Chapter-4: Analysis and Findings

This chapter provides empirical results for the research. It consists of the analysis of the Omani
sample by factor analysis, correlation analysis and multiple regression analysis. It also provides
discussion on the findings and compares the findings between Omani and Expat nationalities. It
concluded by estimating significant differences in variables between the nationalities.

4.1: Demographic description

Out of the 161 qualified respondents there were occupants of ten nations: 82 respondents from
Oman and 57 from expat in particular India, Bangladesh, Philippines and Sri Lankan. 64,6% of all
respondents were females. Age changed from 18 to 67 years of age. The vast majority of the
respondents (47%) have an full time work. Among the other occupation bunches the greatest are
students (15,5%) and those with part time work (14%).

Table 2: Sample characteristics

Characteristics Total sample Omani sample Expat sample
Female 104 (64,6%) 43 (52,4%) 45 (79%)
Gender
Male 57 (35,4%) 39 (47,6%) 12 (21%)
Age range 18-67 19-67 18-63
Age
Median age 28 30.5 24
student 25 (15,5%) 13 (15,8%) 3 (5,3%)
full-time job 76 (47,2%) 29 (35,4%) 42 (73,7%)
part-time job 23 (14,3%) 15 (18,3%) 8 (14%)
Occupation student + job 17 (10,6%) 14 (17%) -
retired 6 (3,7%) 3 (3,7%) 2 (3,5%)
unemployed 10 (6,2%) 6 (7,3%) 2 (3,5%)
other 4 (2,5%) 2 (2,4%) -
Number of respondents 161 82 57
The survey's supply time was limited; along these lines the quantity of respondents was low
and unequal in the examples. Omani sample size of 82 respondents speaks to the number of
inhabitants in Omani online customers. It rises to roughly 4.6 million Omanis (Wikipedia
2019). This example measure relates to margin of error of 11% with 95% confidence interval.
Expat sample of 57 participants shows the margin of error of 14% with 95% confidence
interval for the number of inhabitants in 3.0 million online customers in Oman and GCC.
(Wikipedia 2019). In this way, the Omani example better speaks to its populace, which
legitimizes its decision as a principle test for hypothesis approval.

Omani example is similarly spoken to by males (48%) and females (52%). The age shifts
somewhere in the range of 19 and 67. A large portion of the respondents have an all day work or
low maintenance work. In the Expat test 79% are ladies, age changes from 18 to 63. Dominating
gathering of occupations is all day specialists.

Concerning the frequency on internet shopping, evaluated by the main inquiry in the poll, 42%
of Omani respondents professed to have ordinary shopping background while 58% shopped
online at times. In Expat ordinary customers comprise 49% of the example versus 51% irregular
online purchasers.

4.2. Analysis of the Omani sample

4.2.1. Reliability of the constructs
At the point when a build comprises of a few estimates it is critical to decide it’s internal
reliability, which shows how the things cooperate, regardless of whether they measure
something very similar. Cronbach's alpha coefficient estimates the reliability. Satisfactory
dimension of reliability is demonstrated by an esteem more significant than 0.7 (Easterby-
Smith et al. 2015). Cronbach's alpha coefficients for Omani example are exhibited in the table
below.
Table 3: Cronbach’s alpha for Omani sample

Cronbach
Construct Measures alpha

 I intend to purchase electronic devices in an online

store in the next 12 months
Intention to buy
 It is likely that I will purchase electronic devices in
Consumer electronics .911
an online store in the next 12 months
online
 I want to purchase electronic devices in an online
store in the next 12 months

 Online stores are a suitable place to buy electronics

Attitude towards  Online stores are a reliable and safe place to
buying consumer buy electronics .872
electronics online  Online stores are a better place to buy electronics
than physical stores

