CIS6000 – 20197310
3.11. Pilot Survey Data Analysis
3.11.1 Data Coding
To analyse the collected data, the responses to 5-point Likert scale questions were
coded. Values ranging from 1 to 5 were used to code giving the responses a meaningful
numeric value as depicted below.
Table 3-3 Pilot Data Coding
Responses for the questions in the section for variable Perceived Risk was reversed to
take to the same meaning and suit the variable as the questions were worded in the
positive direction in the survey.
3.11.2 Reliability Analysis
Reliability was analysed to understand if the collection of questions or measurements
per variable consistently reflect the variable. This analysis was done using Cronbach’s
alpha. Further, only the variables with an alpha above 0.70 were considered reliable
according to the standards. Reliability analysis was conducted on every variable
discussed in the study.
51
CIS6000 – 20197310
3.11.2.1 Perceived Ease of Use
Data collected for the variable perceived ease of use initially gave an alpha of 0.781.
Further, eliminating question number 3 which was about comfortably using these
platforms without complications gave an alpha of 0.785 which was considered reliable.
The author has decided not to take forward question number 3 to the final survey due
to the reasons of eliminating it increases the reliability and since the length of the
questionnaire was found to be a weak point.
Figure 3-5 Reliability Statistics - Perceived Ease of Use (Before)
Figure 3-6 Item Total Statistics - Perceived Ease of Use
Figure 3-4 Reliability Statistics - Perceived Ease of Use (After)
52
CIS6000 – 20197310
3.11.2.2 Perceived Usefulness
An alpha value of 0.799 was initially identified for the perceived usefulness variable.
But, after removing question number 6 which was about the acceptance among
merchants, and shops, the variable had an alpha of 0.808 which was considered reliable.
It was decided to eliminate question number 6 from the final questionnaire due to the
reasons of increasing reliability and reducing the impact of questionnaire length.
Figure 3-7 Reliability Statistics - Perceived Usefulness (Before)
Figure 3-9 Item Total Statistics - Perceived Usefulness
Figure 3-8 Reliability Statistics - Perceived Usefulness (After)
53
CIS6000 – 20197310
3.11.2.3 Social Influence
An alpha value of 0.854 was found initially for the social influence variable and after
omitting question number 8 which was about the use of these platforms by people
around, the alpha value was 0.868 which was considered reliable.
It was decided to omit question number 8 because of the increase of reliability and
minimising the effect of the questionnaire length.
Figure 3-10 Reliability Statistics - Social Influence (Before)
Figure 3-11 Item Total Statistics - Social Influence
Figure 3-12 Reliability Statistics - Social Influence (After)
54
CIS6000 – 20197310
3.11.2.4 Perceived Risk
The alpha value for the perceived risk variable was 0.834 which was considered
reliable.
Figure 3-13 Reliability Statistics - Perceived Risk
3.11.2.5 Trust
The variable trust had an alpha of 0.752 which was considered reliable.
Figure 3-14 Reliability Statistics - Trust
3.11.2.6 Awareness
The alpha for the variable awareness was 0.816 according to the pilot survey data,
which was considered reliable.
Figure 3-15 Reliability Statistics - Awareness
55
