Professional Documents
Culture Documents
3.1 Introduction
The research design is based on “philosophy of science” which is established as per the
research question (Lawrence & Lawrence, 2000; Nunnally & Bernstein, 1995). Entire
methodology the philosophy, quantitative methodology, the data collection and analysis
method are originated by the research questions.
Chapter includes the research design, good reason for the usage of quantitative method,
procedure for the data collection and the statistical technique for data analysis. Problems
and solution related to “validity and reliability”, “sample selection”, “measure
development” discussed here.
1
Reliability and Validity
Reliability
To carry out reliability test all construct necessarily have two to three indicators (Joseph,
Ronald & Black, 1998). All construct are developed upon between 4 to 10 items, meeting
the criteria. Pilot testing is performed to measure the reliability by analyzing correlation
among the constructs. Results are satisfactory reliable when item score value is more 0.60
at p<0.05 level of significance.(Nunnally & Bernstein, 1995).
To make research more reliable we take care of systematic reliability measurement
errors. It is categorized into ”sample, transmittal, response, or analysis error” (Green &
Tull, 1988). Sample error is no comparability of the sample with the population(Green &
Tull, 1988). It is overcome by selecting the most renowned apparel brand J.J which is
equally famous among males and females of urban Pakistan. Transmitter error includes
the components of the instrument, involvement of the researcher and respondent’s
response(Green & Tull, 1988). It is minimized by designing and formatting the
questionnaire in a consistent way and used easily understandable words to reduce the
confusion. Response error is reduce by less involvement of researcher, as we used mail
questionnaire that submitted by the respondent thus reduce researcher biasness. Analysis
error occurred by incorrect coding or data entry(Green & Tull, 1988). It is overcome by
designing close ended questionnaire on likert scale.
Validity
Validity makes sure how a well–founded scale measures the construct what it claims to
measure (Green & Tull, 1988). It is the significant measuring tool to determine the
precision of research to overcome the chances of error.
Validity is of three types namely “content, construct and criterion validity” (Neuman
1997).Content validity determines the significance of calculated research sample to
denote the population of overall “items of the construct” (Nunnelly & Bernstain, 1994).
To overcome it, this research used scales presented in earlier researches.
Construct validity is about validating the association “among the items in the
measurement scale”, this association among items must be highly correlated in te scale
(Green, Tull and Albaum 1988). Construct validity are of two kind, convergent validity
and divergent validity (McColl-Kennedy & Fetter, 2001).Convergent validity discover
the relationship among all construct’s items of “perceived risk, brand Involvement,
customer satisfaction, brand trust, brand commitment and attitudinal brand loyalty”
where as Divergent discover negative or no relationship among the items of constructs
(Neuman 1997). For the pilot testing different scholar gave different suggestions,
(Hertzog, 2008) suggested that there should be no straight measurement for the sample.
Michael(1995) suggested that 10–30 respondent can fulfill the criteria; (Connelly, 2008;
Treece, 1982) recommended 10% of sample.
Criterion validity compared the resulted scale with “external measure of the same
construct” (Neuman 1997). Since social sciences constructs are mostly conceptual, it will
2
be hard to discover an appropriate “external measure” (Nunnally and Bernstein 1994),
hence it is not examined.
3
3.6 Scale and Measurement
This questionnaire of this study is designed on the base of research objective(Aaker,
Kumar, & Day, 1998). Collected data is based on “opinion, beliefs and attitude” of the
respondents,. Positive questions are asked which are applicable for all, avoided the bias
question (Cooper & Schindler, 1998; Zikmund, 1994). Survey questions are mostly close
ended which allow to be similar for all the respondents, reduces not only interviewer
“biasness” but also decreases thinking effort for the respondents(Hair, Bush, & Ortinau,
2000). Scale of measurement is included for the respondent “Point of agreement”(Alrack
& Settle, 1995). Likert scale of five categories is designed for the survey question.
Summarized results are illustrated in Table2.
3.8 Summary
Chapter of this research has along with the discussion given reasoning for using the
design approach and validity and reliability of the data. Online and offline mail survey is
used to verify a theoretical model of brand loyalty. All the relationships are taken the
4
previous studies.
