Regression Analysis
Introduction
Regression analysis is a statistical method used to understand the relationship between a dependent
variable and one or more independent variables. It helps predict outcomes and identify trends, making it
a vital tool in fields like finance, economics, and business analytics.
Purpose: State the purpose of regression analysis in research, such as predicting outcomes, identifying
relationships between variables, and informing decision-making.
Regression analysis quantifies the relationship between variables. The key components include:
Dependent Variable (Y): The outcome we aim to predict.
Independent Variable (X): The predictor or factor influencing the dependent variable.
Regression Equation:
Y = β0 + β1X + ε
Y: Dependent variable
X: Independent variable
β0: Intercept
β1: Coefficient of X
ε: Error term
Types of Regression Analysis
1. Simple Linear Regression: Examines the relationship between one independent and one dependent
variable.
2. Multiple Regression: Explores relationships between one dependent and multiple independent
variables.
Applications of Regression Analysis
Regression analysis is widely used in:
Predicting sales based on marketing spend.
Determining factors influencing stock prices.
Understanding the impact of economic indicators on GDP.
Example:
Effect of Internet Speed on Online Shopping Behavior
Objective:
To study the relationship between internet speed, shopping spends, devices connected, ads clicked and
platform [mobile, desktop] (5 independent variables) and monthly income of online shopping
(dependent variable).
Data Collection:
A survey of 10 individuals was conducted.
Regression Model:
Y = β0 + β1 X1 + β2X2+ β3X3+β4X4+β5X5+ ε
Analysis:
Using Excel, the following results were obtained:
Individu Y (monthly X1 (shopping X2(internet X3(devices X4(ads X5{platform
als income) spend) speed) connected) clicked) (mobile=1,
desktop=0}
1 4010 428.3 61 9 1 1
2 702 214.3 24 7 8 0
3 3755 467.55 81 9 12 0
4 4718 500.9 70 8 12 1
5 4996 374.8 30 1 5 0
6 1266 289.9 92 8 7 0
7 4889 474.25 96 8 4 0
8 2827 362.55 84 3 6 1
9 3431 377.75 84 1 13 0
10 697 362.45 97 8 15 1
SUMMARY OUTPUT
Regression Statistics
Multiple R 0.994528
R Square 0.989085
Adjusted R
Square 0.970894
Standard Error 300.8177
Observations 9
ANOVA
Significanc
df SS MS F eF
54.371
Regression 5 24600697.7 4920140 4 0.0038337
Residual 3 271473.885 90491.3
Total 8 24872171.6
Coefficient Standard Upper
s Error t Stat P-value Lower 95% 95% Lower 95.0% Upper 95
0.1250 -
Intercept -1071.99 507.363971 -2.1129 3 -2686.6489 542.6683 2686.648864 542.6683
0.0006
428.3 19.67659 1.28900809 15.2649 1 15.574392 23.77879 15.57439155 23.77878
-
61 -16.62745 4.58566168 -3.626 0.0361 -31.221075 -2.03383 31.22107524 -2.03383
0.0199 -
9 -158.0154 34.7762763 -4.5438 6 -268.68905 -47.3418 268.6890482 -47.3417
0.0301 -
1 -115.2883 29.6351443 -3.8903 2 -209.6006 -20.9761 209.6006034 -20.9760
0.0968 -
1 -561.0875 234.85125 -2.3891 2 -1308.489 186.314 1308.488982 186.3140
Key Coefficients and Interpretations
Intercept (-1071.99): When all independent variables are zero, the baseline predicted income is
negative, indicating no real-world interpretation but serving as a statistical anchor.
X1 (Shopping spend, 19,68): For every unit increase in shopping spend, monthly income
increases by 19.6 units, holding other variables constant.
X2 (Internet speed, -16.63): For every unit increase in internet speed, monthly income decreases
by 16.19 units, which might reflect a counterintuitive relationship or external influencing factors.
X3 (Devices connected, -158.015): As the number of connected devices increases by 1, monthly
income decreases by 163.43 units, suggesting more devices might indicate distractions or non-
productive usage.
X4 (Ads clicked, -115.29): Each additional ad clicked reduces monthly income by 106.6 units,
possibly due to wasted time or impulsive spending.
X5 (Platform, -561.088): Using a mobile platform (vs. desktop) is associated with a reduction in
monthly income by 615.76 units, possibly due to reduced efficiency or usage patterns.
Model Statistics
R-squared (0.989): The model explains 98.9% of the variability in monthly income, indicating
excellent explanatory power.
Adjusted R-squared (0.971): Still very strong, accounting for the number of predictors.
F-statistic (54.37, p=0.0038): The model is statistically significant overall.
Interpretation and Insights
1. Significance of Variables:
Shopping spend (p<0.05): Significant positive impact on income.
Internet speed, devices connected, ads clicked, and platform (p<0.05): Significant negative
impacts.
2. Practical Implications:
While shopping spend positively impacts income, distractions like excessive devices, ads, and mobile
usage negatively influence productivity or income.
Conclusion
The model demonstrates that while some factors like shopping spend positively influence income,
others, such as internet usage habits, device connectivity, and platform choice, can have significant
negative impacts.
The high R-squared value (98.9%) indicates that the model fits the data well. However, the results also
highlight potential behavioral inefficiencies, suggesting the need for better time management, controlled
device usage, and prioritization of productive internet use.
This analysis provides a foundation for further exploration, possibly incorporating additional variables
like education, profession, or work hours to improve understanding.