MULTIPLE REGRESSION ANALYSIS OF THE EFFECT OF EXCHANGE

RATES AND UNEMPLOYMENTS ON THE NIGERIA ECONOMY GROWTH

BY

BADA DAMILOLA F/HD/23/3740033

ADEYEMO DEBORAH O. F/HD/23/3740034
EWEDEYI SADIAT YETUNDE F/HD/23/3740035
TAIWO VICTORIA MEDESE F/HD/23/3740036

YABA COLLEGE OF TECHNOLOGY

SCHOOL OF SCIENCE
DEPARTMENT OF STATISTICS

COURSE TITLE:
RESEARCH METHODOLOGY

COURSE CODE:
STA 326

LECTURER: MRS AJIBOYE, Y. O.

 Abstract

Regression analysis is a widely used statistical technique to build a model from a set of data on
two or more variables. Linear regression is based on linear correlation, and assumes that change
in one variable is accompanied by a proportional change in another variable. Simple linear
regression, or bivariate regression, is used for predicting the value of one variable from another
variable (predictor); however, multiple linear regression, which enables us to analyse more than
one predictor or variable, is more commonly used. This paper explains both simple and multiple
linear regressions illustrated with an example of analysis and also discusses some common errors
in presenting the results of regression, including inappropriate titles, causal language,
inappropriate conclusions, and misinterpretation
 TABLE OF CONTENTS
Title Page
Abstract
Introduction
1.1 Background Of The Study
1.2 Statements Of The Problem
1.3 Aim And Objectives
1.4. Research Questions
1.5 Research Hypothesis

2.0 Literature Review

3.0 Methodology
3.1 Introduction
3.2 Research Design
3.3 Data Sources
3.4 Data Collection Method
3.5 Data Analysis Techniques
3.6 Model Specification
3.7 Estimation Technique
3.8 Diagnostic Tests

4.0 Data Presentation And Analysis

5.0 Conclusion

References
 INTRODUCTION
1.1 Background of the study

Regression analysis is one of the most frequently used tools in market research. In its simplest
form, regression analysis allows market researchers to analyze relationships between one
independent and one dependent variable. In marketing applications, the dependent variable is
usually the outcome we care about (e.g., sales), while the independent variables are the instruments
we have to achieve those outcomes with (e.g., pricing or advertising). Regression analysis can
provide insights that few other techniques can. The key benefits of using regression analysis are
that it can:

1. Indicate if independent variables have a significant relationship with a dependent variable.

2. Indicate the relative strength of different independent variables’ effects on a dependent
variable.
3. Make predictions.

Knowing about the effects of independent variables on dependent variables can help market
researchers in many different ways. For example, it can help direct spending if we know
promotional activities significantly increase sales. Knowing about the relative strength of effects
is useful for marketers because it may help answer questions such as whether sales depend more
on price or on promotions. Regression analysis also allows us to compare the effects of variables
measured on different scales such as the effect of price changes (e.g., measured in $) and the
number of promotional activities.

Regression analysis can also help to make predictions. For example, if we have estimated a
regression model using data on sales, prices, and promotional activities, the results from this
regression analysis could provide a precise answer to what would happen to sales if prices were to
increase by 5% and promotional activities were to increase by 10%. Such precise answers can help
(marketing) managers make sound decisions. Furthermore, by providing various scenarios, such
as calculating the sales effects of price increases of 5%, 10%, and 15%, managers can evaluate
marketing plans and create marketing strategies.
Understanding Regression Analysis

Regression models are generally noted as follows:

y = α + β1 x1 + e

What does this mean? The y represents the dependent variable, which is the variable you are trying
to explain. The α represents the constant (sometimes called intercept) of the regression model and
indicates what your dependent variable would be if all of the independent variables were zero. The
independent variable is indicated by x1. β1 (pronounced as beta) indicates the (regression)
coefficient of the independent variable x. This coefficient represents the gradient of the line and is
also referred to as the slope. A positive β1 coefficient indicates an upward sloping regression line,
while a negative β1 indicates a downward sloping line.

In regression analysis, we can calculate whether this value (the β1 parameter) differs significantly
from zero by using a t-test.