5
CHAPTER 4: DATA RESULTS
Results
This chapter will discuss the outcome of Descriptive analysis, Correlation and Regression
and mediating Analysis.
Descriptive Statistics
Cronbach
Alpha Mean Std. Deviation Skewness Kurtosis
Brand Commitment 0.947 2.6323 .96006 0.94 .247
Brand Trust 0.921 2.2774 .78134 1.7 .247
Customer Satisfaction 0.947 2.4111 .89414 2.25 .247
Brand Involvement 0.913 2.8010 1.00335 -0.55 .247
Brand Perceived Risk 890 3.3196 .81026 -0.56 .247
Attitudinal Brand 2.4962 .84868 .247
0.892 1.16
Loyalty
Behavioral Brand 2.5268 .90198 .247
0.923 1.54
Loyalty
6
As mentioned in Table 4 Brand Commitment has the highest Cronbach alpha value (α=
0.947, Mean = 2.63, Std. Dev = 0.96) while Brand Perceived Risk has lowest value (α =
0.890, Mean = 3.3, Std. Dev = 0.81). All the construct’s Alpha Cronbach values are
greater than 0.7 signifying satisfactory internal reliability (Bryman & Bell, 2015). Brand
Involvement has the lowest skewness (SK=-0.55) and brand trust has highest skewness
(SK= 1.7). All the construct’s skewness values lies between the range of +1.96 to -1.96
that shows construct have the normality (Bryman and Bell, 2015). While the lowest
kurtosis value is of attitudinal brand loyalty (KR= -1.89) and highest is of Brand
Commitment (KR=-3.8). The Kurtosis values of most variables lie below +3 to -3 which
show constructs have the normality (Kallner, 2018).
Multiple regression analysis is used to determine the combine outcome of all predictors
on Attitudinal Brand loyalty. Results are sum up in Table 6.
7
Coefficients Coefficients
B Std. Error Beta
1 (Constant) .367 .106 3.456 .001
Brand Commitment .213 .039 .241 5.440 .000
Brand Trust .170 .057 .156 2.958 .003
Customer Satisfaction .152 .062 .160 2.445 .015
Brand Involvement .348 .050 .411 6.912 .000
Brand Perceived Risk -.048 .037 -.046 -1.302 .194
a. Dependent Variable: Attitudinal Brand Loyalty, R2=.745, Adj. R2=.741, F(388,5-223.547<0.005)
The results explain that the predictors (Brand Commitment, Brand Trust, Customer
Satisfaction, Brand Involvement, Brand Perceived Risk) aggregately explains 74 .of the
variance F =223.5, p=0< 0.05.
Hypothesis 1 states that Brand trust positively impact attitudinal brand loyalty, the results
support the hypothesis refers to the table 6. Hypothesis 2 states that Brand commitment
positively impact on attitudinal brand loyalty, the results support the hypothesis refers to
the table 6. Hypothesis 3 states that Brand perceived risk positively impact on attitudinal
brand loyalty, the results support the hypothesis refers to the table 6
Hypothesis 4 states that Brand Involvement positively impact on attitudinal brand loyalty,
the result support the hypothesis refers to the table 6. Hypothesis 5 states that Customer
Satisfaction positively impact on attitudinal brand loyalty, the result support the
hypothesis refers to the table 6.
It is shown from table 6 that Attitudinal Brand loyalty is getting significantly influence by
Brand Involvement (ß = .411, p<.05), followed by Brand Commitment (ß = .241, p<.05),
Brand Trust (ß = .156, p<.05), Customer Satisfaction (ß = .160, p<.05). However
Attitudinal brand loyalty is not significantly influenced by Brand Perceived Risk (ß =
-.046, p>.05).
Hypothesis 9 states that attitudinal brand loyalty positively impact on behavioral brand
loyalty, the results support the hypothesis refers to the table 8. It is shown in table 4.6 the
predictors (Attitudinal Brand loyalty) 79 of the variance F =1455.8, p=0< 0.05. It is
shown from table 8 that Behavioral Brand loyalty is getting significantly influence by
Attitudinal brand loyalty (ß = .889, p<.05).
