Study of online store customer behaviour

The primary aim was to analyze the sample data to highlight key aspects about the usage behavior of customers. The available sample data has been collected from 50 customers and includes information regarding age, gender, time and money spent on navigation, and errors. The browsing time for each customer on average is not different from the claimed 12 minutes. Besides, there does not seem to exist any significant association between the browsing time and the spent amount. Navigation errors seem to be driven by the underlying time, owing to which a significant difference in average errors has been concluded. The gender does not influence the browsing time for the customer. Also, the age of the customer and the average spent amount tend to highlight a positive correlation. Besides, the average errors in navigation by a customer tend to vary from the claimed value of 10.

RESEARCH METHODS
STUDENT ID:
[Pick the date]

(1) The respective hypotheses are highlighted as shown below.
(a) Claim to test: Average time spent on web page browsing is 12 minutes.
H0: The average time on web browsing does not differ significantly from 12 minutes.
H1: The average time on web browsing does differ significantly from 12 minutes
Claim to test: Average expenditure and time spent on web page browsing show statistically
significant linear association.
H0: Correlation coefficient is insignificant and hence can be assumed to be zero.
H1: Correlation coefficient is significant and hence cannot be assumed to be zero.
(b) Claim to test: The average navigations errors are not the same for the two timings.
H0: The navigation errors during AM and PM time do not differ in any significant manner.
H1: The navigation errors during AM and PM time do differ in significant manner.
(c) Claim to test: The average time spent on browsing tends to be different for the two
genders.
H0: The average time spent on browsing does not significantly differ for the two genders.
H1: The average time spent on browsing does significantly differ for the two genders.
(d) Claim to test: There is significant association between the spending and the age of the
customer
H0: The customer age and average expenditure do not show any significant association and
hance the correlation coefficient can be assumed as zero.
H1: The customer age and average expenditure does show significant association and hance
the correlation coefficient cannot be assumed as zero.
(2) Claim to test: Average time spent on web browsing by a customer is 10 minutes.
1

(3) The appropriate statistical tests which ought to eb conducted for testing of the above
hypothesis are listed below.
(4) The relevant results of the various hypothesis test as obtained through the aid of SPSS
software are summarised below. The level of significance is taken as 5% for each of the
following tests.
(a)
The p value > level of significance and hence there is no rejection of null hypothesis. The
original claim is supported by the sample data.
2

The p value < level of significance and hence there is rejection of null hypothesis. The
original claim is supported by the sample data.
(b)
The p value < level of significance and hence there is rejection of null hypothesis. The
original claim is supported by the sample data.
3

(c)
The p value> level of significance and hence there is no rejection of null hypothesis. The
original claim is not supported by the sample data.
(d)
The p value < level of significance and hence there is rejection of null hypothesis. The
original claim is supported by the sample data.
4

(e)
The p value < level of significance and hence there is rejection of null hypothesis. The
original claim is refuted by the sample data.
5

(5) Title: Study of online store customer behaviour
(6) Abstract: The primary aim was to analyse the sample data so as to highlight key aspects
about the usage behaviour of customers. The available sample data has been collected from
50 customers and includes information regards age, gender besides time and money spent on
navigation coupled with errors. The browsing time for each customer on an average is not
different from the claimed 12 minutes. Besides, there does not seem to exist any significant
association between the browsing time and the spent amount. Navigation errors seem to be
driven by the underlying time owing to which significant difference in average errors has
been concluded. The gender does not influence the browsing time for the customer. Also,
age of customer and the average spent amount tend to highlight a positive correlation.
Besides, the average errors in navigation by a customer tend to vary from the claimed value
of 10.
(7) The sampling technique choice is pivotal with regards to deriving population
characteristics using the sample data as an unrepresentative sample can skew the results and
adversely impact reliability. As a result, it is apparent that choosing a probability based
sampling technique would be preferable over non-probability based techniques. This would
help in minimising bias. Owing to the consumer behaviour being driven by key factors such
as age and gender, the simple random sampling would not be a suitable choice as the
underlying sample that is selected may be biased. This could lead to reduction in reliability
of results derived from statistical analysis (Lieberman et. al., 2013). Therefore, it is preferable
that s stratified random sampling technique is used for selection of the sample. In this
method, various groups are formed based on the key attributes and then random sampling is
used to select the respective sample while ensuring that their representation in the sample
closely mirrors the actual population of interest (Medhi, 2016).
(8) The initial step for carrying out the above study would be to study the literature and
highlight the research questions which would be addressed through the study. The second
step is to select the suitable sample with the aid of the stratified random sampling so that a
representative sample can be obtained. For testing the various hypotheses framed for
answering the research questions, data would be collected from the pilot study customers.
The next step would be to process and clean this raw data so that it can be directly used for
statistical analysis (Hastie, Tibshirani and Friedman, 2014). Using the data obtained and the
underlying hypothesis, it needs to be ascertained if the hypothesis support the claim or refute
6

the same. Based on this, conclusion and discussion would arise. Additionally, the study
limitations are pointed and possible follow up studies that would be helpful on the subject
should be mentioned (Lehman and Romano, 2016).
7

References
Hastie, T., Tibshirani, R. and Friedman, J. (2014) The Elements of Statistical Learning. 4th
ed. New York: Springer Publications.
Lehman, L. E. and Romano, P. J. (2016) Testing Statistical Hypotheses. 3rd ed. Berlin :
Springer Science & Business Media.
Lieberman, F. J., Nag, B., Hiller, F.S. and Basu, P. (2013) Introduction To Operations
Research. 5th ed. New Delhi: Tata McGraw Hill Publishers.
Medhi, J. (2016) Statistical Methods: An Introductory Text. 4th ed. Sydney: New Age
International.
8