The Error (Residual) in Regression Analysis

The last element of the regression notation, the e, denotes the error (or residual) of the equation.
The term "error" is commonly used in research, while SPSS uses the term "residuals." If we use
the word "error," we discuss errors in a general sense. If we use "residuals," we refer to specific
output created by SPSS. The error is the distance between each observation and the best fitting
line.
To clarify what a regression error is, consider that the error is the difference between the regression
line (which represents our regression prediction) and the actual observation. The predictions made
by the “best” regression line are indicated by ^y (pronounced y-hat). Thus, the error for the first
observation is:
e₁ = y₁ − ^y₁
In the example above, we have only one independent variable. This is called bivariate regression.
If we include multiple independent variables, we call this multiple regression. The notation for
multiple regression is similar to that of bivariate regression. If we were to have three independent
variables, such as the index of promotional activities (x₁), price of competitor 1 (x₂), and the price
of competitor 2 (x₃), our notation would be:
y = α + β₁ x₁ + β₂ x₂ + β₃ x₃ + e
We need one regression coefficient for each independent variable (i.e., β₁, β₂, and β₃). Technically,
the βs indicate how a change in an independent variable influences the dependent variable,
assuming all other independent variables are held constant.
Executing a Regression Analysis
Now that we have introduced some basics of regression analysis, it is time to discuss how to
execute a regression analysis. We outline the key steps in the process.
First, we need to consider the data requirements for regression analysis, which determine if
regression analysis can be used. After addressing these data requirements, we move on to
specifying and estimating the regression model.

1.2 STATEMENTS OF THE PROBLEM

Nigeria faces significant challenges due to persistent exchange rate volatility, which has profound
impacts on its economic growth and development. The depreciation of the naira has increased
import costs, leading to higher prices and reduced purchasing power for Nigerians. This instability
has discouraged foreign investment, worsened poverty and inequality, and limited access to
education and healthcare, especially for the poor. The volatility also affects economic
opportunities for youth, raising the risk of social unrest. The Central Bank of Nigeria struggles to
manage this issue, as the country relies heavily on oil exports, which are influenced by global oil
price fluctuations. A lack of coherent exchange rate policies and failure to address structural issues
have hindered Nigeria's overall economic progress.

1.3 AIM AND OBJECTIVES

The main aim is to examine the effect of exchange rates and unemployment on Nigeria economy.
The following objectives are:

1. To investigate the effect of exchange rates and unemployment on GDP of Nigeria.

2. To examine the relationship between exchange rates, unemployment and Nigeria's GDP.
1.4. RESEARCH QUESTIONS

1. What is the relationship between exchange rates, unemployment and GDP in Nigeria?

2. What is the effect of exchange rates and unemployment on the economic growth of Nigeria?

1.5 RESEARCH HYPOTHESIS

Hypothesis 1

H0: There is no significant relationship between exchange rates, unemployment rate and GDP in
Nigeria.

H1: There is a significant relationship between exchange rates and economic growth in Nigeria.

Hypothesis 2

H0: Exchange rates and unemployment rate has no significant effect on the economic growth of
Nigeria.

H1: Exchange rates and unemployment rate has a significant effect on the economic growth of
Nigeria..
2.0 LITERATURE REVIEW

This study examines the effects of exchange rate on unemployment and Nigeria's economic
growth.

2.1 CONCEPTUAL REVIEW

A conceptual review analyzes the underlying concepts, theories, and frameworks related to a
specific research topic or phenomenon. It explores the abstract ideas, definitions, and relationships
that shape our understanding of the topic.

2.2 CONCEPTUAL FRAMEWORK

Exchange rate is defined as the price of one country's currency expressed in terms of another
country's currency. It can be expressed as either nominal or real exchange rate. The real effective
exchange rate (REER) is also significant, accounting for the nominal effective exchange rate
divided by a price deflator index. There are ongoing debates regarding the relationship between
exchange rate and unemployment in economic literature. Some studies have explored whether
such a relationship exists, while others aim to understand its nature. Notable works include those
by Nyahokwe and Ncwadi, Shaari, Hussain, and Abdul Rahim (2013), Chang, Mohammadi and
Gholami (2008), Frenkel and Ros, Milas and Legrenzi, and Djivre and Ribon. Most studies, except
for Mohammadi and Gholami (2008), suggest a relationship between exchange rate and
unemployment. For example, Nyahokwe and Ncwadi (2013) focused on the impact of real
exchange rate volatility on unemployment in South Africa. They found that real exchange rate
accounted for the largest variation in unemployment rates, concluding that fluctuations in
unemployment were primarily due to real exchange rate shocks, rather than other factors like
interest rates or economic growth.

