, Customer Satisfaction is
Overall Percentage 68.3 worthless, Customer loyalty is priceless:
a. The cut value is .500 How to make customers love you, keep them
coming back and tell everyone they know.
Based on the results obtained by the researcher, Austin: Bard Press, 1998.
the following are the conclusions: [4] D. W. Finn and C. W. Lamb, “An Evaluation
Total 51 11 62
Private Sector Predicte No 14
56 196
d group 0
Ye
5 8 13
s
Total 14
64 209
5
Homemaker/Hous Predicte No 48 25 73
e Wife d group Ye
4 5 9
s
Total 52 30 82
Student Predicte No
51 16 67
d group
Total 51 16 67
Professionals Predicte No 10 11 21
d group Ye
11 10 21
s
Total 21 21 42
Total Predicte No 31 12
437
d group 0 7
Ye
45 60 105
s
Total 35 18
542
5 7
Appendix D
Predicted group * Will you wait to purchase when
they are opening the store/yet to open in the
Morning? * SERVQUAL – Cross tabulation
Will you wait
to purchase
during
opening
SERVQUAL hours
Score No Yes Total
-1 to Predicted No
29 3 32
- 0.75 group
Total 29 3 32
-0.75 to Predicted No 71 21 92
-0.5 group Yes 4 13 17
Total 75 34 109
-0.5 to Predicted No 92 39 131
-0.25 group Yes 1 10 11
Total 93 49 142
-0.25 to Predicted No 96 42 138
0 group Yes 28 21 49
Total 124 63 187
0 to Predicted No 20 19 39
0.25 group Yes 12 14 26
Total 32 33 65
0.25 to Predicted No 2 3 5
0.5 group Yes 0 2 2
Total 2 5 7
Total Predicted No 310 127 437
group Yes 45 60 105
Total 355 187 542