 People that are important to me (family, friends)

would buy electronic devices online
 People that are important to me (family, friends) .694
Subjective norm with
would recommend me to buy electronic devices if item deleted
regards to buying
online –
electronics online
 Media (advertisement, online reviews, social .845
media posts etc.) encourages me to buy electronics
online

Perceived behavioral  I know where and how to buy electronics online

control of buying  I feel that buying electronics online is not
.840
consumer goods problematic
online  I have enough time to buy electronics online

 It would be easy for me to navigate online stores

and make a payment if I wanted to buy electronics
Attitude toward Using
online
online stores When
 I find it enjoyable to use online stores for .634
buying consumer
buying electronics
electronics
 I can save time and/or money if I buy
electronics through an online store

 A high degree of trust exists in my family .598

Cultural Environment
of trust  People in my community trust each other if item deleted -
 I live in a high trust society .672
The constructs of Intention, Attitude towards purchasing buyer electronics on the web and
Perceived behavioral control have high Cronbach's alpha estimations of more significant than
0.8. Subjective norm construct had a 0.694 coefficient, and evacuation of the third measure
(media impact) would expand it effectively to 0.845, which proposes that the proportion of
media impact does not completely incorporate with different proportions of Subjective
standard. Along these lines, I chose to dispense with it from the develop. Same rationale was
utilized with respect to the develop of Cultural condition of trust. It created the impression that
the proportion of trust in the family was not unequivocally identified with different measures,
and its expulsion expanded the unwavering quality, so I chose to dispose of it also, which
expanded the builds Cronbach's alpha to 0.672. Subsequently, there were two develops with
unwavering quality of marginally beneath 0.7 (Attitude toward utilizing on the web stores
when purchasing purchaser gadgets (0.634) and Cultural condition of trust (0.672). Regardless
of lower unwavering quality, I chose to continue with these develops as a result of their
hypothetical criticalness. Analysts additionally propose that lower dimensions of Cronbach's
alpha can be acknowledged in showcasing research (Park and Jun 2003).
4.2.2 Factor analysis
Firstly, I evaluated factor loading for every individual develop to decide how separate things
load together (Appendix 2). The outcome demonstrated strong loadings from 0.69 to 0.94. As
indicated by Hair et al (1998), when sample size is around 85, as on account of Omani example,
huge factor loadings would be those of 0.6 and higher, so the outcome of Omani example show
great dimensions of factor loadings.

