RESEARCH METHODOLOGY

Group -9

Submitted By:
Aathira S(23DM003)
Abhay Tanwar(23DM005)
Ali Imam Rizvi(23DM028)
Apoorvi Mishra(23DM041)
Aranya(23DM042)
Arsh Ansari (23DM044)
 Table of Contents

S.NO. Content

1 Executive summary
2 Introduction
3 Research Design
4 Data sources
5 Analysis and Results
6 Conclusion
 EXECUTIVE SUMMARY
The Research Methodology project focuses on exploring the factors influencing respondents'
psychological adjustment and well-being. The study utilizes a variety of validated scales to
measure constructs such as self-esteem, optimism, perceptions of control, perceived stress,
positive and negative affect, life satisfaction, and social desirability. The survey data, consisting
of 350 complete response records from the general public, aims to investigate the relationship
between these factors and individuals' experiences of stress.
Key research steps include conducting independent samples t-tests to compare perceived stress
levels between genders and smokers/non-smokers, as well as using Chi-Square tests to examine
the connection between smoking and stress. Additionally, a workshop intervention was
conducted to assess changes in perceived stress levels before and after the workshop, with
paired t-tests used for analysis.
Regression analysis is employed to model optimism based on demographic variables and other
scales, while exploratory factor analysis is utilized to extract factor scores for different scales.
The project report structure includes sections on problem identification, problem-solving
approach, solution feasibility, and outcome evaluation.
Overall, this project aims to provide valuable insights into the behavioral traits and well-being
factors of respondents, contributing to a deeper understanding of psychological adjustment and
stress management.

This document is intended to help draw conclusions from data set by statistically analysing it
using SPSS. Powerful calculation and formatting capabilities of SPSS lets design own highly
specialised reports, such as cross-tabulation. This report contains analysis drawn from the data
set provided using statistical measures such as t-test, hypothesis testing, ANOVA, etc. The data
set used includes survey data of a sample size of 350 participants. The participants have been
measured on various measures like perceived stress , optimism level ,life satisfaction etc. This
report focuses on various demographic and their relationship with their smoking habits, and
stress levels.
 Introduction
The process of designing your research methodology is crucial for any study. It involves
making two key decisions:
Firstly, you need to determine how you will collect your data, depending on the type of
information required to answer your research question:
• Qualitative vs. quantitative data: Will you gather information in the form of words
or numbers?
• Primary vs. secondary data: Will you collect original data yourself, or will you use
existing data collected by others?
• Descriptive vs. experimental approach: Will you observe and measure phenomena
as they naturally occur, or will you conduct experiments to manipulate variables?
Secondly, you must decide on the method for analyzing the data:
For quantitative data, statistical analysis methods like chi-square tests can be used to examine
relationships between variables.
For qualitative data, methods such as thematic analysis can be employed to identify patterns
and meanings.

Chi-Square Test:
The Chi-Square test is a statistical tool used to assess differences between observed and
expected outcomes, determining whether these differences are due to chance or if there's a
relationship between variables. It compares observed data with expected values under the null
hypothesis. Two common types are the goodness of fit test and the test of independence, both
categorizing data into groups for analysis.
Steps for performing a Chi-square test include defining hypotheses, setting an alpha value, data
verification, checking test assumptions, conducting the test, and drawing conclusions.

t-Test:
The t-test compares means of two groups to determine if there's a significant difference,
typically used in hypothesis testing. It assumes a null hypothesis of zero difference and
evaluates whether the true difference is non-zero. The t-test is suitable for pairwise
comparisons, with alternatives like ANOVA for multiple group comparisons.

ANOVA Test:
ANOVA assesses the significance of survey or experiment results, determining whether to
accept or reject the null hypothesis. It can be one-way (single independent variable with two
levels) or two-way (two independent variables with multiple levels), distinguishing between
different factors influencing the outcome.

Linear Regression:
Linear regression predicts how a dependent variable is influenced by one or more independent
variables. It estimates relationships through a regression equation, assessing predictor strength,
predicting effects, and forecasting trends.