9
Table 9A: Relationship with Attitudinal Brand Loyalty
Unstandardized Standardized
Coefficients Coefficients
Model B Std. Error Beta t Sig.
1 (Constant) .098 .063 1.550 .122
Customer Satisfaction(IV) .232 .038 .230 6.190 .000
Attitudinal Brand loyalty(MED) .749 .040 .704 18.945 .000
Dependent variable: Behavioral brand loyalty, R2=809, Adj R2=808, F(386,2-817.281<0.005)
The results show that customer satisfaction and behavioral brand loyalty relationship is
significant. After incorporating attitudinal brand loyalty (mediator) as a predictor the
relationship of customer satisfaction and attitudinal brand loyalty is also significant refer
to table 9A. The results further show that after incorporation of attitudinal brand loyalty
as a predictor, the beta value of customer satisfaction has significantly decreased. Thus it
is inferred that attitudinal brand loyalty mediates customer satisfaction and behavioral
brand loyalty (Chiou & Droge, 2006b; Anand K Jaiswal & Niraj, 2011).
Table 10: Direct Relationship between Brand Trust and Behavioral Brand Loyalty
Standardized
Unstandardized Coefficients Coefficients
Model B Std. Error Beta t Sig.
1 (Constant) .556 .093 5.946 .000
Brand Trust .865 .039 .750 22.285 .000
2 2
a. Dependent Variable: Behavioral Brand Loyalty, R =.562, Adj. R =.561, F(388,1-496.6<0.005)
The results show that brand trust and behavioral brand loyalty relationship is significant.
After incorporating attitudinal brand loyalty as a predictor the relationship of brand trust
and attitudinal brand loyalty is also significant refer to table 10 A. The result also show
10
after incorporation of attitudinal brand loyalty as a predictor the beta value of brand trust
has significantly decreased. Thus it is inferred that attitudinal brand loyalty mediates
brand trust and behavioral brand loyalty(Liu et al., 2011).
Table 11: Direct Relationship between Brand Commitment and Behavioral Brand
Loyalty
Standardized
Unstandardized Coefficients Coefficients
Model B Std.Error Beta t Sig.
1 (Constant) .556 .081 6.938 .000
Brand Commitment .745 .029 .793 22.622 .000
2 2
a. Dependent Variable: Behavioral Brand loyalty, R =.629, Adj. R =.628, F(388,1-656<0.005)
The results show that brand commitment and behavioral brand loyalty relationship is
significant. After incorporating attitudinal brand loyalty as a predictor the relationship of
brand commitment and attitudinal brand loyalty is also significant refer to table 11 A.
The result further also shows that after the incorporation of attitudinal brand loyalty as a
predictor, the beta value of brand commitment has significantly decreased. Thus it is
inferred that attitudinal brand loyalty mediates brand commitment and behavioral brand
loyalty (George & Stavros, 2013).
.
11
CHAPTER 5: DISCUSSION AND CONCLUSION
5.1 Discussion
Consumer purchasing is complex behavior, there might be many factors that can help
persuading consumer to purchase. The aim of the thesis is to better understand both the
dimensions (attitudinal and behavioral) of brand loyalty together. For achieving this aim
a new model is developed based on multidimensional brand theory, attitudinal and
behavioral brand model, the results have answered all the objectives.
5.2 Brand Trust and Attitudinal Brand Loyalty
The first objective of this thesis was to assured that brand trust positively impact on
attitudinal brand loyalty. This research question is answered through H1 that stated that
brand Trust has a positive impact on attitudinal brand loyalty, Brand Trust (ß = .156,
p<.05), The results are consistent with the earlier studies (Aaker, 1991a; Assael, 1998;
Betty & Kahle, 1988; Jacoby & Chesnut, 1978; Chaudhuri & Holbrook, 2001).
Correlation results also show strong positive association among Brand Trust and
Attitudinal Brand Loyalty (Table 6).