Later I led exploratory factor investigation for every one of the measures together utilizing
Varimax turn (table 4.1.2). Just estimations of higher than 0.3 were incorporated into the yield to
outline just significant outcomes and make the table increasingly readable. The quantity of
elements (components) to hold (3) depended on Kaiser's basis of Eigen value over 1 and the
screen test (Yong and Pearce 2013). The way that number of segments does not compare to
wanted number of builds (6) can imply that a few measures are more related than others. The
quantity of measures in this examination was negligible in light of the fact that the objective was
to make a short and clear poll to get more reactions. More measures can fortify each develop and
make it increasingly particular from others. We can see that dependent on the most grounded
stacking measures the three parts catch distinctive articulations: electronics, component 2 -
statements on other people’s opinions, component 3 – statements on cultural environment of
trust.
Table 4: Factor analysis for Omani sample
Rotated Component Matrix
Component
1 2 3
Intention: I intend to purchase electronic devices
.682 .338
in an online store in the next 12 months
Intention: It is likely that I will purchase
electronic devices in an online store in the next 12
.745 .362
months
Intention: I want to purchase electronic devices in
.731
an online store in the next 12 months
Attitude: Online stores are a suitable place to buy
.728 .345
electronics
Attitude: Online stores are a reliable and safe
.795 .322
place to buy electronics
Attitude: Online stores are a better place to buy
.545 .451
electronics than physical store
Subjective norm: People that are important to me
(family, friends) would buy electronic devices
.844
online
Subjective norm: People that are important to me
(family, friends) would recommend me to buy
.883
electronic devices online
PBC: I know where and how to buy electronics
.744
online
PBC: I feel that buying electronics online is not
.785
problematic
PBC: I have enough time to buy electronics
.773
online
TAM: It would be easy for me to navigate online
stores and make a payment if I wanted to buy
.678
electronics online
TAM: I find it enjoyable to use online stores for
.700
buying electronics
TAM: I can save time and/or money if I buy
.544
electronics through an online store
.797
CET: People in my community trust each other
.845
CET: I live in a high trust society
44.6% 10.6% 8.1%
Explained variance
Component 1 clarifies 44.6% of the variance and is loaded by measures that speak to goal, frame
of mind to web based shopping, perceived behavioral control and attitude to utilizing on the
online stores. Component t 2 clarifies 10.6% of the difference and shows loading of measures
that intention, attitude to internet shopping and subjective norm. Component 3 clarifies 8.1% of
the difference and is loading by measures of cultural environment of trust.
The three proportions of intention are emphatically loaded onto component 1 (from 0.69 to 0.75)
with weaker cross-loadings to component 2 (from 0.34 to 0.36). Their internal reliability was
demonstrated with a high Cronbach's alpha estimation of 0.9 (table 4.1.1), so they can be figured
into one develop attitude towards buying consumer electronics online, and represented by
variable attitude.
The three measures of perceived behavioral control to internet shopping are strongly loaded to
component 1 (from 0.55 to 0.8) with weaker cross-loading to component 2 (from 0.35 to 0.45).
Cronbach's alpha estimation of 0.87 demonstrated their internal reliability, so it legitimizes that
they can be figured into one construct - Attitude towards purchasing buyer gadgets on the web,
and spoken to by factor Attitude.
The two proportions of subjective norm emphatically on the component 2 (0.84 and 0.88) and
have Cronbach's alpha of 0,85. Along these lines, they can be figured into one construct –
subjective norm with regards to buying consumer electronics online, and represented by variable
subjective norm.
The three proportions of perceived behavioral control show solid loading on component 1 (from
0,74 to 0,78) and high inner reliability with Cronbach's alpha of 0.84, so they can be figured into
one construct perceived behavioral control of buying consumer goods online, which was
represented by variable PBC. The three proportions of attitude to online stores (TAM) have huge
loading on component 1 (from 0.54 to 0.7). With Cronbach's alpha of 0.63 their internal
reliability can't be viewed as high; in any case, they don't have different cross-loadings, which
demonstrates that they are predictable in their importance. Along these lines, with the end goal of
this examination I choose to figure them together into one construct - Attitude toward using on
line stores when purchasing shopper electronics, which was spoken to by factor Attitude_TAM.
The two proportions of cultural environment of trust load firmly to concept 3 (0.79 and 0.85) and
have high inside internal reliability (0.67), so I process them into one develop – Cultural
environment of trust, which was spoken to by factor CET.
4.2.3 Correlation analysis

Correlation matrix (table 4.1.3) shows relationships between variables assessed with Pearson’s
coefficient.

Table 5: Correlation matrix for Omani sample

Subjective
Factors Intention Attitude PBC Attitude_TAM CET
norm

Intention 1
Attitude .679** 1
Subjective .408** .485** 1
norm
PBC .561** .623** .362** 1
Attitude_TAM .687** .716** .394** .670** 1
CET -.049 -.099 .101 .059 -.019 1
**. Correlation is significant at the 0.01 level (2-tailed).

PBC – Perceived behavioral control, CET – cultural environment of trust, Attitude TAM – attitude
towards using online stores

The output demonstrates that the greater part of the factors have noteworthy connection. Just
Cultural environment trust (CET) has no huge connection with different parameters. It
demonstrates that in the Omani example there was no relationship distinguished between trust in
the public arena and parts of web based shopping.