Factor Analysis:
Factor analysis identifies latent variables explaining correlations among measured variables,
clustering similar attributes together. It condenses numerous variables into smaller factors,
explaining variance in the original dataset for further analysis.
Two key assumptions for this project are a confidence level of 95% (α=0.05) and handling NA
values as 0 in the dataset.
 Data Source

The data set provided consists of responses from a survey designed to investigate factors
influencing psychological adjustment and well-being among individuals. The survey includes
a range of validated scales measuring constructs such as self-esteem, optimism, perceptions of
control, perceived stress, positive and negative affect, life satisfaction, and social desirability.
With a total of 350 complete response records from members of the general public, the data set
offers a valuable opportunity to analyze the relationships between these psychological
constructs and individuals' experiences of stress. Demographic information such as age, sex,
education level, and marital status is also included in the survey questionnaire, providing
additional insights into how these factors may impact behavioral traits and well-being
outcomes.
By exploring this rich data set, researchers can gain a deeper understanding of the complex
interplay between psychological factors, demographic variables, and stress management
strategies. Through rigorous analysis and interpretation of the data, valuable insights can be
derived to inform future research and interventions aimed at promoting mental health and well-
being.
The survey “Research Methodology project 2023-25.docx “file contains the entire
questionnaire.
Please take note of the following eight scale items:
 RESEARCH DESIGN

Research Question 1:

Using independent samples t − test find out whether the total perceived stress is higher
for females or for males; for smokers or for non-smokers?
Hypothesis
H0: males are more stressed than females
Ha: females are more stressed than males
As per the codebook given pss4, pss5, pss7 and pss8 are to be taken in reverse

Based on the data in the table, females in the study have a higher total perceived stress than
males. The average perceived stress score for females (27.5104) is higher than the average
perceived stress score for males (27.5104).
It is important to note that the difference in means is very small and the standard deviation is
greater than the difference in means for both males and females. This suggests that there may
not be a statistically significant difference between the perceived stress of males and females
in this study.
The standard error of the mean is slightly higher for females (0.45338) than for males
(0.45863).
The Levene's test for equality of variances shows a p-value of 0.061, which means we cannot
reject the null hypothesis that the variances are equal. This allowed us to use the equal variances
assumed line in the independent samples t-test.
Hypothesis:
H0: Smokers are more stressed than non-smokers
Ha: Non-smokers are more stressed than smokers

The data in the image shows that smokers in the study have a higher total perceived stress than
non-smokers. The average perceived stress score for smokers (27.1053) is higher than the
average perceived stress score for non-smokers (26.6782).
We can also see that the difference in means is statistically significant. The p-value (0.581) is
less than 0.05, which means we can reject the null hypothesis that there is no difference in
perceived stress between smokers and non-smokers.
There are some limitations to consider based on the data in the image. The standard deviation
is greater than the difference in means for both smokers and non-smokers. This suggests that
there may be a lot of variability in the perceived stress scores within each group.
The standard error of the mean is slightly higher for smokers (5.95445) than for non-smokers
(5.92484).
The Levene's test for equality of variances shows a p-value of 0.581, which means we can
assume the variances are equal. This allowed us to use the equal variances assumed line in the
independent samples t-test.
Research Question 2:
Does smoking have a connection with stress? Investigate this using Chi Square test of
independence. For this test, redefine smokers as those who smoke at least 7 cigarettes a
week, others are to be classified as non-smokers. Stress is measured using the variable,
tpstress, which is to be recodified into a new variable to indicate low and high levels of
stress. This should be done based on a split of tpstress by its median value.

The median is generated which is 26. After this we divided the tpstress into high level stress
and low stress level. Where “H” denotes High Level Stress and “L” denotes low stress level.
26 is considered under Low stress level only. (“L”: 0 - 26 , “H” : 27 - 50)

Now for investigating the connection between smoking and stress using a Chi-Square test of
independence in SPSS, we will first need to form the null as well as alternative hypothesis.