2.4 EMPIRICAL REVIEW

Aidi, Saidu, and Suleiman (2018) explored the relationship between exchange rate volatility and
the performance of Nigeria's manufacturing sector using quarterly data from 1980Q1 to 2016Q4.
They applied the OLS multiple regression technique and used Exponential Generalized
Autoregressive Conditional Heteroskedasticity (EGARCH) to measure exchange rate volatility.
Their findings indicated an inverse relationship between exchange rate volatility and
manufacturing sector performance in Nigeria.

Nsofor, Takon, and Ugwuegbe (2017) studied the effects of exchange rate volatility on economic
growth in Nigeria using data from 1981 to 2015. They employed the GARCH (1,1) model to
estimate exchange rate volatility and the Generalized Method of Moments (GMM) to analyze the
impact on economic growth. The results showed that exchange rate volatility and foreign direct
investment (FDI) had a negative and significant impact on Nigeria's economic growth, while
government expenditure and external reserves had a positive and significant impact.

Iyeli and Clement (2017) examined the effects of exchange rate volatility on economic growth in
Nigeria from 1970 to 2011, using data sourced from the Central Bank of Nigeria Statistical
Bulletin. They employed Johansen Co-integration techniques to assess short- and long-term
effects. Their findings revealed a positive relationship between exchange rate volatility and GDP
in the long run, with oil revenue and exchange rates contributing positively to GDP.

Ugochukwu (2015) evaluated the impact of exchange rate fluctuation on Nigeria's economic
growth using annual data from 1980 to 2012. The study applied the Augmented Dickey-Fuller
(ADF) test and co-integration analysis. The results suggested that exchange rate volatility had a
negative short-term effect on economic growth in Nigeria, with a similar negative long-term
relationship between the two variables.

Uduakobong and Enobong (2015) analyzed the relationship between exchange rate movements
and economic growth in Nigeria using annual data from 1970 to 2011. Their study found a positive
but insignificant relationship between exchange rate and economic growth. Furthermore, no
causality between the two variables was observed. They recommended that the government adopt
appropriate fiscal and monetary policies to stabilize exchange rates and foster economic growth.

Emerah, Adeleke, and David (2015) investigated the relationship between GDP, exchange rates,
imports, exports, and inflation in Nigeria, using data from 1986 to 2013. Their analysis revealed a
significant positive relationship between GDP, exchange rates, and exports. The F-statistic
indicated that all independent variables were jointly significant in explaining Nigeria's economic
growth, concluding that exchange rate volatility affects the growth of the Nigerian economy during
the period under review.
 METHODOLOGY

3.1 INTRODUCTION

This chapter outlines the methodology used to investigate the relationship between exchange rates,
unemployment, and economic growth in Nigeria. The chapter discusses the research design, data
sources, data collection methods, and data analysis techniques used in the study.

3.2 RESEARCH DESIGN

This study employs a quantitative research design, using secondary data to examine the
relationship between exchange rates, unemployment, and economic growth in Nigeria. The study
uses a time-series analysis approach, examining data 2000 to 2024.

3.3 DATA SOURCES

The data used in this study are obtained from secondary sources, including: The sources of the
data used is from CBN.

3.4 DATA COLLECTION METHOD

The data are collected from the above-mentioned sources using online databases and publications.

3.5 DATA ANALYSIS TECHNIQUES

The data are analyzed using descriptive statistics, correlation analysis, and regression analysis.
Specifically:

1. Descriptive statistics are used to summarize the data and provide an overview of the variables.

2. Correlation analysis is used to examine the relationships between the variables.

3. Regression analysis is used to examine the impact of exchange rates and unemployment on
economic growth of Nigeria.

3.6 MODEL SPECIFICATION

The study uses a multiple linear regression model to examine the relationship between exchange
rates, unemployment, and economic growth in Nigeria. The model is specified as follows:
GDP = β0 + β1EXCH + β2UNEMP + ε

Where:

- UNEMP is the unemployment rate

- EXCH is the exchange rate

- GDP is the GDP growth rate

- β0 is the intercept or constant term

- β1, β2, are the slope coefficients representing the change in the dependent variable for a one-unit
change in each independent variable

- ε is the error term representing the random variation in the dependent variable not explained by
the independent variables

3.7 ESTIMATION TECHNIQUE

The model is estimated using the Ordinary Least Squares (OLS) technique.

