Download as docx, pdf, or txt
Download as docx, pdf, or txt
You are on page 1of 14

Quantitative Analysis:

Data Collection & Questionnaire Development:


Data was collected using non-probability convenience sampling via a questionnaire
survey form. Each Likert scale question in the survey belonged to a specific construct.
For each construct, a specific set of questions were developed to analyze and assess
the likelihood of the respondents for each of the constructs.

A 7-point Likert scale was used for each of these constructs. For each construct,
participants would suggest their agreeableness to the desired motive based on the
questions. The responses were then added to calculate their total score.

We used a seven-point Likert scale to test each question for the constructs. The seven-
point Likert scale shows a higher level of accuracy because of increased sensitivity.
Moreover, it is easier to use and helps the researcher gauge a better understanding of
the respondent's evaluation of questions.

Our questionnaire consisted of 29 questions. 4 questions were included on the


construct of intrinsic motivation and engagement. 3 questions were included on the
construct of satisfaction. 2 questions were included on the construct of intention,
interest, feedback, perceived usefulness, and perceived ease of use. 1 question
was included on the construct of self-efficacy. (Refer to appendix for the entire
questionnaire.)

To ensure satisfactory and reliable results, the questionnaire was circulated around the
age group of 18-25. Moreover, in order to reach a conclusion for our hypothesis, we
used validated scales to ensure the accuracy of survey results.

The questions at the beginning of the survey were used as filter questions in
determining whether the respondent met the criteria to fill the survey or not. Later on,
the questions became more specific and were meant to consciously target the insights
of respondents, to gauge a better understanding of their online learning experience and
perception. Different sets of questions pointed to different constructs of our study.

This helped us gain a meaningful understanding of how each construct was related to
students’ online learning experiences. Below are the results of our quantitative analysis
which emerged after conducting basic and advanced tests through SPSS.
Sampling Design:
● Population Description – The target population for this study were all the
students who had experienced E-learning.

● Sample Unit/Element – Undergraduate students

● Sampling Frame – 176 respondents/students. The scope of this study was


limited to only participants from Pakistan. The responses were collected online
from different cities in Pakistan, where the majority of the students belonged to
universities in Karachi. Moreover, the students belonged to a wide range of
universities however, majority of them belonged to IBA. The students also
belonged to different program of study, however, majority of them were pursuing
BBA.

● Sampling Method – For this study, due to time and financial limitations, we used
non-probability sampling in the form of convenience sampling. Our goal was to
collect the maximum number of responses for our online survey.

● Response Rate and Final Sample Size – A total of 176 online responses were
collected during a period of less than 1 month.

● Respondent Profile – all respondents had quite similar profiles.

● Fieldwork – The most important part of the study was the circulation of online
responses through social media platforms such as WhatsApp, Messenger, and
on different public forums on Facebook. Once the responses were collected, we
performed our quantitative analysis which was used to conclude our findings.

Descriptive Analysis
The graph, tables, and pie charts below demonstrate the demographics of our
respondents. Our questionnaire received 176 responses, out of which 36.4% were
males and 62.9% were females. Nearly half of our respondents belonged to the age
group of 21-22 (58.3%), 36.4% of the respondents belonged to the age group category
of 18-20 and the remaining belonged to the category of 23-25. Since our target sample
size was 18-25 years of age, the respondents proved to be the right representation of
our population. The responses were collected online from different cities in Pakistan,
where the majority of the students belonged to universities in Karachi. Moreover, the
students belonged to a wide range of universities however, majority of them belonged to
IBA. The students also belonged to different program of study, however, majority of
them were pursuing BBA.
Factor Analysis
The SPSS software was used to interpret and analyze the gathered data. There were
no missing variables however, there were a few outliers.
KMO and Bartlett’s Test:
In order to check the adequacy of the sample size, we conducted the KMO and
Bartlett’s Test. The KMO value was greater than 0.6 (i.e. 0.789) for our 176 respondents
and the significance of Bartlett’s Test of Sphericity was less than 0.05 (i.e. <0.001).

Hence, it is satisfactory for our research and is significant. Therefore, our factor analysis
is appropriate.
Variance Analysis
We can see from the below-attached results of variances, that the SPSS software has
extracted the initial 2 eigenvalues (factors or components), and their cumulative
percentage was 54.144%. Thus, based on the given data, it can be concluded that the
initial 2 factors explain 54.144% of the total variances.

Scree Plot
The scree plot shows 8 components in the downward curve shown below. The slope is
leveling off at the eigenvalue of ‘1’. Only component ‘one’ should be retained in the
given data in the scree plot because it has an eigenvalue score above one.

Component Matrix
Most of these components in the component matrix illustrate a significant value which is
greater than 0.4 on the first component. However, very few items load on the remainder
of the components which further supports our conclusion from the scree plot.

Reliability Test
A reliability test was conducted to evaluate and examine if each of the constructs
formed are internally consistent or not. This was analyzed by checking the Cronbach
Alpha’s value of each construct. In order to be acceptable, the value of Cronbach Alpha
must be greater than 0.7.

Interest

The Cronbach's Alpha value of 0.056 shows that the scale used is not reliable. The
number of items on the scale is 2, and the correlation is not internally consistent and
reliable as the Cronbach’s Alpha value is less than 0.7.

Feedback

The Cronbach's Alpha value of 0.423 shows that the scale used is not reliable. The
number of items on the scale is 2, and the correlation is not internally consistent and
reliable as the Cronbach’s Alpha value is less than 0.7.

Perceived usefulness

The Cronbach's Alpha value of 0.480 shows that the scale used is not reliable. The
number of items on the scale is 2, and the correlation is not internally consistent and
reliable as the Cronbach’s Alpha value is less than 0.7.
Perceived ease of use

The Cronbach's Alpha value of 0.714 shows that the scale used is reliable. The number
of items on the scale is 2, and the correlation is internally consistent and reliable as the
Cronbach’s Alpha value is greater than 0.7.

Satisfaction

The Cronbach's Alpha value of 0.900 shows that the scale used is reliable. The number
of items on the scale is 3, and the correlation is internally consistent and reliable as the
Cronbach’s Alpha value is greater than 0.7.
Intention

The Cronbach's Alpha value of 0.756 shows that the scale used is reliable. The number
of items on the scale is 2, and the correlation is internally consistent and reliable as the
Cronbach’s Alpha value is greater than 0.7.

Engagement

The Cronbach's Alpha value of 0.577 shows that the scale used is not reliable. The
number of items on the scale is 4, and the correlation is not internally consistent or
reliable as the Cronbach’s Alpha value is less than 0.7.
Intrinsic motivation

The Cronbach's Alpha value of 0.643 shows that the scale used is not reliable. The
number of items on the scale is 4, and the correlation is internally not consistent and not
reliable as the Cronbach’s Alpha value is less than 0.7.

Cross Tabulation Analysis


Age-Gender Tabulation
The data below illustrates that there was a wide gap in the proportion of female
participants to the proportion of male participants. There were more (nearly double the
amount) of female respondents than male. For each age category of age groups, the
number of female participants were greater than male participants. Participants
belonging to age group ‘18-20’ and ‘21-22’ were in more than double of the proportion of
the participants belonging to age group ‘23-25’.
Age-Program of study Tabulation
The data below illustrates that there was a wide gap in the proportion of junior year
participants and the rest of the participants. There were more (nearly double the
amount) of junior level respondents than sophomores. In addition, the amount of
respondents in junior year were more than double the amount as compared to freshmen
and seniors.

You might also like