Documente Academic
Documente Profesional
Documente Cultură
c&e((icie)" A*=: R
4
F v!ue
". )ependent variable N ISH
1onstant ..$"
Service quality O functional (SH:"&6 ..#*P ..-&.P ""#.-&
-. )ependent variable N ISH
1onstant .."$
Service quality O functional (SH:--, i.e.,
S;3A8;3:6
"."&P ..-#$P "0/.(&
Correlation Coefficients:
SH:"& P ISH N ..$(P % SH:-- P ISH N ../$P
CoteK Significance level isK P pQ...".
@he above discussion thus brings us to two important conclusions. :irstly, the five
dimension structure as proposed by the developers of S;3AH?< scale (on which S;3A8;3:
scale is based6 is not valid across service industries and countries. @his finding is quite in
"&
conformity with the results obtained by several other researchers in the past (e.g., 'abakus and
'oller, "##-% 1arman "##.% 5hou, -..$6. @he obvious inference is that the dimensionality of
service quality scale is both conte9t and country specific% and, hence, is not directly transferable
across industries and countries. Secondly, the thirteen items retained in the present study for
measuring service quality (functional aspect6 are not adequate enough to capture customers>
overall service quality perceptions. @his finding points to a pressing need for adding additional
itemsLdimensions to 1ronin and @aylor>s ("##-6 S;3A8;3: scale.
R&!e &( >Ou"c&'e3 C&'-&)e)" )* Di'e)%i&)-%-eci(ic A)!#%i% i) Pre*ic"i)0 Cu%"&'er%3
Over!! Service Qu!i"# Perce-"i&)%
Since the literature suggests that outcome quality (SHI6 is an important determinant of
perceived service quality, it was included as another independent variable in the regression equation.
@he results presented in @ables & and $ show an improvement in the ad=usted 3
-
value from .. -&.
(when SH:"& alone is used as independent variable6 to ..-*( when both SH:"& and SHI are used
as independent variables. 'oth the predictor variables are statistically significant, with outcome
component emerging as an equally important determinant. lmost similar standardised regression
coefficients of ..&- and ..-( bear a testimony to the equal importance of this variable. significant
improvement in ad=usted 3
-
value with the addition of ISH variable suggests that the outcome
component is an important determinant of service quality and needs to be taken into account while
measuring consumer perceptions of service quality in the case of fast food restaurants.
Cotwithstanding improvements in the results, the fact remains that even these two factors taken
together are not able to adequately e9plain variations present in the customers> overall service
quality perceptions - ad=usted 3
-
value
being =ust ..-*(. @his once again implies that efforts be made
in future to identify additional itemsLdimensions in respect of both the functional and outcome
service quality components so as to be able to improve the predictive capabilities of two multi-item
service quality scales.
s suggested in the literature, a dimension-specific analysis of service quality scale (SH:"&6
was undertaken along with the outcome service quality component as another e9planatory variable.
3esults corresponding to equation & are quite revealing (see @able $6. If the four functional service
quality dimensions, empathy and support are not significant e9planatory variables. 7n terms of
standardized beta coefficients, outcome quality (SHI6 emerges as the most important determinant of
service quality, closely followed by tangibility (@C6. @he variable dependence ();86 is also a
significant variable, but it trails far behind the tangibility and outcome components. @he two
collinearity statistics - @I7 and A7: - being greater than ..". and less than ". respectively imply an
"$
absence of multicollinearity among the independent variables (Hair at al., "##/6, thus adding to our
confidence in the results.
T1!e ?: Over!! Service Qu!i"# OSQ6 )* I"% Re!"i&)%+i- .i"+ Fu)c"i&)! 5SQF2<6 )*
Ou"c&'e Service Qu!i"# 5SQO6 - Re0re%%i&) Re%u!"%
E,:
N&: Vri1!e
C&e((icie)"%
A*=: R
4
F
v!ue @R
4
F
v!ue
&( @R
4
TOI
1
VIF
c
Stand-
ardised
". )ependent variable N ISH
1onstant ..--
:unctional quality (SH:"&6 ..0$P ..&-P ..0/ ".//
Iutcome quality (SHI6 ..&(P ..-(P ..-*(P **..* ...$#P -*..& ..0/ ".//
-. )ependent variable N ISH
1onstant .."- -
@angibility (@C6 ..$&P ..-#P ..*& ".&*
;mpathy (;+86 ...0 ...$ ..*& ".&(
)ependence ();86 .."$PP ..".PP ..*" ".$-
Support (S?86 -.... -...& ..(" ".-&
Iutcome quality SHI6 ..$-P ..&.P ..&./P &/.00 - - ..0- "./#
&. )ependent variable N ISH
1onstant .."- -
@angibility (@C6 ..$-P ..-(P ..(. ".-/
)ependence ();86 .."/PP ..""PP ..*$ ".&0
Iutcome quality SHI6 ..$/P ..&-P ..&.0P /#."# - - ..*( ".-#
CotesK a. Significance levels areK P pQ..." and PP 8Q.../.
b. @I7 refers to tolerance score for the independent variable.
c. A7: refers to variance inflated factor.