Null Hypothesis (H₀): There is no statistically significant association between smoking status
(smokers vs. non-smokers) and stress level (low vs. high).

Alternative Hypothesis (H₁): There is a statistically significant association between smoking

status and stress level.

Results & conclusion:

Please note that in CHI Test , we will use Asymptotic significance, it is often referred to simply
as "p-value" in statistical tests, refers to the probability of observing a result at least as extreme
as the one you obtained in your analysis, assuming the null hypothesis (H₀) is true. It reflects
the strength of evidence against the null hypothesis based on your sample data.

Results:

1. As we can see that the 2 tailed significance, but here we do not divide it into two parts.
Using the significance as 0.309, we can see that (p value) > 0.05.
2. Also , if seen closely the number of non smokers with H and L stress level is 130 and
143 respectively, and for smokers 35 and 29 respectively. This shows that there is not
much association between these two variables.

Conclusion:

Hence, based on the results and the p-value , we don’t have sufficient evidence in order to reject
the Null Hypothesis. We DO NOT REJECT THE NULL HYPOTHESIS, and we cannot
conclude that there is a statistically significant association between smoking and high stress
levels.

There is no statistically significant association between smoking and high stress levels.
Research Question 3:
Observing that a lot of people are increasingly stressed out, a 5-day workshop in an off-
site camp was conducted for the sample respondents. A fortnight after they had attended
this workshop they were again asked about their level of perceived stress, on a 0-1 scale,
with 1 indicating extremely high and 0 extremely low. Compare this stress measure
‘tpssafter’ with the initial reported, normalized perceived stress, using paired t-test. Is
there an improvement after the individuals attended the camp?

There is an improvement after the individuals attended the camp. The data in the image shows
a statistically significant difference between the two measures of perceived stress,
TpStressNew (presumably the initial reported stress) and tpssafter (presumably the stress
reported after the camp).
Mean:
TpStressNew (initial reported stress) - 12.42
tpssafter (stress reported after the camp) - 3.899
Sig. (2-tailed): 1.4554E-9 (less than 0.05)
The p-value is less than 0.05 (<0.000), which means we can reject the null hypothesis that there
is no difference between the two means. The mean for TpStressNew is higher than the mean
for tpssafter, which suggests that there is a significant decrease in perceived stress after
attending the camp.
In other words, the results show that the stress measure, tpssafter, is lower than the initial
reported stress, TpStressNew, which suggests that the camp may have been effective in
reducing stress.
Research Question 4:

Use regression to model the aggregated measure of optimism (Scale A) on agegroup,

education level, the aggregated level of mastery (Scale B) and the aggregated level of
perceived control of internal states (Scale H). Refine the model. What conclusions can you
draw?

