Data “Repurchase intention”

Variable description:
• “Web”: The e-commerce website that participants made purchases
• “Gender”: Gender of participants (1 = Male; 2 = Female)
• “Age”: Age range of participants (1 = under 25; 2 = 25 – 30; 3 = above 30)
• “PFreq”: Purchase frequency of participants (1 = once per month; 2 = twice per month; 3 = three
times per month; 4 = four times or above per month)
• “Income”: Monthly income of participants (1 = under VND 5 million; 2 = VND 5 million - VND 10
million; 3 = above VND 10 million)
• “Education”: Education level of participants (1 = high school; 2 = college; 3 = university or higher)
• “TRU”: Trust in online shopping (1 = low trust, 2 = medium trust, 3 = high trust)
• “RP”: Repurchase intention (Likert 7-point: 1= totally disagree & 7 = totally agree) measured
by 4 items (RP1, RP2, RP3, RP4)
• “WOM” Word of mouth (Likert 7-point: 1= totally disagree & 7 = totally agree) measured by 5
items (WOM1, WOM2, WOM3, WOM4 & WOM5)
• “IQ”: Information quality of retailing website (Likert 7-point: 1= totally disagree & 7 = totally
agree) measured by 5 items (IQ1, IQ2, IQ3, IQ4 & IQ5)

Questions: Analysing the data and interpret the results to meet the following manager’s information
needs:
1. The manager wishes to know which retailing website consumers make purchase most frequently and
least frequently.
2. In order to employ STP (segmenting – targeting – positioning), the manager wishes to know
characteristics of the e-shopper in details (level of trust, gender, age, purchasing frequency, income
and education level).
3. Before employing data analysis, the manager wishes to check the reliability of the scales measuring
WOM, IQ and RP based on Cronbach alpha.
4. The manager wishes to identify basic statistical figures of WOM, IQ and RP including mean, median,
mode, range, standard deviation, skewness and kurtosis.
5. The manager wishes to know how trust varies due to different level of age, purchase frequency and
income.
6. The manager thinks that WOM may influence heavily on consumer in online purchase, the
manager wishes to test the hypothesis stating that mean of WOM in the population exceeds 5.
7. The managers think that male and female consumer may not similar in evaluating quality of
provided information and repurchase intention. Thus, the manager wishes to check whether
repurchase intention and perceived information quality are statistically different or not across
gender.
 OUTPUT
Question 3: Check the reliability of the scales measuring WOM, IQ and RP based on Cronbach alpha.

DEFINITION
In quantitative research, measuring large factors can be
complex, so we need detailed scales which consist of
multiple observed variables (questions) to accurately and
specifically each factor.

For instance, if we want to measure the factor "promotion-

training" affecting employee satisfaction, we might use
questions like:

 PT1: "I am trained with essential skills to complete tasks successfully."

 PT2: "The company provides promotion opportunities based on skill."
 PT3: "The company ensures fairness in promotion opportunities."
 PT4: "Employees are aware of promotion requirements."
However, not all these observed variables reflect the concept of the factor they intend to measure
 Cronbach's Alpha reliability test is the tool for assessing whether the observed variables within a
factor are reliable, effective and should be included in the scale  ensure consistency among observed
variables within a scale  the scale accurately measures the factor.

Cronbach's Alpha coefficient ranges between 0 and 1.

 0 = little to no correlation among the observed variables
 1 = perfect correlation

CRITERIA
1. Cronbach’s Alpha
- Ideally, Cronbach’s Alpha ≥ 0.7, but for exploratory studies, > 0.6 is acceptable.
- Higher values indicate greater scale reliability. But, to prevent variables measuring the same
aspect too similarly (redundancy), Cronbach’s Alpha < 0.95.

2. Corrected Item-Total Correlation: measures how each observed variable correlates with others
within the scale.
- Corrected Item-Total Correlation ≥ 0.3
- Higher values means that the observed variable correlates well with other variables in the scale
(consistency). Lower values indicate that the observed variable is less correlated with other
variables and less effective in measuring the factor  reconsider/remove.

