Cronbachs alpha reliability test

Cronbachs Alpha calculation can be categorized into various status which is shown

as below table:

Cronbach's alpha Internal consistency

0.9 Excellent

0.8 Good

0.7 Acceptable

0.6 Questionable

0.5 Poor

< 0.5 Unacceptable

Table 1 - Cronbach's Alpha rule based on George and Mallery

Factor Number of Item Cronbach's Alpha

Price 5 0.713
Brand 5 0.702
Compatibility 5 0.742
Product Features 5 0.732
Relative Advantage 5 0.745
Social Influence 5 0.741
Purchase Intention 5 0.765
Table 2 - Reliability test for consumers purchase intention on smartphone

From the data analysed in the research, we can see that all the factors have the

Cronbachs Alpha as range between 0.70 and 0.77, which are considered to be

acceptable for the research. Or in the order words, those items were well designed and

data is reliable.

1
 Reliability Statistics
Cronbach's
Alpha N of Items
.702 5

Item-Total Statistics

Corrected Cronbach's
Scale Mean if Scale Variance Item-Total Alpha if Item
Item Deleted if Item Deleted Correlation Deleted
BR1 15.91 4.108 .425 .666
BR2 15.89 3.588 .636 .574
BR3 15.68 3.839 .586 .601
BR4 15.89 4.052 .388 .684
BR5 15.87 4.494 .286 .720
Table 3 Reliability test of Brand (1 run)
st

2
After re-run the SPSS reliability test for the items BR1, BR2, BR3, BR4, we

continued to remove the item of BR4 as its Corrected Item-Total Correlation is 0.297

and the Cronbach's Alpha if Item Deleted is 0.786. In the last run including of BR1,

BR2 and BR3, all the value were well construct and will be used in the next step of

calculating the Exploratory Factor Analysis (EFA)

Reliability Statistics

Cronbach's
Alpha N of Items

.720 4

Item-Total Statistics

Scale Variance Corrected Cronbach's

Scale Mean if if Item Item-Total Alpha if Item
Item Deleted Deleted Correlation Deleted

BR1 11.97 2.759 .520 .652

BR2 11.95 2.454 .675 .554
BR3 11.74 2.738 .584 .616
BR4 11.94 3.103 .297 .786

Table 4 - Reliability test of Brand (2nd run)

3
Reliability Statistics

Cronbach's
Alpha N of Items

.786 3

Item-Total Statistics

Scale Variance Corrected Cronbach's

Scale Mean if if Item Item-Total Alpha if Item
Item Deleted Deleted Correlation Deleted

BR1 8.05 1.574 .573 .765

BR2 8.03 1.419 .680 .647
BR3 7.82 1.588 .625 .710

Table 5 - Reliability test of Brand (3rd run)

To sum up, all the items of the research will have a valid reliability shown in the

below table. More details of other items could be found in the Appendix.

Factor Number of Item Cronbach's Alpha

Price 5 0.713
Brand 5 0.786
Compatibility 5 0.742
Product Features 5 0.732
Relative Advantage 5 0.745
Social Influence 5 0.741
Purchase Intention 5 0.765
Table 6 - The final table of Cronbachs alpha reliability test

1.1 Exploratory Factor Analysis (EFA)

In this section, the researcher present the results of the Exploratory factor analysis

(EFA), which might explain the number of underlying factors or dimensions of each

construct. By using the technique, we will drop some invalid items and group all

items into relevant groups. Varimax rotation would be used to minimize the number

4
of variables that have high loading in each factor. In the EFA calculation, there are

several criteria to be follow with:

The KMO value must be equal or greater than 0.5

Factor maximum absolute value loading must be greater than 0.5

The difference between the maximum absolute value and minimum absolute

value of loading factors must be greater than or equal to 0.3 for any item

(Jabnoun and Al-Tamimi, 2003)