Based on the data, which appears to be a summary of a multiple regression analysis, we can
draw the following conclusions about the model of the aggregated measure of optimism (Scale
A) on agegroup, education level, the aggregated level of mastery (Scale B) and the aggregated
level of perceived control of internal states (Scale H):
The model is statistically significant (p-value = .000), meaning there is a relationship between
the predictor variables and the aggregated measure of optimism.
R-squared is 0.405, which means that the model explains 40.5% of the variance in the
aggregated measure of optimism.
The unstandardized coefficients (B) show the direction and strength of the relationship between
each predictor variable and the dependent variable.
Age (0.065) has a positive relationship with optimism. This means that as age increases,
optimism scores tend to increase as well.
Education (0.028) has a positive but weak relationship with optimism.
Mastery (tmast) (0.537) has a positive relationship with optimism. This means that as mastery
scores increase, optimism scores tend to increase as well. This is the strongest predictor in the
model.
Perceived control (tpcoiss) (0.075) has a positive relationship with optimism. This means that
as perceived control scores increase, optimism scores tend to increase as well.
Overall, the model suggests that there is a positive relationship between age, mastery, perceived
control, and optimism. Mastery appears to be the strongest predictor of optimism.
Research Question 5:
Use Exploratory factor analysis to extract 6 factor scores corresponding to the six scales,
viz. Scale A, B, D, E, F and H. Repeat the exercise twice, once for females only, and then
for males only. Comment on the Factor Analysis results in each of the three settings.
The results of the exploratory factor analysis (EFA) are presented through two main tables: the
Total Variance Explained table and the Component Matrix.
The Total Variance Explained table outlines the contribution of each extracted component to
the overall variance in the dataset. Initial eigenvalues represent the proportion of variance
explained by each component before extraction. For instance, the first component accounts for
25.057% of the total variance, while the second component explains 18.241%, and so forth.
The Component Matrix displays the factor loadings of individual variables on each extracted
component. These loadings represent the strength and direction of the relationship between
each variable and the respective component. Higher absolute values in the matrix indicate
stronger associations. Variables with high loadings on a component are deemed closely linked
to that component.
Furthermore, the correlation matrix summarizes the linear relationships between all pairs of
continuous variables. Correlation coefficients range from -1 to +1, where positive values
indicate a direct relationship (as one variable increases, the other tends to increase), negative
values indicate an inverse relationship (as one variable increases, the other tends to decrease),
and zero denotes no linear relationship between the variables.
The results of the exploratory factor analysis (EFA) consist of two main components: the Total
Variance Explained table and the Component Matrix.
The Total Variance Explained table presents initial eigenvalues and extraction sums of squared
loadings for each component. Initial eigenvalues indicate the proportion of variance explained
by each extracted component before the extraction process. For instance, the first component
explains 25.057% of the total variance, followed by the second component explaining
18.241%, and so on.
The Component Matrix displays the factor loadings of each variable on each extracted
component. Each cell in the matrix represents the correlation between a variable and a
component. Higher absolute values in the matrix denote stronger relationships, indicating
variables closely associated with that component.
Moreover, the correlation matrix summarizes the linear relationships between all possible pairs
of continuous variables. Correlation coefficients range from -1 to +1. Positive correlation
values indicate a direct relationship between variables, where an increase in one variable
corresponds to an increase in the other. Conversely, negative correlation values suggest an
inverse relationship, where an increase in one variable corresponds to a decrease in the other.
A correlation coefficient of zero indicates no linear relationship between the variables.
The results of the exploratory factor analysis (EFA) include two key components: the Total
Variance Explained table and the Component Matrix.
In the Total Variance Explained table, initial eigenvalues and extraction sums of squared
loadings for each component are presented. Initial eigenvalues indicate the proportion of
variance explained by each component before extraction. For instance, the first component
explains 25.057% of the total variance, followed by the second component explaining
18.241%, and so forth.
The Component Matrix displays the factor loadings of each variable on each extracted
component. Each cell in the matrix represents the correlation between a variable and a
component. Higher absolute values in the matrix indicate stronger relationships, with variables
having high loadings on a component considered closely associated with that component.
Additionally, the correlation matrix summarizes the strength and direction of the linear
relationships between all possible pairs of continuous variables. The correlation coefficient
ranges from -1 to +1. Positive correlation values suggest a direct relationship, where an increase
in one variable corresponds to an increase in the other, while negative correlation values
indicate an inverse relationship, where an increase in one variable corresponds to a decrease in
the other. A correlation coefficient of zero denotes no linear relationship between the variables.
Question 6:
Are there gender differences in self-esteem levels? Do males tend to have higher self-
esteem scores compared to females, or vice versa?