Different factors have positive connection that ranges among feeble and solid. Concerning
Intention, it has solid relationship with Attitude and Attitude TAM and moderate connection with
Subjective standard and PBC. It implies that expansion in those factors prompts increment in aim
to purchase electronics on the web.
Among different factors, the most grounded relationship is among Attitude and Attitude_TAM,
which demonstrates that mentality to purchasing on the web emphatically relies upon disposition
to utilizing innovation at the same time and the other way around. Additionally, there are solid
connections amongst Attitude and PBC, and Attitude_TAM and PBC.

So, the correlation matrix shows many links between intention and other variables and gives
grounds to multiple regression analysis. Cultural environment of trust was kept in the model for
theory validation purposes and for further comparison between Omani and Expat.

4.2.4 Multiple linear regression and hypotheses testing

I utilized multiple linear regression to test the five hypothesis that were outlined in the
hypothetical piece of the examination. Proposed research model was worked to test the
relationship between the dependent variable (Intention) and independent factors. The resultof the
examination is exhibited in the table.

The model’s adjusted R2 equals 0,56, which indicates that 56% of the variation is explained by
this model, so the predicting power of the model is quite high. Statistical significance of the
coefficients is determined by level of p (Sig.), where it has to be below 0,05 to confirm
significance of the coefficients.

Multicollinearity issues were checked through the VIF (Variance inflation factor). When its level
is above 10, it is usually a sign of multicollinearity problem in the model. This is not the case for
the model in this research, as every VIF value is below this threshold. So, there is no
multicollinearity issue in the model.
Table 6: Multiple linear regression analysis for Omani sample

Unstandardized Standardized t p Statistics

Model Coefficients Coefficients (Sig.) VIF
B Std. Error Beta
(Constant) -.082 .959 -.086 .932
Attitude .332 .153 .259 2.166 .034 2.618

Subjective norm .108 .116 .082 .930 .356 1.418

PBC .134 .191 .077 .700 .486 2.208

Attitude_TAM .592 .190 .366 3.110 .003 2.546
CET -.001 .137 -.001 -.007 .995 1.092
Age -.002 .008 -.016 -.202 .841 1.135
Gender -.539 .178 -.235 -3.024 .003 1.108
Occupation -.009 .049 -.015 -.193 .848 1.066
Dependent variable: Intention

R2=0.559
PBC – Perceived behavioural control, CET – cultural environment of trust, Attitude_TAM – attitude
towards using online stores

Regression output allowed to obtain the following results for hypotheses testing.
H1 expressed that there is a positive connection between attitudes towards purchasing customer
electronic on the web and and individual’s intention to buy consumer electronics online Based on
the results from multiple linear regression output for Attitude (t=2,17; p=0,034), H1 is accepted.
It implies that there is a relationship between attitude to online shopping of electronics and
intention to buy electronics in online stores.. The investigation demonstrates that behavior
towards purchasing electronics online has impact on explanation of consumer’s intention. The
estimation of b-coefficient can be translated like so: an expansion in frame of mind by 1 point
adds to increment in aim by 0,33, every single other factor continuing as before.
H2 stated: “There is a positive relationship between subjective norm and individual’s intention to
buy consumer electronics online”. Subjective norm has no significance is the model (t=0,93;
p=0,356), so this hypothesis is rejected. In spite of the moderate correlation between's parameters
(0,4), which demonstrates a positive relationship, subjective norm isn't valued accurately critical in
clarification of purchaser's goal to purchase hardware in online stores. Thus, shopper's observations
about opinions of other individuals on internet shopping are not demonstrated as a solid factor of
effect on aims to buy with regards to this investigation.

H3 stated that there is a positive relationship between perceived behavioral control and
individual’s intention to buy consumer electronics online. Based on obtained values (t=0,7;
p=0,486) this hypothesis is rejected. Despite the reasonable correlation between intention and
perceived behavioral control (0,56), the analysis shows that perceived behavioral control has no
statistically significant power to explain intention to buy consumer electronics online.