Total variance explained must be greater than or equal to 50%

The KMO test of the research found a value of 0.669, which is above of the minimum

requirements. The Barletts test of schericity was recorded of 0.000, which is

significant at the level of 0.01, this means that factor analysis could be used in this

study and the scale is usable

KMO and Bartlett's Test

Kaiser-Meyer-Olkin Measure of Sampling
.669
Adequacy.
Bartlett's Test of Approx. Chi-Square 2405.214
Sphericity df 378
Sig. .000

Rotated component matrix would show the items under each factor to be grouped in

relevant items. The minimal value of each item on the table should be 0.5. Regarding

to the research, all items were loaded in the range of 0.557 and 0.849 under new nine

dimension describe as below table:

5
 Rotated Component Matrixa
Component
1 2 3 4 5 6 7 8 9
RA4 .782
RA1 .671
RA5 .642
RA2 .624
RA3 .603
FE1 .775
FE2 .695
FE5 .679
FE4 .677
PR2 .797
PR1 .731
PR5 .589
PR3 .586
BR2 .843
BR3 .833
BR1 .771
CO1 .837
CO2 .799
CO3 .732
SI1 .829
SI2 .793
SI5 .557
SI3 .849
SI4 .738
FE3 .635
PR4 .629
CO4 .760
CO5 .611
Extraction Method: Principal Component Analysis.
Rotation Method: Varimax with Kaiser Normalization.
a. Rotation converged in 13 iterations.
Table 7 Exploratory Factor Analysis (EFA) for independent variables

The new factor after the EFA test shall be categorized with new items as below:

6
 Factor Variables included Name Label

Factor 1 RA1, RA2, RA3, Relative advantage RA

RA4, RA5

Factor 2 FE1, FE2, FE4, FE5 Product Features FE

Factor 3 PR1, PR2, PR3, PR5 Price PR

Factor 4 BR1, BR2, BR3 Brand BR

Factor 5 CO1, CO2, CO3 Compatibility CO

Factor 6 SI1, SI2, SI5 Social influence SI

Factor 7 SI3, SI4 Social interaction SIN

Factor 8 FE3, PR4 Performance PE

Factor 9 CO4, CO5 Preference PN

Table 8 Group of EFA

A similar test has been applied to the dependent variables of purchase intention in
order to examine if all the items are grouped into one or not:

KMO and Bartlett's Test

Kaiser-Meyer-Olkin Measure of Sampling
.739
Adequacy.
Bartlett's Test of Approx. Chi-Square 297.187
Sphericity df 10
Sig. .000
Component Matrixa
Component
1
PI2 .774
PI4 .731
PI1 .721
PI5 .700
PI3 .686
Extraction Method: Principal Component Analysis.
a. 1 components extracted.
Table 9 Exploratory Factor Analysis (EFA) for independent variables

7
As the above table, we could discover that the KMO is 0.739, meanwhile, the

Bartletts test of sphericity is 0.000. The both values show that the variable PI was

retained and was expectedly group into one group as proposed.

1.2 Pearson Correlation Coefficient

Pearson Correlation Coefficient is to indicate the direction, strength as well as the

significant of the bivariate relationships between the variables measured on interval

scale (Hair et al., 2007)