12
Stone & Mason, 1995). Correlation results also do not show strong positive association
between Brand Perceived Risk and Attitudinal Brand Loyalty. (Table 6)
13
answered through H8 which stated that customer satisfaction has a significant positive
effect on behavioral brand loyalty, Customer Satisfaction (ß = .145, p<.05).
The results are consistent with the earlier studies (Ganesh et al., 2000; Lee & Lee, 2013;
Traylor, 1981; Zeithaml, 1981c). Correlation result also shows strong positive
relationship between customer satisfaction and behavioral brand loyalty (Table 7).
5.12 Mediating role of Attitudinal Brand Loyalty between Brand Trust and
Behavioral Brand Loyalty
The eleventh objective of this study is to assured the mediating role of attitudinal brand
loyalty between brand trust and behavioral brand loyalty. This research question is
answered through H11 which stated that attitudinal brand loyalty mediates the effect of
brand trust on behavioral brand loyalty. Results in table 10 signify that brand trust has
positive effect on behavioral brand loyalty when interacted directly (ß = .750, p<.05), on
the other hand after the inclusion of attitudinal brand loyalty as mediating variable the
beta value of customer decreased significantly (ß = .191, p<.05) refer to table 10 A. This
clearly indicates the mediating role of attitudinal brand loyalty. The findings are
consistent with the earlier studies (Liu et al., 2011).
14
5.13 Mediating role of Attitudinal Brand Loyalty between Brand Commitment and
Behavioral Brand Loyalty
The twelve objective of this thesis is to assured the mediating role of attitudinal brand
loyalty among brand commitment and behavioral brand loyalty. This research question is
answered through H12 which stated that attitudinal brand loyalty mediates the effect of
brand commitment on behavioral brand loyalty. Results in table 11 signify that brand
commitment has positive impact on behavioral brand loyalty when interacted directly (ß
= .793, p<.05), on the contrary after the inclusion of attitudinal brand loyalty as
mediating variable the beta value of customer decreased significantly (ß = .264, p<.05)
refer to table 11 A. This clearly indicates the mediating role of attitudinal brand loyalty.
The results are consistent with the earlier studies (George & Stavros, 2013).
15
only help to inclined customers towards the brand but also make them purchase
repeatedly, finding is reliable with the previous study (Dholkia, 1997).The outcome also
signify the importance of brand trust as it is also positively related with behavioral brand
loyalty which is consistent with the previous findings (Anuwichanont, 2011a; Hanzaee &
Andervazh, 2012; Matzler et al., 2008a; Ok, Choi, & Hyun, 2011).
Research findings support the view that attitudinal brand loyalty is positively related with
behavioral brand loyalty. Previous studies raised query on embracing attitudinal brand
loyalty in brand loyalty related (Dekempe, Stenkamp,Mellans and Abele 1997; Ehrenbarg
1997a), this study concluded that although attitudinal brand loyalty is not closet predictor
of behavioral brand loyalty, it is a significant predictor of behavioral brand loyalty.
Findings also confirmed the mediating role of attitudinal brand loyalty which is a proof of
key role of attitudinal brand loyalty in developing brand loyalty. Attitudinal brand loyalty
mediate the effect of customer satisfaction on behavioral brand loyalty which is
consistent previous study (Jaiswal & Niraj, 2011). Brand trust has direct and
indirect(mediated) relation with behavioral brand loyalty which prove the mediating role
of attitudinal brand loyalty and is reliable with the previous judgment. (Anuwichanont,
2011b; Gommans et al., 2001; Matzler et al., 2006b, 2008b; Ok et al., 2011). Attitudinal
brand loyalty mediate the impact of brand commitment on ehavioral brand loyalty which
is consistent with the previous study (George & Stavros, 2013).
All the results and their findings are positively significant and reliable with the proposed
model. Data findings concluded good performace of J dot as they are focused on brand
involvement and brand committed. Jdot customers are quite satisfied and they have firm
trust on Jdot.
16
forming consumer attitude for the brand. It is worth noted that if the marketers need to
increase brand loyalty they should focus more on increasing the customer satisfaction.
17
************** THANKS FOR YOUR TIME *******************
18