Sunteți pe pagina 1din 11

AM521412

Scientific theory and data analysis, autumn 2017 Submitted by : Subash Adhikari

Statistical assignment 1


A simple model of relationships between four variables in the context of customers of web-based product providers (e-tailors) is given
below. The variables are:
E-tail repurchase intention,
Service quality,
Community network involvement (web-based)
and Product consumption satisfaction.
The variable Community network involvement is a dummy variable with the value 0 if the customer has been inactive on the online
network, and value 1 if the customer has been an active online network user (communicating both with representatives of the firm and
other customers).

Based on survey data collected from 225 customers of one particular online provider of health related products, you are asked to do an
analysis of this model.

The SPSS data file e-tail-data.sav contains 14 indicators (questions) measured on a Likert scale from 1 to 7, and 1 nominal (dummy) variable.
All the indicators are described in the data file.
All the following tasks must be done:









1. Do a descriptive analysis and an inspection of the appropriateness of the data. If necessary, clean the data.

Solution

Inspection of frequency test of each variable is necessary in order to check the whether there is any errors or not in the data.



As assigned in Likert scale the measure is from 1 to 7 so the value of variable should not exceed 7. We have to inspect thoroughly in each
variable to check if there is any error or not (that means whether the values lie between 1-7 or not). So, as we see in the table that some
variables have maximum value of 9. This revels that there exists an error.

So, after checking each and every variable we can find that respondent no 82 and 200 have got wrong variable .So we need to clear those
wrong or error variable and again run the frequency table .


2

After running the frequency table again with omitting errors the table is now right. So now we can do descriptive analysis.

Discriptive Analysis




The prime motive of using descriptive analysis is to summarize the data. As we can see minimum value, maximum value ,mean and
standard deviation .Though the total number of respondent was supposed to be 225 but we can see the following chart has 223 number of
respondent .This is because we omitted /clear the 2 respondent because of error in max/min value .

3



2) Do a factor and reliability analysis of the 13 indicators measured on a scale from 1-7 (NB! use them all in one factor model, do not
split up). Describe, comment and conclude!

Solution






As we can see in the result that Kaiser-Meyer-Olkin Measure of Sampling have the value of 0.896 and Bartlitts Test of Sphericity is (P=.000)
this signifies that the following result of factor analysis is appropriate. Because the value of Kaiser-Meyer-Olkin Measure of Sampling should
be 0.0 or above as the result is above the assigned its alright similarly Bartlitts Test of Sphericity should be 0.5 or smaller to that. so, both
the outcome are appropriate.

4




We can see that there is weak correlation in variable number 15 because it is supposed to be more than 0.5 measure. So to show the strong
coorelation we need to remove the variable in 15 and analyze again.




So after omitting the variable 15 and computing again we can see that variable below has shown strong correlations in the above image.

6



Rotated component matrix is the key output of principal components analysis. It contains estimates of the correlations between each of the
variables and the estimated components. So here we can see the component 1,2 and 3 and there are several variables that falls under
these components. In the table we can see that there are several factors that are each strongly loaded . We consider those factors strongly
loaded of they are above .4 on the first two component. Hence, we can see that each and every factors are strongly co related with the 3
components provided. The component 1 satisfaction have 9 factors component 2(service quality) have 4 factors and component 3(
repurchase intention ) have 5 factors .

7

The Reliability Analysis procedure calculates a number of commonly used measures of scale reliability and also provides information about
the relationships between individual items in the scale. From the above rotated component matrix, we have to segregate the three
components into respective related variables. To start the analysis, begin by CLICKING on the Analyze menu, select the Scale option, and
then the Reliability Analysis sub-option.Then by naming the any component in scale variable we can allocate variables that have correlation
into the items.






The first table we need to look at in an output is the Reliability Statistics table. This gives Cronbachs alpha coefficient. We are looking for a
score of over .7 for high internal consistency. In this case component SATSFACTION, RE-PURCHASE INTENTION AND SERVICE QUALITY has
value = .881,0.847,0.854 respectively which shows the components are reliable .
8


3. Prepare the data for hypothesis testing by doing a data reduction (make summated scales
based on the result from the factor model). Explain briefly what you have done!

Solution




As per the question we are required to reduce the data and prepare the data for hypothesis. Taking reference of the question number 2 we
have segregated three super variable. Then compiled those 3 variable with other variable that have correlation with them .We have
computed variable using the mean by indicating target variable (eg satisfaction ).This result in reduction of the data making summated scale
based on the result from factor mode.



9


4. Test for the difference in repurchase intention mean value between customers with high
and low online community network involvement, respectively. Describe, comment and conclude!

Solution

As per the question we are supposed to test for the difference in repurchase intention mean value between customers with high
and low online community network involvement, respectively. So we are using independent-samples t-test because it is more commonly
used for two groups and here we are comparing between two group (i.e with high and low online community network involvement). To run
an Independent Samples t Test in SPSS, click Analyze > Compare Means > Independent-Samples T Test. In Test variable we have to put
REPURCHASE variable and dummy variable in grouping variable.(specify value as 0 and 1 as value in grouping variable).


10


As we can see, there are two parts that provide different pieces of information: (A) Levenes Test for Equality of Variances and (B) t-test for
Equality of Means. The significance level in levenes Test for equity of variance is 0.396 whereas in t-test for equity of means there is
significance level of 0.001. As per the assumption p > 0.05 we can see that the equity of variance is greater than 0.05 (p>0.05) so we cant
reject the null hypothesis. It reveals that the variance in repurchase intention of low involvement is not significantly different than
that of high involvement group. Whereas significance for equity of means is less than 0.05(p <0.05) so we can reject the hypothesis and it
reveals that there is difference between the Means. This revels that, the mean score of high involvement group is higher than that of low
involvement group.

11

S-ar putea să vă placă și