8
 Correlations

RA FE PR BR CO SI SIN PE PN PI
RA Pearson
1 .207** .137* .183** .180** .007 .237** .111 .393** .459**
Correlation
Sig. (2-tailed) .002 .043 .007 .008 .912 .000 .104 .000 .000
N 218 218 218 218 218 218 218 218 218 218
FE Pearson
.207** 1 .028 .177** .319** -.047 .178** .353** .272** .447**
Correlation
Sig. (2-tailed) .002 .679 .009 .000 .487 .008 .000 .000 .000
N 218 218 218 218 218 218 218 218 218 218
PR Pearson
.137* .028 1 .195** -.062 .042 -.021 .369** -.035 .061
Correlation
Sig. (2-tailed) .043 .679 .004 .359 .540 .757 .000 .611 .371
N 218 218 218 218 218 218 218 218 218 218
BR Pearson
.183** .177** .195** 1 .225** -.012 .086 .193** .156* .170*
Correlation
Sig. (2-tailed) .007 .009 .004 .001 .865 .208 .004 .021 .012
N 218 218 218 218 218 218 218 218 218 218
CO Pearson
.180** .319** -.062 .225** 1 -.117 .060 .232** .417** .205**
Correlation
Sig. (2-tailed) .008 .000 .359 .001 .085 .381 .001 .000 .002
N 218 218 218 218 218 218 218 218 218 218
SI Pearson
.007 -.047 .042 -.012 -.117 1 .412** .022 .057 .008
Correlation
Sig. (2-tailed) .912 .487 .540 .865 .085 .000 .752 .401 .911
N 218 218 218 218 218 218 218 218 218 218
SIN Pearson
.237** .178** -.021 .086 .060 .412** 1 .141* .108 .179**
Correlation
Sig. (2-tailed) .000 .008 .757 .208 .381 .000 .038 .111 .008
N 218 218 218 218 218 218 218 218 218 218
PE Pearson
.111 .353** .369** .193** .232** .022 .141* 1 .233** .285**
Correlation
Sig. (2-tailed) .104 .000 .000 .004 .001 .752 .038 .001 .000
N 218 218 218 218 218 218 218 218 218 218
PN Pearson
.393** .272** -.035 .156* .417** .057 .108 .233** 1 .537**
Correlation
Sig. (2-tailed) .000 .000 .611 .021 .000 .401 .111 .001 .000
N 218 218 218 218 218 218 218 218 218 218
PI Pearson
.459** .447** .061 .170* .205** .008 .179** .285** .537** 1
Correlation
Sig. (2-tailed) .000 .000 .371 .012 .002 .911 .008 .000 .000
N 218 218 218 218 218 218 218 218 218 218
**. Correlation is significant at the 0.01 level (2-tailed).
*. Correlation is significant at the 0.05 level (2-tailed).

Table 10 - Pearson Correlation Testing

The above table shows the correlations between independent variables and dependent

variable. There are two factors found to be insignificant with the dependent variables

Purchase intention: Price (PR) and Social Influence (SI) have the P value of 0.371 and

9
0.911 which are considerably higher than 0.05. The other factors with P value under

0.05 are acceptable.

The relation between Relative advantage (RA) and Purchase intention (PI) is

considered at a medium rate as the correlation value is 0.459. This value fall in a

range of 0.41 to 0.70, which is said to be a moderate strength.

A similar relation also has been found between the Product Features (FE) and

Purchase intention. While the correlation value falls between 0.41 to 0.70, at the

number of 0.447. It is concluded that there is a moderate correlation between Product

Features and Purchase intention.

There is a relationship between Brand (BR) and Purchase intention. However, as the

correlation value is just 0.170, in the range of 0.00 to 0.20, this relationship is just

considered as a slightly rate.

Meanwhile, there are stronger relationship between Compatibility (CO) and Purchase

intention. While falling in the range of 0.21 to 0.40, the correlation value is 0.205.

In conclusion, there is definite a relationship between Compatibility and Purchase

intention, however, at a small strength.

There is evidence to consider a relation between Preference (PN) and Purchase

intention. As the correlation value is 0.537, it falls between 0.41 to 0.70. This

indicates there is a moderate relationship between Preference and Purchase intention.

Social interaction (SIN) and Purchase intention are at a negligible relationship. The

correlation value appears to be 0.179, hence, there is just a slight relationship between

the two variables

10
There is surely a relationship between Performance (PE) and Purchase intention. As

the correlation value is 0.285, we can conclude that there is definite a relationship

between the two variables, nevertheless, this relationship is in a small strength.