3.8 DIAGNOSTIC TESTS

The study conducts diagnostic tests to check for the assumptions of the regression analysis,
including:

1. Multicollinearity test

2. Heteroscedasticity test

3. Normality test
 DATA PRESENTATION AND ANALYSIS

Data Presentation

Year GDP growth (annual %) Unemployment Official

exchange rate
2000 5.015935 3.87 101.6973
2001 5.917685 3.827 111.2313
2002 15.32916 3.618 120.5782
2003 7.347195 3.594 129.2224
2004 9.250558 3.556 132.888
2005 6.438517 3.69 131.2743
2006 6.059428 3.718 128.6517
2007 6.59113 3.767 125.8081
2008 6.764473 3.771 118.5667
2009 8.036925 3.751 148.88
2010 8.005656 3.739 150.2975
2011 5.307924 3.77 153.8625
2012 4.230061 3.76 157.5
2013 6.671335 3.711 157.3117
2014 6.309719 3.903 158.5526
2015 2.652693 4.137 192.4403
2016 -1.61687 4.498 253.492
2017 0.805887 4.826 305.7901
2018 1.922757 5.065 306.0837
2019 2.208429 5.206 306.921
2020 -1.79425 5.712 358.8108
2021 3.647187 5.393 401.152
2022 3.251681 3.827 425.9792
2023 2.860215 3.074 227.8937
 Correlations
Gross Unemployme Exchange rate
Domestic nt
Product
Pearson Correlation Gross Domestic Product 1.000 -.634 -.571
Unemployment -.634 1.000 .340
Exchange rate -.571 .340 1.000
Sig. (1-tailed) Gross Domestic Product . .000 .002
Unemployment .000 . .052
Exchange rate .002 .052 .
N Gross Domestic Product 24 24 24
Unemployment 24 24 24
Exchange rate 24 24 24

Interpretation:

The correlation analysis reveals the relationships between Gross Domestic Product (GDP),
Unemployment, and the Exchange Rate. There is a moderate negative correlation between GDP
and Unemployment (r = -0.634), which is statistically significant (p = 0.000). This suggests that
as GDP increases, unemployment tends to decrease, and vice versa. Similarly, a moderate negative
correlation is observed between GDP and the Exchange Rate (r = -0.571), with statistical
significance (p = 0.002). This indicates that as GDP grows, the exchange rate tends to fall, implying
that a stronger economy may lead to a weaker currency, or conversely, a weaker economy to a
stronger currency.

However, the relationship between Unemployment and the Exchange Rate shows a weak positive
correlation (r = 0.340), which is not statistically significant (p = 0.052). This means that while
there is a slight tendency for unemployment and exchange rates to move in the same direction, this
relationship is not strong enough to be considered meaningful based on this data.
 Model Summaryb
R R Adjusted Std. Error of Change Statistics
Square R Square the Estimate R Square F Change df1 df2 Sig. F
Change Change
.738a .545 .502 2.553 .545 12.584 2 21 .000
a. Predictors: (Constant), Exchange rate, Unemployment
b. Dependent Variable: Gross Domestic Product

R = 0.738: This is the multiple correlation coefficient, indicating a strong positive relationship
between the predictors (Exchange Rate and Unemployment) and the dependent variable (GDP).
An R value of 0.738 suggests that the model explains a significant portion of the variability in
GDP.

R Square = 0.545: This is the coefficient of determination, which tells us that 54.5% of the variance
in GDP can be explained by the combined effect of Exchange Rate and Unemployment. This is a
good proportion, indicating that the model has a moderate to strong explanatory power.

Adjusted R Square = 0.502: This adjusted value accounts for the number of predictors in the model
and gives a more accurate measure of how well the model generalizes to the population. The slight
decrease from R Square (from 0.545 to 0.502) suggests that the model's explanatory power remains
fairly robust, even after accounting for the number of predictors.

Std. Error of the Estimate = 2.553: This is the standard error of the regression model's predictions.
It indicates the average distance between the actual and predicted values of GDP. A lower value
would indicate better predictive accuracy.

R Square Change = 0.545: This indicates that the inclusion of the predictors (Exchange Rate and
Unemployment) results in a 54.5% increase in the explained variance in GDP compared to a
baseline model with no predictors. This is a significant change, demonstrating that the predictors
have a strong impact on GDP.
F Change = 12.584: This F-statistic tests whether the model as a whole is statistically significant.
The value of 12.584 indicates that the predictors (Exchange Rate and Unemployment) collectively
have a significant effect on GDP.

df1 = 2, df2 = 21: These are the degrees of freedom for the model. df1 (2) corresponds to the
number of predictors in the model, and df2 (21) is the residual degrees of freedom, calculated as
the total number of observations minus the number of predictors minus 1.

