Methods
- Data visualization is a very important aspect in the field of data
analytics. The human brain processes visual information quicker and
more efficiently than mere numbers or reports, and so, a few
visualizations have been used here.
- A Likert Scale is an orderly scale from which respondents choose the
option that best supports their opinion, and has been used to measure
life satisfaction, job satisfaction and beneficiality of the lockdown
as it helps in capturing the intensity of the respondents’ feeling
towards the factor.
- The Kruskal-Wallis test has been extensively used in the study for
analyzing the relationship between various parameters. It is a
non-parametric test used for comparing two or more independent samples
with null hypothesis being the equality of medians of all groups. This
test was opted keeping in mind the non-normality of the data, thus
giving Kruskal-Wallis test an edge over the parametric alternative of
ANOVA. Once the Kruskal-Wallis test has been applied, the next step
would be to identify the particular pairs of samples that have medians
significantly different from each other. Here, Dunn’s Post Hoc test
comes handy. It is one of the non parametric alternatives to Tuckey’s
Post Hoc test. The null hypothesis under Dunn’s test is the absence of
significant difference between the medians of the groups.
- In the study, Mann-Whitney U test has also been used, which is a
non-parametric test of the null hypothesis that the probability that a
randomly selected value from one population is less than a randomly
selected value from a second population is equal to the probability of
being greater. It is the non-parametric alternative to independent
samples t-test.
The Kruskal-Wallis, Dunn’s Post Hoc test and Mann-Whitney U and have
been applied to study the relationships between various parameters such
as occupation, age, income level, marital status, family structure, life
satisfaction, job satisfaction, beneficiality of lockdown ,time spent on
social media and time spent on oneself.
- Wilcoxon Signed-Rank test is a non-parametric statistical hypothesis
test used to compare two related samples, matched samples, or repeated
measurements on a single sample to assess whether their population
mean ranks differ. It is the non-parametric alternative to Paired
Student’s t-test. The test is applicable to non-normal data, which is
the case here. It has been used to study the impact of lockdown on
time spent on oneself and social media, job satisfaction and life
satisfaction.
- The term Level of Significance, denoted by α, is the probability of
rejecting the null hypothesis when it is true. It shows how likely a
pattern in the data is due to chance and is determined before
conducting the experiment. In this study, all statistical tests were
carried out at 5% Level of Significance.
- It was seen that categories of some parameters consisted of a very
small number of data points. Such data points were excluded from
further analysis. These included:
1 respondent opting gender as ‘Other’
1 Divorcee and 3 Widowed respondents
23 respondents not currently residing in India
3 respondents of age less than 15
6 respondents having reported unrealistic hours spent on social media
and oneself
- The parameters ‘Time spent on social media’ and ‘Time spent on
oneself’ were converted into class intervals for the convenience of
analysis. The intervals were-’Less than 1’ , ’1-3’, ’3-5’ and ‘More
than 5’ (in hours) .
- The first three groups under Educational Qualification were pooled
during analysis as the sample size was low.
- All the variables were included in the analysis.
- Data and codes have not been made available for access. Analyses were
performed using the open source programming language Python.
Limitations
- The data was collected online, since conducting a field survey was
difficult, given the situations due to COVID-19. Thus, there is a
definite bias in the data as only those with access to smartphones and
internet connection have been a part of the survey. People from remote
areas might not have been properly represented in the study.
- If an individual or their close ones have been tested COVID-19
positive could be a factor that might bring about interesting results.
- Geographic location of an individual has not been considered as a
factor. Based on whether or not the location is a highly/mildly
COVID-19 infected zone, a significant change in the well-being of
individuals might show up.
- Whether somebody is closely dealing with COVID-19 patients /
quarantined people at the workplace could be another factor that could
bring about interesting insights.