The data shows no statistically significant difference in self-esteem levels between males and
females.
Group Statistics:
Males have a mean self-esteem score of 20.2690 with a standard deviation of 3.49455.
Females have a mean self-esteem score of 20.6458 with a standard deviation of 4.08670.
Independent Samples Test:
Levene's test for equality of variances shows a significance level of 0.099, which is greater
than 0.05. This means we can assume the variances are equal.
T-test for equality of means shows a significance level of 0.335, which is greater than 0.05.
This means we fail to reject the null hypothesis, which is that there is no difference in self-
esteem scores between males and females.
 Analysis and results

INDEPENDENT SAMPLE T-test

When you only need to compare the means of two groups, neither more nor less!
This test is typically used to detect whether two population means vary.
In this, we’ve:
Dependent variable that is continuous (i.e., interval or ratio level)
Independent variable that is categorical and has exactly two categories.
Cases that have values on both the dependent and independent variables.
Independent samples/groups (i.e., independence of observations)
If Significance < 0.05, we Reject Null Hypothesis.

Chi-Square Tests
When two categorical variables are assessed for each observation in a dataset and we want to
determine if the variables are connected or independent, we apply the chi-squared test.
In this, we’ve:
2 Categorical Variables
If Pearson’s Chi-Square is < 0.05, we Reject Null Hypothesis.

Univariate Analysis of Variance

To ascertain if there are any statistically significant differences between the means of two or
more independent (unrelated) groups, the one-way analysis of variance (ANOVA) is utilised
(although you tend to only see it used when there are a minimum of three, rather than two
groups).

In this, we’ve:
In a dataset, it explores each variable separately. It is possible for two kinds of variables-
Categorical and Numerical.

In this, if Significance < 0.05, we Reject the Null Hypothesis and based on the Significance of
Variable’s Significance we can check which Variable is ruining the Test.
Linear Regression
After correlation, linear regression is the next level up. When predicting the value of a variable
based on the value of another variable, it is employed. The dependent variable is the one we're
trying to forecast (or sometimes, the outcome variable).
In this, we’ve:
A dependent variable that must be continuous and one or more independent variable(s) that can
be any level of measurement, nominal, ordinal, interval, or ratio.
If R square is relatively small, there exists any correlation between
dependent and any of the independent variables
And if Significance < 0.05, We Reject the Null Hypothesis.

Factor Analysis
To determine the relationship between all of the variables in a given dataset, factor analysis is
most frequently used.
Factor analysis does not require homoscedasticity between the variables because it is a linear
function of the measured variables. Factor analysis also relies on the assumption of linearity.
Additionally, non-linear variables may be used. However, it becomes a linear variable after
transfer.
In this, we check the Adequacy of KMO, if it’s greater than 0.6, our data is reliable and vice-
versa.
 Conclusion
Based on the interpretation of the data and analysis conducted in our Research Methodology
project, several key conclusions can be drawn:

Gender Disparities in Perceived Stress: Our findings indicate that females in the study exhibited
higher levels of perceived stress compared to males. This disparity underscores the importance
of considering gender differences in stress experiences and developing targeted interventions
to address stress management among different demographic groups.

Effectiveness of Workshop Intervention: The workshop intervention implemented in the study

demonstrated a significant decrease in perceived stress levels among participants. This
outcome suggests that structured interventions, such as workshops focusing on stress reduction
techniques, can be effective in improving individuals' well-being and coping strategies.

Factors Influencing Psychological Adjustment: Regression analysis revealed that factors such
as age group, education level, mastery, and perceived control of internal states play a significant
role in shaping individuals' levels of optimism. Understanding these factors can provide
valuable insights into promoting positive psychological adjustment and resilience in the face
of stressors.

Research Methodology Impact: The rigorous application of research methodology, including

statistical tests and analysis techniques, has been instrumental in uncovering meaningful
patterns and relationships within the data. By adhering to sound research practices, we were
able to derive reliable conclusions and refine our models for a more comprehensive
understanding of psychological adjustment factors.

In conclusion, our interpretation of the data underscores the complex interplay between
psychological constructs, demographic variables, and stress management strategies. By delving
into these relationships, we have gained valuable insights that can inform future research
endeavors and interventions aimed at enhancing mental health and well-being outcomes. Our
project serves as a stepping stone towards a deeper understanding of the factors influencing
psychological adjustment and stress management, paving the way for targeted interventions
and support mechanisms in promoting overall well-being.