Sig. F Change = 0.000: The p-value for the F-test is less than 0.05, indicating that the model is
statistically significant. This means that the relationship between the predictors (Exchange Rate
and Unemployment) and GDP is unlikely to have occurred by chance, and the model is a good fit
for the data.

Interpretation:

The regression model explains 54.5% of the variance in GDP, and the predictors (Exchange Rate
and Unemployment) significantly improve the model's ability to predict GDP, as indicated by the
strong R-squared value and the statistically significant F-test (p = 0.000). The model's fit is good,
and the predictors are important contributors to understanding GDP in this context.

ANOVAa
Model Sum of Squares df Mean Square F Sig.

1 Regression 164.054 2 82.027 12.584 .000b

Residual 136.889 21 6.519
Total 300.944 23
a. Dependent Variable: Gross Domestic Product
b. Predictors: (Constant), Exchange rate, Unemployment

Sum of Squares (SS): This measures the variation in the dependent variable (Gross Domestic
Product, GDP) that is explained by the model and the error (residuals).
Regression SS (164.054): This represents the variation in GDP that is explained by the model (i.e.,
the relationship between GDP, Exchange Rate, and Unemployment).
Residual SS (136.889): This represents the unexplained variation in GDP after accounting for the
predictors in the model.
Total SS (300.944): This is the total variation in GDP, which is the sum of the regression and
residual sums of squares.
Degrees of Freedom (df): The degrees of freedom help in understanding the amount of information
available for estimation.
Regression df (2): This corresponds to the number of predictors in the model (Exchange Rate and
Unemployment).
Residual df (21): This corresponds to the number of observations minus the number of predictors
minus 1 (N - p - 1 = 24 - 2 - 1).
Total df (23): This is the total number of observations minus 1 (N - 1 = 24 - 1).
Mean Square (MS): The mean square is the sum of squares divided by the corresponding degrees
of freedom.
Regression MS (82.027): This is calculated by dividing the Regression Sum of Squares by its
degrees of freedom (164.054 ÷ 2).
Residual MS (6.519): This is calculated by dividing the Residual Sum of Squares by its degrees of
freedom (136.889 ÷ 21).
F-statistic (12.584): The F-statistic tests whether the regression model provides a significantly
better fit than the null model (a model with no predictors). It is calculated as the ratio of the Mean
Square for Regression to the Mean Square for Residual (82.027 ÷ 6.519). The higher the F-statistic,
the stronger the evidence that the model explains a significant portion of the variation in GDP.
Sig. (p-value) = 0.000: The p-value is less than 0.05, indicating that the regression model as a
whole is statistically significant. In other words, the relationship between Exchange Rate,
Unemployment, and Gross Domestic Product is unlikely to have occurred by random chance.
Statistical Interpretation:
The ANOVA results indicate that the overall regression model is statistically significant (F =
12.584, p = 0.000), suggesting that the combination of Exchange Rate and Unemployment explains
a significant portion of the variation in Gross Domestic Product (GDP). The Regression Sum of
Squares (164.054) indicates that the model explains a substantial amount of the variation in GDP,
and the Residual Sum of Squares (136.889) shows the unexplained variation. The p-value of 0.000
(which is less than 0.05) confirms that the predictors, Exchange Rate and Unemployment, have a
statistically significant impact on GDP, making this regression model a good fit for predicting
GDP. Thus, we can conclude that the model significantly improves the prediction of GDP, and the
relationship between Exchange Rate, Unemployment, and GDP is not due to random chance.