The findings from the first three hypotheses show that only one component of the Theory of
planned behavior, attitude, can be used for explaining intention in this model.

H4 stated that there is a positive relationship between attitude toward using online stores and
individual’s intention to buy consumer electronics online. Based on the data (t=3,11; p=0,03) this
hypothesis was accepted. It implies that the attitude toward utilizing on the online stores affects
the clarification of purchaser expectation to purchase. The estimation of b-coefficient can be
interpreted like so: an expansion in frame of mind towards utilizing on the web stores by 1 point
adds to intention by 0,592, every other factor continuing as before. Attitude toward utilizing on
the online stores has the most astounding impact on goal in this model.

This demonstrates the significance of the Technology acceptance model when shopping conduct
is inspected in online setting and supports the examinations that interface TAM to look into in
online purchaser conduct.

H5 stated: “There is a positive relationship between cultural environment of trust and

individual’s intention to buy consumer electronics online”. Based on the obtained values (t=-
0,007; p=0,995) it was rejected.
Therefore, the cultural trust environment is not statistically significant in explaining the intent to
buy electronics online. Furthermore, the correlation analysis evidence indicates that there is no
relationship between the structures. Therefore, we can conclude that the concept of confidence in
culture does not have explanatory power of statistical significance in the analysis of consumer
intent to buy online. But this may be the case for the Omani sample. Oman is a country with the
highest levels of social trust (Newton 2001). But a fairly recent activity, such as online shopping,
can be seen as exceeding the usual individual's tendency to trust..

Attitude towards buying consumer H1 (b=0.332)

electronic online

Subjective norm

[Perceived behavioral control of

buying consumer electronics online Intention to buy
consumer electronics
online
Attitude toward using online stores H4 (b=0.592)

Cultural environment

Figure 4. Hypotheses testing results

Regression also showed the importance of one control variable in this model, sex variable
(t = -3.024; p = 0.03). Equivalent to the value of b-coefficient 0,539, which means that males
have a stronger intention to buy online electronics than females. This can be explained by the
fact that women tend to perceive the risk when buying online more than men (Garbarino and
Strahilevitz 2004).

For age and occupation control variables, they have been shown to be insignificant. I tried to
assemble them (profession: student, work, not working; age: groups by age), but I did not give
any significant impact on the regression model.
4.3. Cross-national analysis between Oman and Expat
This section aims to find the answer for the second research question 2: Is there a difference in
intention to purchase electronics online and factors that contribute to different nationalities?

4.3.1 Discussion of the findings from separate analyses.

The result of the separate analysis of Omanis and expatriates revealed a significant achievement
for the validity of the proposed theoretical model. With high explanatory capacity (R2 = 0.56),
multiple regression analysis of the Omani sample allows acceptance of two basic hypotheses (H1
and H2). For the sample of expatriates, only one hypothesis (H1) can be accepted, but the
explanatory power of the model was low (R2 = 0.17) (Annex 3). Also, the correlation matrix for
the Omani sample showed that the intent was associated with all other structures except for the
cultural trust environment. The correlation matrix for a sample of expatriates showed that the
intention was only related to the position of buying consumer electronics over the Internet.

Table 7: Results of hypotheses testing in Oman and Expat

Omani Expat
H1: There is a positive relationship between attitude
towards buying consumer electronics online and accepted accepted
individual’s intention to buy consumer electronics online.
H2: There is a positive relationship between subjective
norm and individual’s intention to buy consumer electronics rejected rejected
online.
H3: There is a positive relationship between perceived
behavioral control and individual’s intention to buy rejected rejected
consumer electronics online
H4: There is a positive relationship between attitude
toward using online stores and individual’s intention accepted rejected
to buy consumer electronics online.
H5: There is a positive relationship between cultural
environment of trust and individual’s intention to buy rejected rejected
consumer electronics online.
4.3.2 Difference in level of variables between Oman and Expat
Analysis of level of variables in cross-cultural research estimates differences in specific
variables in the nationalities compared (Ghhauri and Grønhaug 2010). One way of doing so is
by comparing the means of each variable across countries. Figure 5 allows to visualize means of
the variables in Omani and Expat, considering that statements were valued from 1 (strongly
disagree) to 5 (strongly agree).