PROCEDURE IN SPSS
Analyze  Scale  Reliability analysis  Input the list of items  In “Statistics”, tick “Scale if items
deleted”.
OUTPUT
1. WOM
Reliability Statistics
Cronbach's Alpha N of Items
.932 5

Item-Total Statistics
Scale Mean if Scale Variance if Corrected Item- Cronbach's Alpha if
Item Deleted Item Deleted Total Correlation Item Deleted
WOM1 19.940 27.956 .780 .924
WOM2 19.670 28.607 .726 .934
WOM3 20.310 27.206 .807 .919
WOM4 20.140 25.253 .898 .901
WOM5 20.220 25.830 .892 .903

2. IQ
Reliability Statistics
Cronbach's Alpha N of Items
.870 5

Item-Total Statistics
Scale Mean if Scale Variance if Corrected Item- Cronbach's Alpha if
Item Deleted Item Deleted Total Correlation Item Deleted
IQ1 17.5600 21.501 .617 .861
IQ2 17.9500 19.543 .772 .824
IQ3 18.2400 18.972 .805 .815
IQ4 18.3000 19.000 .655 .858
IQ5 17.4300 21.541 .653 .853

3. RP
Reliability Statistics
Cronbach's Alpha N of Items
.820 4

Item-Total Statistics
Scale Mean if Scale Variance if Corrected Item- Cronbach's Alpha if
Item Deleted Item Deleted Total Correlation Item Deleted
RP1 14.8000 13.980 .661 .765
RP2 15.0600 14.421 .626 .781
RP3 15.1400 14.142 .669 .762
RP4 14.7900 13.501 .619 .787

 The Cronbach's Alpha values of WOM, IQ, RP are 0.932, 0.870, 0.820 respectively, which are all
above 0.7.
 The corrected item- Total Correlation for each item within the factors are all above 0.3  Each
observed variable within these scales shows adequate correlation with other variables in its factor.
 The scales for WOM, IQ, and RP are reliable and valid.
Question 6: Test the hypothesis stating that mean of WOM in the population exceeds 5.

UNDERSTANDING THE DATA

 Columns WOM1 to WOM5: represent the scores given to each platform across five different
aspects of WOM. Each aspect is measured on a Likert scale from 1 to 7.
 Rows: Each row corresponds to a different e-commerce platform.
 We want to know the average perception across all five aspects of WOM for each platform.
 Calculate the mean of WOM1 to WOM5 (horizontally) for each platform  an overall WOM score for
that platform  Create a new column “WOM_Avg”

T-TEST PROCEDURE IN SPSS

1. Formulate Hypotheses
 H0: The population mean of WOM is smaller or equal to 5 (μ ≤ 5).
 H1: The population mean of WOM is greater than 5 (μ > 5).
2. Select Formula and Significance Level
 One-sample t-test.
 Significance Level: Typically set to 0.05.
3. Collect Data
 Compute the WOM_Avg for each platform.
 Step in SPSS:
o Go to Transform  Compute Variable...
o Numeric Expression: (WOM1 + WOM2 + WOM3 + WOM4 + WOM5) / 5
4. Run t-test
 Step in SPSS:
o Go to Analyze  Compare Means  One-Sample T Test...
o Select WOM_Avg as the test variable.
o Set Test Value to 5, Confidence level to 95%
o Click OK to run the t-test.
5. Output
One-Sample Statistics

N Mean Std. Deviation Std. Error Mean

WOM_Avg 100 5.0140 1.28747 .12875

 The sample mean of WOM_Avg (μ) is 5.0140, which is just slightly above the test value of 5.
 One-Sample Test
Test Value = 5
Mean 95% Confidence Interval of the Difference
t df Sig. (2-tailed)
Difference Lower Upper
WOM_Total .109 99 .914 .01400 -.2415 .2695

6. Make Decision: 2 approaches

 t-value Approach: Compare t-value = 0.109 with the critical value (α = 0.05 and df = 99).
 p-Value Approach: Compare the p-value (0.914) with the significance level (0.05).
o Since 0.914 > 0.05, we fail to reject H0.
7. Conclude
 We do not have enough evidence to reject H0 that the mean WOM score is equal to 5 (H0 can be
true). Therefore, we cannot conclude that the mean WOM score exceeds 5 (H1 can be false).
 The manager thinks that WOM may heavily influence consumer online purchases. The Likert scale
responses (1 = Totally Disagree to 7 = Totally Agree) would be the customers' responses to the
statement "WOM may influence heavily on consumer in online purchase."
1 2 3 4 5 6 7
|--------|--------|--------|--------|--------|--------|
Totally disagree Neutral Totally agree
 A score of 5 suggests a slight agreement. WOM scores are not significantly higher than 5, indicating
that consumers somewhat agree with the statement that WOM influences their online purchases,
but the influence is not as strong as the manager hypothesized.
Question 7: Check whether repurchase intention and perceived information quality are statistically
different or not across gender.
 Conduct two separate tests:
a) Check whether repurchase intention is statistically different across gender.
b) Check whether perceived information quality is statistically different across gender.

a) Check whether repurchase intention is statistically different across gender.