Coefficientsa
Model Unstandardized Standardized t Sig.
Coefficients Coefficients
B Std. Error Beta
(Constant) 18.381 3.286 5.593 .000
Unemployment -2.687 .844 -.498 -3.182 .004
Exchange rate -.011 .004 -.402 -2.567 .018
a. Dependent Variable: Gross Domestic Product

95.0% Confidence Interval for B Collinearity Statistics

Lower Bound Upper Bound Tolerance VIF
11.547 25.215
-4.443 -.931 .885 1.130
-.020 -.002 .885 1.130

The Coefficients table provides insight into the individual contributions of Unemployment and the
Exchange Rate in predicting Gross Domestic Product (GDP). The unstandardized coefficient for
the constant (18.381) indicates that when both Unemployment and the Exchange Rate are zero,
the expected value of GDP is 18.381. For Unemployment, the unstandardized coefficient is -2.687,
which means that, holding the Exchange Rate constant, for every 1-unit increase in
Unemployment, GDP is expected to decrease by 2.687 units. This suggests a negative relationship
between Unemployment and GDP. Similarly, the coefficient for the Exchange Rate is -0.011,
indicating that for every 1-unit increase in the Exchange Rate, GDP is expected to decrease by
0.011 units, also suggesting a negative relationship between the two.

The standardized coefficients (Beta) show the relative strength of each predictor. The Beta value
for Unemployment is -0.498, while the Beta value for the Exchange Rate is -0.402. These values
suggest that Unemployment has a slightly stronger effect on GDP than the Exchange Rate, as the
absolute value of the Unemployment Beta is higher. Both predictors have negative effects on GDP,
with Unemployment having a slightly more substantial impact.
The t-values and significance levels further confirm the statistical significance of both predictors.
The t-value for Unemployment is -3.182, with a p-value of 0.004, which is less than the 0.05
significance threshold, indicating that Unemployment has a statistically significant negative effect
on GDP. Similarly, the t-value for the Exchange Rate is -2.567, with a p-value of 0.018, which is
also less than 0.05, confirming that the Exchange Rate significantly affects GDP as well.

The 95% Confidence Intervals for the coefficients provide further evidence of the reliability of
these estimates. For Unemployment, the interval is (-4.443, -0.931), and for the Exchange Rate,
the interval is (-0.020, -0.002). Since both intervals do not include zero, we can be confident that
the relationships are significant and not due to random chance.

Finally, the collinearity statistics, including the Tolerance (0.885) and VIF (1.130), suggest that
there are no issues with multicollinearity between the predictors, as both values fall within
acceptable ranges (Tolerance > 0.1 and VIF < 10).

In summary, both Unemployment and the Exchange Rate have statistically significant negative
effects on GDP, with Unemployment having a slightly stronger impact. The model does not suffer
from multicollinearity, and the coefficients are reliable for making inferences about the
relationships between the variables.

Collinearity Diagnosticsa
Model Dimension Eigenvalue Condition Variance Proportions
Index (Constant) Unemployme Exchange rate
nt
1 1 2.816 1.000 .00 .00 .03
2 .172 4.043 .03 .02 .92
3 .012 15.252 .97 .98 .05
a. Dependent Variable: Gross Domestic Product

Interpretation:

The Condition Index of 15.252 in the third dimension, combined with the high Variance

Proportions for Unemployment (0.98) and Exchange Rate (0.97), suggests that there is some

degree of multicollinearity between these two predictors. While the condition index is not
exceedingly high (which would indicate severe multicollinearity), the high values for Variance

Proportions point to the fact that Unemployment and Exchange Rate are highly correlated with

each other, potentially causing multicollinearity issues.

The first dimension, with a Condition Index of 1.000, indicates that there is no significant

multicollinearity here, and the second dimension also does not show severe multicollinearity,

with a condition index of 4.043. Therefore, the primary concern for multicollinearity lies in the

third dimension, where Unemployment and Exchange Rate show high collinearity.

Residuals Statisticsa
Minimum Maximum Mean Std. N
Deviation
Predicted Value -.91 7.37 5.05 2.671 24
Residual -5.112 7.964 .000 2.440 24
Std. Predicted -2.230 .868 .000 1.000 24
Value
Std. Residual -2.002 3.119 .000 .956 24
a. Dependent Variable: Gross Domestic Product

The residuals have a mean of 0, which suggests that there is no systematic bias in the model’s
predictions. This is a good sign, as it indicates that the model is neither overestimating nor
underestimating GDP on average.

The standard deviation of the residuals (2.440) indicates that the model’s predictions deviate, on
average, by 2.44 units from the actual values. This is a reasonable level of error in the context of
predicting GDP, but it also suggests room for improvement.

The standardized residuals range from -2.002 to 3.119, which is within the typical range for
residuals, although the value of 3.119 is somewhat large. Residuals that exceed |3| are typically
considered potential outliers. In this case, the model might benefit from further diagnostics or
addressing potential outliers.
The standardized predicted values and standardized residuals both have a mean of 0 and a standard
deviation close to 1, which suggests that the model is performing well in terms of scaling and the
spread of the residuals is consistent with expectations.