1.3 Multiple Regression Analysis

Weiers (2008) stated that multiple regression analysis is an analysis with one

dependent variable and two or more dependent variables participate in. In the other

hand, Zikmund et al., (2010) demonstrate it to be an association in which the effects

of two or more independent variables on a single, interval-scaled dependent variable

which are investigated continuously.

Model Summaryb

Std. Error of the

Model R R Square Adjusted R Square Estimate
1 a
.677 .458 .440 .43019
a. Predictors: (Constant), PN, SIN, BR, PE, FE, RA, CO
b. Dependent Variable: PI

According to the Model Summary table, the R Square value, which is determined for

the explained variation (variance) in Purchase intention from the seven factors

including Relative advantage, Product Features, Brand, Compatibility, Social

interaction, Performance and Preference is approximately 45.8%. This means that

there is also 55.2% unexplained still left in the study. Meanwhile, the multiple

correlation coefficient (R) provides the strength of the relationship between the

dependent variable and independent variables is 0.677

11
 ANOVAa
Sum of Mean
Model Squares df Square F Sig.
1 Regression 32.883 7 4.698 25.384 .000b
Residual 38.863 210 .185
Total 71.745 217
a. Dependent Variable: PI
b. Predictors: (Constant), PN, SIN, BR, PE, FE, RA, CO

The p-value (Significant of 0.000) in the ANOVA table is less than 0.05. Its means

that the alternative hypothesis as the independent variables are significantly explains

the variance in consumers' level is supported by the data and will be accepted.

Unstandardized Standardized
Coefficients Coefficients

Std.
Model B Error Beta t Sig.
1 (Constant) .094 .369 .254 .800
RA .254 .058 .252 4.411 .000
FE .329 .065 .291 5.070 .000
BR .020 .052 .021 .387 .699

CO -.123 .061 -.119 -2.033 .043

SIN .017 .047 .019 .355 .723

PE .100 .064 .086 1.553 .122
PN .360 .057 .382 6.315 .000
Table 11 - Multiple regression testing

At last, the Coefficients table provides that the following independent variables are

not significant to predict the dependent variable including: Brand (p-value of 0.699),

Social interaction (p-value of 0.723) and Performance (p-value of 0.122). Those

values are excluded while they are higher than the alpha level of 0.05. In contrast,

Relative advantage (RA), Product Features (FE), Compatibility (CO) and Preference

12
(PN) are significant to predict the Purchase intention. Those independent variables

have p-value less than alpha of 0.05. Hence, the relationship between these variables

can be formed by the following equation, which is supported by the above table:

Purchase intention = 0.094 + 0.254(Relative Advantage) + 0.329(Product

Features) - 0.123(Compatibility) + 0.360(Preference)

1.4 Test of Hypothesis

After several stages of testing, the initial hypothesis has been seen to change as the

dependent variables transformed into various groups. The test of hypothesis would be

modified to match with the data collected and previous analyzing. In particular, they

were examined as follow:

Hypothesis 1

H0: There is no impact from compatibility towards purchase intention of Smartphone

among office workers in Hochiminh City

H1: There is an impact from compatibility towards purchase intention of Smartphone

among office workers in Hochiminh City

Reject H0, if p<0.05

As the multiple regression table shows that the p-value of compatibility is 0.043,

which is less than the significant level of 0.05. Hence, H0 was rejected, it means that

compatibility has an impact towards purchase intention.

Hypothesis 2

H0: There is no impact from Product Features towards purchase intention of

Smartphone among office workers in Hochiminh City

13
H1: There is an impact from Product Features towards purchase intention of

Smartphone among office workers in Hochiminh City

Reject H0, if p<0.05

As the multiple regression table shows that the p-value of Product Features is 0.000,

which is less than the significant level of 0.05. Hence, H0 was rejected, it means that

Product Features has an impact towards purchase intention.