;mpathy and support dimensions have not only low beta values, but are also not statistically
significant. @his probably has happened due to nature of the service industry under consideration. s
already mentioned, fast food restaurants offer highly standardized service, thus leaving no or little
scope for customization. Hence, the empathy has not emerged as a significant determinant. @he other
variable Dsupport> (capturing the influence of availability of support to the employees from
management and accuracy of records6 is not only psychometrically unsound for being less valid and
reliable, but it also does not seem to be holding much relevance. vailability of support to employees
is an internal matter and is of no direct relevance to the customers in forming their service quality
perceptions. ;ven the item accuracy of records seems of little importance in the case of fast food
restaurants because all of them currently make use of cash registers and hand out printed receipts
with all the necessary details to enable the customers check the accuracy of bills, thus leaving no
scope for the accuracy to vary across the restaurants.
Since empathy and support are not found significant e9planatory variables, these were
dropped from further analysis in the study. regression analysis performed on the remaining
"/
variables (equation $6 produced results which are more or less similar to those obtained in regard to
equation &.
Service Qu!i"# % ) A)"ece*e)" &( Cu%"&'er S"i%(c"i&): A00re0"ive )* Di'e)%i&)-
%-eci(ic A)!#%i%
1ustomer satisfaction (1S6 was regressed on various variants of service quality measure to
assess the role of service quality as an antecedent of customer satisfaction perceptions. @he results
reported in @able / shows that overall service quality (ISH6 is a ma=or and significant determinant
of customer satisfaction in the fast food restaurants. n ad=usted 3
-
value of ..//" shows that the
overall service quality perceptions are able to e9plain //.". per cent of the variations present in the
customer satisfaction. Since the customers form their satisfaction =udgments on the basis of service
quality and price perceptions, the variable price equity (8;6 was also introduced in the regression
equation. @he results reveal a marginal but significant rise in the ad=usted 3
-
value. 'oth the
e9planatory variables are statistically significant, with ISH remaining as the ma=or determinant.
8rice turns out to be as a considerably less important factor.
7n order to assess the usefulness of using multi-item service quality scales in place of ISH
scale, customer satisfaction was regressed on SH:-- and SH:"& alternately along with service
outcome (SHI6 component. 3esults corresponding to equation - in @able / show a rather poor fit of
SH:-- to customer satisfaction perceptions. ;ven the use of variable SH:"& causes a further
(though marginal6 decline in ad=usted 3
-
value. 8rice equity in both the equations remains a
significant but marginal determinant. @he lower ad=usted 3
-
values in respect of both the SH:"& and
SH:-- scales imply need for improving these two scales in future.
dimension-specific analysis (substituting SH:"& by its two ma=or dimensions, viz., @C
and );86 does help improving the ad=usted 3
-
value to a level that was attained earlier (equation &6.
@aken together, the four antecedents, viz., @C, );8, SHI and 8;, are able to e9plain $..$ per cent
of variations present in the customer satisfaction perceptions. :urthermore, all the antecedents are
significant. Service outcome quality (SHI6 turns out to the most important determinant, closely
followed by tangibility dimension (@C6. @he other two variables, viz., );8 and 8;, are
individually about half as strong as the other two variables (i.e., SHI and @C6 individually are in
their impact on customer satisfaction.
@he above analysis thus shows that service quality does affect customer satisfaction, with
service outcome playing a ma=or role in the case of fast food restaurants. :urther the results show
that a dimension-specific analysis is a better alternative to link service quality to customer
"0
satisfaction by pointing out as to which of the several service quality dimensions have stronger
impact on customer satisfaction than the aggregative analysis using only the summed service quality
(SH:"&6 score is able to do. @he variable price equity, however, is found to be playing a significant
but marginal role in influencing the customer satisfaction with fast food restaurants.
T1!e 8: Service Qu!i"# % A)"ece*e)" &( Cu%"&'er S"i%(c"i&) 5CS6 - Re0re%%i&) Re%u!"%
E,:
N&: Vri1!e
C&e((icie)"%
A*=:
R
4
F
v!ue @R
4
F
v!ue
&( @R
4
TOI
1
VIF
c
B
Stand-
ardised B
". )ependent variable N 1S
1onstant ..#* -
Iverall SH (ISH6 ..0*P ..*$P ..//-P $(0.&* - - - -
-. )ependent variable N 1S
1onstant ..**
Iverall SH (ISH6 ..0$P ..*"P ..#$ "..0
8rice equity (8;6 ...#P .."&P ../0*P -/(.0- ..."*P "/..& ..#$ "..0
&. )ependent variable N 1S
1onstant -..$# -
:unctional quality (SH:--,
i.e., S;3A8;3:6
..*(P ..$-P ..// ".(.