Figure 5: Comparison of means: Omani and Expat

We can see that there are some differences in most variables. This shows that expats seem more
willing to buy electronics online. They also show a more positive attitude to buy and a higher
value for autonomy in online shopping. Oman assessed the role of behavioral control more
positively. The level of attitude of using online stores is almost the same level. The level of the
cultural environment of trust is higher in Oman. But this is just a visual comparison of means.
There is a need to test the statistical significance of this difference.

I chose to use the non-barometric Mann-Whitney U test due to different sizes of Omani and
expatriate specimens and very small sample sizes. I have separately evaluated the key variables
(table below) and the measures that comprise them (appendix 3).
Table 8: Mann-Whitney test results for the constructs

p
Mann- (Asym
Mean Sum of
Country N Whitney p. Sig.
Rank Ranks
U (2-
tailed))
Omani 82 64.85 5317.50
Intention 1914.500 .067
Expat 57 77.41 4412.50
Omani 82 61.76 5064.50
Attitude 1661.500 .003*
Expat 57 81.85 4665.50
Omani 82 62.29 5107.50
Subjective Norm 1704.500 .005*
Expat 57 81.10 4622.50
Omani 82 75.26 6171.50
PBC 1905.500 .056
Expat 57 62.43 3558.50
Omani 82 70.66 5794.50
Attitude_TAM 2282.500 .813
Expat 57 69.04 3935.50
Omani 82 79.58 6525.50
CET 1551.500 .000*
Expat 57 56.22 3204.50

* Indicates a significant difference between the three groups at the level of 0.05 tested by Mann–Whitney
U test

Mann Whitney's test shows that construct of intention does not show a significant difference
between groups (p = 0.067> 0.05). This means that people in Oman and Expat have evaluated
the intention of buying electronics online in a similar way, so it seems that cultural differences
do not affect the intention. However, in-depth analysis of discrete elements (Appendix 3) shows
that among the three elements of intent, one of them shows a significant difference between
groups ("I'm likely to buy electronics from an online store within 12 months"). It reflects a
different perception of the potential of online purchases in Oman and Expat, which can occur
because of cultural differences.
There are three structures that show a great difference between Oman and Expat - the attitude
toward buying electronics online, the personal standard, and the cultural environment of trust.
Omanis have a different attitude towards buying electronics online (p = 0.003) for expatriate
residents. But there is a measure that supports such a difference ("e-shops are a better place to
buy electronics from the actual store"). This means that there are different opinions in online
stores as an alternative to physical stores in Oman and expatriates. This may depend on many
cultural issues, for example, increasing risk in local online stores, increasing loyalty to some
stores, and so forth.

As for building a cultural environment of trust, cultural differences were predicted and
confirmed through analysis (p = 0.00). However, opinions only differ about the term "I live in a
high confidence society" significantly. This research is supported by trust in the community,
which Oman is generally estimated to be one of the highest levels of social trust in the world
(Newton 2001). At the same time, expatriates are developing countries that suffer from public
mistrust in political and government institutions (Stepanenko 2006).

The subjective rule (p = 0.005) varies between countries. All measures for this construction show
a significant difference between the two groups. He pointed out that perceptions of the views of
others on the purchase of electronics online are different between countries.

Behavioral control and attitudes towards the use of online stores do not show a significant
difference between groups based on the Mann Whitney test. Separate phrases vary within
structures, but do not contribute significantly to construction. It shows that people in Oman and
expatriates believe that they control the process and that the technological aspects of using online
stores are similar.