Hypothesis 3

H0: There is no impact from brand towards purchase intention of Smartphone among

office workers in Hochiminh City

H1: There is an impact from brand towards purchase intention of Smartphone among

office workers in Hochiminh City

Reject H0, if p<0.05

As the multiple regression table shows that the p-value of brand is 0.699, which is

higher than the significant level of 0.05. Hence, H0 was accepted, it means that brand

has no impact towards purchase intention.

Hypothesis 4

H0: There is an impact from social interaction towards purchase intention of

Smartphone among office workers in Hochiminh City

H1: There is an impact from social interaction towards purchase intention of

Smartphone among office workers in Hochiminh City

Reject H0, if p<0.05

14
As the multiple regression table shows that the p-value of social interaction is 0.723,

which is higher than the significant level of 0.05. Hence, H0 was accepted, it means

that social interaction has no impact towards purchase intention.

Hypothesis 5

H0: There is an impact from performance towards purchase intention of Smartphone

among office workers in Hochiminh City

H1: There is an impact from performance towards purchase intention of Smartphone

among office workers in Hochiminh City

Reject H0, if p<0.05

As the multiple regression table shows that the p-value of performance is 0.723,

which is higher than the significant level of 0.05. Hence, H0 was accepted, it means

that performance has no impact towards purchase intention.

Hypothesis 6

H0: There is an impact from preference towards purchase intention of Smartphone

among office workers in Hochiminh City

H1: There is an impact from preference towards purchase intention of Smartphone

among office workers in Hochiminh City

Reject H0, if p<0.05

As the multiple regression table shows that the p-value of preference is 0.000, which

is less than the significant level of 0.05. Hence, H0 was rejected, it means that

preference has an impact towards purchase intention.

15
Hypothesis 7

H0: There is an impact from relative advantage towards purchase intention of

Smartphone among office workers in Hochiminh City

H1: There is an impact from relative advantage towards purchase intention of

Smartphone among office workers in Hochiminh City

Reject H0, if p<0.05

As the multiple regression table shows that the p-value of relative advantage is 0.000,

which is less than the significant level of 0.05. Hence, H0 was rejected, it means that

relative advantage has an impact towards purchase intention.

Hypothesis 8

H0: No factors (Brand, Relative Advantage, Compatibility, Performance, Preference,

Social Interaction and Product Features) have the most significant influence on the

purchase intention of smartphone among office workers in Hochiminh City

H1: At least one factor (Brand, Relative Advantage, Compatibility, Performance,

Preference, Social Interaction and Product Features) has the most significant influence

on the purchase intention of smartphone among office workers in Hochiminh City

Reject H0, if p < 0.05

There are four factors influence on the purchase intention of smartphone among the

office workers in Hochiminh City: Relative advantage, Product Features,

Compatibility and Preference. Those three factors have p-value of under the

significant level of 0.05. Step back to the table of Regression above, preference will

16
be the factor most significant influence on purchase intention of smartphone among

office workers in Hochiminh City as its highest Beta value of 0.382

17
 CHAPTER 2. CONCLUSION

2.1 Conclusion

This study explores the factors influence on the purchase intention of smartphone. 218

respondents have been selected to measure the purchase intention of smartphone.

Questionnaires were provided, which will collect data into 5 groups of independent

variables.

The respondents of this study including of 50% males and 50% females

unintentionally. Meanwhile, most of respondents are at the age of 22 to 29 with 64%.

Most of them have an average income less than 8,000,000 VND (32% in detail). The

Product Features of connect to internet was founded to be an important items as its

mean of 4.29. In contrast, the least mean of item value was 2.63 of Friends and

family are very helpful to me in making decision of buying smartphone in Social

influence.

The data analyze also presented some irrelevant points in the study. Two items were

excluded from the factor Brand after three times of Cronbachs alpha test. This stage

has also demonstrated that all factors had the reliability value in range of 0.7 to 0.8,

which is considered as a good rate.