Iutcome quality (SHI6 ..&"P ..-/P ..0. ".0(
8rice equity (8;6 .../P ...*P ..$.$P #..&* - - ..(/ "."(
$. )ependent variable N 1S
1onstant -.."0 -
:unctional quality (SH:"&6 ../#P ..&&P ..0. ".0*
Iutcome quality (SHI6 ..&#P ..&"P ..0$ "./*
8rice equity (8;6 ...*PP ...#PP ..&*0P *(.(* ..(/ "."*
/. )ependent variable N 1S
1onstant -..-$P -
@angibility (@C6 ..&/P ..-0P ..(. ".-/
)ependence ();86 .."(P .."$P ..0# ".$$
Iutcome quality SHI6 ..$$P ..&/P ..*0 ".&-
8rice equity (8;6 ...*PP ..".PP ..$.$P 00.-. - - ..($ "."#
CotesK a. Significance levels areK P pQ..." and PP 8Q.../.
b. @I7 refers to tolerance score for the independent variable.
c. A7: refers to variance inflated factor.
Service Qu!i"# )* Cu%"&'er S"i%(c"i&) % Pre*ic"&r% &( Be+vi&ur! I)"e)"i&)%:
A00re0"ive )* Di'e)%i&) S-eci(ic A)!#%e%
Higher service quality has been empirically linked to higher satisfaction and favourable
behavioural intentions both directly as well as interactively with customer satisfaction in the past
studies reviewed earlier. @wo sub-constructs used for tapping the behavioural intentions in the
present study includeK restaurant patronisation intentions (876 and recommendation intentions (376.
7n order to assess whether service quality and customer satisfaction affect behavioural intentions in a
curvilinear and interactive manner, higher order regression equations with provision for interactions
between service quality and customer satisfaction were run. However, all the quadratic regression
"*
equations as well as those with interaction terms did not turn out to be producing satisfactory fits due
to high collinearity present among the higher order and cross products of direct measures of
independent variables. Hence, it was decided to use only the first order basic regression equations.
@he results corresponding to equation " in @able 0 show a significant and strong influence of
customers> overall service quality (ISH6 perceptions on their restaurant patronization intentions. 'ut
with the inclusion of customer satisfaction as another independent variable, ad=usted 3
-
value gets
significantly increased to ..&&., and both the ISH and 1S emerge as significant predictors.
However, when ISH is replaced by SH:-- - its twenty-two multi-item counterpart% the results turn
out to be poorer. Cot only there is a fall in the value of ad=usted 3
-
, the variable SH:-- also becomes
insignificant suggesting albeit erroneously that functional service quality (i.e., S;3A8;3:6 is not a
determinant of customers> patronization intentions. @his probably is occurring due to poor validity of
the scale pointed out earlier in connection with the factor analysis undertaken in the study. @he
results, however, show an improvement when the variable SH:"& is used instead of SH:--.
dimension-specific analysis of functional service quality as undertaken in equation /
brings the results closer to the ones obtained earlier in equation & where ISH has been used as a
measure of functional service quality. value of ..&$- of ad=usted 3
-
suggests that the four variables
taken together (viz., @C, );8, SHI and 1S6 are able to e9plain &$.- per cent of variations in
customer patronization intentions, with customer satisfaction being the most important and
significant determinant. @he three service quality related variables O @C, );8 and SHI - are also
significant, but these are individually almost half as important as the customer satisfaction variable
alone is (see @able 06. lack of multicollinearity among the independent variables as evident from
@I7 value being higher than ..". and A7: value being less than ". further add to the reliability of
regression results arrived at our end.
3esults relating to customer restaurant recommendation intentions appear almost similar to
those obtained in regard to customer patronization intentions patter (see @able *6. Iverall service
quality perceptions constitute an important determinant of the recommendation intentions, but the fit
gets significantly improved with the inclusion of customer satisfaction as another independent
variable. 'etween the two multi-item functional service quality scales, SH:"& is providing better
results than SH:--. Iutcome service quality is a significant and ma=or determinant of the
recommendation intentions. dimension-specific analysis of functional service quality shows that
that all the four variables present in equation / are significant, with 1S being the most important
determinant, followed by SHI and @C in that order. );8, i.e., dependence, turns out to be the least
important determinant.
"(
T1!e A: Service Qu!i"# )* Cu%"&'er S"i%(c"i&) 5CS6 % Pre*ic"&r% &( Re%"ur)"
P"r&)i%"i&) I)"e)"i&)% 5PI6 - Re0re%%i&) Re%u!"%
E,:
N&: Vri1!e
C&e((icie)"%
A*=:
R
4
F
v!ue @R
4
F
v!ue
&( @R
4
TOI
1
VIF
c
S")*-
r*i%e*
A*=:
R
4
F
v!ue @R
4
F
v!ue
&( @R
4
TOI
1
VIF
c
S")*-
r*i%e*