UNDERSTANDING THE DATA

 Columns RP1 to RP4: Represent the scores given to each platform across four aspects of
repurchase intention. Each aspect is measured on a Likert scale from 1 to 7.
 Column Gender: Represents the gender of participants (1 = Male; 2 = Female).
 We want to know the average perception across all four aspects of repurchase intention for each platform.
 Calculate the mean of RP1 to RP4 (horizontally) for each platform  an overall repurchase intention
score for that platform  Create a new column “RP_Avg”.

T-TEST PROCEDURE IN SPSS

1. Formulate Hypotheses
 H0: The population mean of repurchase intention (RP) is equal for both genders (μ1 = μ2).
 H1: The population mean of repurchase intention (RP) is different for each gender (μ1 ≠ μ2).
2. Select Formula and Significance Level
 Independent Samples t-test.
 Significance Level: Typically set to 0.05.
3. Collect Data
 Compute the RP_Avg for each platform.
 Step in SPSS:
o Go to Transform  Compute Variable...
o Numeric Expression: (RP1 + RP2 + RP3 + RP4) / 4
4. Calculate t Statistic
 Step in SPSS:
o Go to Analyze  Compare Means  One-Sample T Test...
o Test Variable: Move RP_avg to the Test Variable(s) box.
o Grouping Variable: Move Gender to the Grouping Variable box.
o Click Define Groups: enter 1 for Group 1 (Male) and 2 for Group 2 (Female)
o Click OK to run the t-test.
 5. Output
Group statistics

Gender N Mean Std. Deviation Std. Error Mean

1.0 28 5.1875 1.32746 .25087
RP_Avg
2.0 72 4.9028 1.16398 .13718

Independent Samples Test

Levene's Test for
Equality of t-test for Equality of Means
Variances
` 95% Confidence
Sig. (2- Mean Std. Error Interval of the
F Sig. t df Difference
tailed) Difference Difference
Lower Upper
Equal variances
.003 .953 1.055 98 .294 .28472 .26976 -.25061 .82005
assumed
RP_Avg
Equal variances
.996 44.062 .325 .28472 .28592 -.29149 .86094
not assumed

6. Make Decision: 2 steps

Step 1: Levene's Test - Test for Variance Differences Between Two Groups
 Sig < 0.05: Equal variances not assumed
 Sig > 0.05: Equal variances assumed
Step 2: t-test - Test for Mean Differences Between Two Groups
 Sig < 0.05: Reject H0
 Sig > 0.05: Fail to reject H0

In this problem:
Step 1: Levene's Test: Sig.= 0.953 > 0.05  Equal variances assumed
Step 2: t-test: Sig.= 0.325 > 0.05  Fail to reject H0

7. Conclude
 We do not have enough evidence to reject H0 that the mean repurchase intention (RP) score is
the same for both genders. Therefore, we cannot conclude that the mean repurchase intention
score differs between male and female consumers.
 This indicates that gender does not significantly influence repurchase intention.
 b) Check whether perceived information quality is statistically different across gender.

T-TEST PROCEDURE IN SPSS

1. Formulate Hypotheses
 H0: The population mean of information quality (IQ) is equal for both genders (μ1 = μ2).
 H1: The population mean of information quality (IQ) is different for each gender (μ1 ≠ μ2).
2. Select Formula and Significance Level
 Independent Samples t-test.
 Significance Level: Typically set to 0.05.
3. Collect Data
 Compute the RP_Avg for each platform.
 Step in SPSS:
o Go to Transform  Compute Variable...
o Numeric Expression: (IQ1 + IQ2 + IQ3 + IQ4 + IQ5) / 5
4. Calculate t Statistic
 Step in SPSS:
o Go to Analyze  Compare Means  One-Sample T Test...
o Test Variable: Move IQ_avg to the Test Variable(s) box.
o Grouping Variable: Move Gender to the Grouping Variable box.
o Click Define Groups: enter 1 for Group 1 (Male) and 2 for Group 2 (Female)
o Click OK to run the t-test.
5. Output
Independent Samples Test
Levene's Test for
Equality of t-test for Equality of Means
Variances
` 95% Confidence
Sig. (2- Mean Std. Error Interval of the
F Sig. t df Difference
tailed) Difference Difference
Lower Upper
Equal variances
.747 .389 -.014 98 .988 -.00357 .24700 -.49373 .48659
assumed
RP_Avg
Equal variances
-.013 42.981 .989 -.00357 .26549 -.53900 .53185
not assumed

6. Make Decision: 2 steps

Step 1: Levene's Test: Sig.= 0.389 > 0.05  Equal variances assumed
Step 2: t-test: Sig.= 0.989 > 0.05  Fail to reject H0
7. Conclude
 We do not have enough evidence to reject H0 that the mean perceived information quality (IQ)
score is the same for both genders. Therefore, we cannot conclude that the mean information
quality score differs between male and female consumers.
 This indicates that gender does not significantly influence information quality.