Several factors have been examined if they have impact on the purchasing intention of

smartphone in this study. From the very beginning, the researchers spend efforts

analyzing this relationship between the dependent variables and the independent

variables including: Price, Brand, Compatibility, Relative Advantage, Product

Features and Social Influence. Nevertheless, collected data present that office workers

in Hochiminh City has been affected by Relative Advantage, Product Features,

Compatibility and Preference. While the first three variables as listed are proposed at

18
first, the last variables was surprisingly discovered after the validity test. The variable

was a part of the previous items found in Product Features and Price. Furthermore,

Preference is also known as the variable which most influence the purchase intention

with the highest Beta in Multiple Regression Calculation. However, Compatibility

was also found to have negative relationship with purchase intention. Meanwhile, the

rest three factors were known to play a vital role in affecting the dependent variable.

Since then, adjustments of those three factors will lead to changes of purchasing

intention in reality.

Evidences from the study demonstrated that, smartphone firms might be inherited

from the buying decision if they could enhance the effect of Relative Advantage,

Product Features, Compatibility and Preference into real product, as well as into their

marketing strategy.

2.2 Practical implication

From the academic findings, practical implication is also recommended, which might

support the managerial decisions of smartphone manufacturers in term of product

development, marketing and communication.

While Relative Advantage has been proven to have impact on purchase intention, it is

suggested to enhance the advantages of smartphone towards the old feature phones.

Instead of perceiving a just-enough-to-use phone, consumers nowadays require more

hi-tech complements, more fashion and style and higher integration. Invest on product

development with leading value continuously might raise an advantage for

smartphone, especially, in the segment of office workers, who are at intensive use of

smartphone.

19
Product Features is yet another field to consider in product development. For office

workers, who are examined to prefer function of gaming, taking photos and

connecting to internet, those phone with ready feature to fulfill their requirements

might take a easily step to approach those group of consumers. As well as focus on

these features for firms marcom campaign. Especially, when the manufacturer has

their ability on offering high-quality of those features, take them as their unique

selling points to promote as a key road to conquer the consumers.

Last but not least, Preference has been analyzed to be the most significant impact on

the purchase intention, which means their items regarding the use of applications and

price per quality would be considered as most important things to smartphone

producers. It is essential to understand that office workers, who tend to have higher

education comparing to other group of society, may spend time and efforts in

searching information and review for a smartphone. Since then, those people logically

require a competitive price in connection with the quality they will perceive.

Furthermore, as employees work in environment of computing machines, applications

with productivity would be an useful tool into their smartphone. Firms might pay

attention to provide such appropriate price with meaningful software to satisfy their

consumers.

In addition to the above implications, it is necessary for firm to collaborate with their

consumers. Review and feedback from consumers for a smartphone function, design,

price and build quality is required to build a strong relation between manufacturer and

consumers, as well as to deeply understand the consumers continuously.

20
2.3 Limitation and further recommendation

The data analyze has concerned about the precision of factors and items included.

There are items and factors removed from the initial proposal, which have let to

changes of analyzing factors and items. However, as the purpose to understand the

relationship between factors influencing the purchase intention of smartphone, later

analyze has significant discover to unhide the researchers questions. Alongside the

usefulness findings, this research has several limitation to be discussed.

Firstly, with a narrow scope of office workers in Hochiminh City with small samples,

this research does not have ability to generalize the results in term of the whole

country. It is recommended to have further studies in different cities and province

with larger samples. Since then, more accurate insights of smartphone consumers

would be draw out to represent for the Vietnamese smartphone market.

Secondly, the study sample has not include the other groups as teenager or students,

who are very fond of smartphone and have highly intention to use smartphone.

Further explanation for more segment insight should give a wider picture for firms to

consider while planning to launch new products.

Thirdly, various stages of the consumers buying decision have not included in this

research. Those stages also have important impact on the purchase intention of

consumers, it is suggested that researchers should also consider and fulfill other

determinants.

21
22
23