Interpretation of a Histogram Following a Normal Distribution:

1. Shape of the Histogram:

o A histogram that follows a normal distribution typically has a bell-shaped curve,
where the majority of the data points are clustered around the center, and the
frequency of values gradually decreases as you move away from the center in both
directions.
o If your histogram follows this shape, it suggests that the residuals (differences
between observed and predicted values) are normally distributed, which is a
desirable property in regression analysis.
2. Central Tendency:
o Since the residuals have a mean of 0, the histogram should be centered around 0
on the x-axis. This means that, on average, the model’s predictions are not biased
toward overestimating or underestimating the dependent variable (GDP).
 o The normal distribution suggests that the model’s errors are distributed
symmetrically around 0, with most of the errors being small and a few outliers
(larger errors) on either side.
3. Spread of Residuals:
o The spread of the residuals is indicated by the standard deviation, which was
2.440 in the interpretation above. If the histogram follows a normal distribution, it
will show that most residuals are within ±2 or ±3 standard deviations from the mean
(0).
o In a normal distribution, approximately 68% of the residuals will fall within one
standard deviation, 95% will fall within two standard deviations, and 99.7% will
fall within three standard deviations. So, if the histogram shows a large
concentration of residuals close to 0 and tapers off symmetrically, it aligns with this
expectation.
4. Outliers:
o In a perfectly normal distribution, you would expect very few extreme residuals far
away from 0 (i.e., outliers). Outliers would appear as isolated bars on the far ends
of the histogram.
o If your histogram shows a few extreme values far from the mean (positive or
negative), this could suggest the presence of outliers. In the context of regression,
extreme residuals might indicate that some data points are not well predicted by the
model and could require further examination.
5. Skewness and Kurtosis:
o If the histogram is perfectly normal, it should have no skew (i.e., the left and right
tails are symmetrical) and no excess kurtosis (i.e., the peaks and tails are of normal
height).
o If the histogram is skewed (either left or right), it could indicate that the model is
systematically overestimating or underestimating the dependent variable. A
positively skewed distribution would suggest that the model is underpredicting
some values, while a negatively skewed distribution would indicate
overprediction.
 o If the histogram has heavy tails (i.e., it’s more peaked and has more extreme values
than a normal distribution), it could suggest that the model is underperforming for
certain data points and that there may be heteroscedasticity (non-constant variance
of residuals).

Conclusion:

A histogram that follows a normal distribution of residuals suggests that the regression model is
well-specified, with errors being symmetrically distributed around zero and no significant patterns
in the residuals. This is a good sign that the model is making reliable predictions and meets the
assumptions of linear regression, particularly the assumption of normally distributed errors. If the
histogram shows some deviation from normality (such as skewness or heavy tails), this may
require further investigation or adjustments to the model.
The P-P plot for standardized residuals in a regression analysis is a graphical tool used to assess
the normality of the residuals and identify potential outliers. On the plot, the x-axis represents the
observed probability values, while the y-axis shows the expected probability values. The diagonal
line on the plot represents the ideal scenario where residuals are perfectly normal. If the residuals
are normally distributed, the points on the plot will align closely with the diagonal line. In the case
described, the points generally follow the diagonal line, indicating that the residuals are
approximately normal, which supports the validity of the regression model. However, the plot also
reveals some outliers, which are points that deviate significantly from the diagonal. These outliers
suggest the presence of potential data errors or observations that do not fit the regression model
well. It is important to investigate these outliers further to determine whether they are due to data
inaccuracies or represent actual anomalies in the data

Normality of Residuals: The foundation of valid regression analysis lies in the normal distribution
of residuals – the differences between the observed and predicted values of the dependent variable.
A normal Predicted Probability (P-P) plot serves as a diagnostic tool, where a normal distribution
of residuals aligns closely with the plot’s diagonal line.

Homoscedasticity: This assumption concerns the distribution of residuals. For your data to be
homoscedastic, the residuals should be evenly spread across the range of predicted values,
resembling a “shotgun blast” of points. This uniform distribution ensures that the variance of errors
is constant. Heteroscedasticity, the opposite condition, manifests as a patterned spread of residuals
(e.g., cone or fan-shaped), indicating variance inconsistency.

Linearity: The assumption of linearity posits a direct, straight-line relationship between predictor
and outcome variables. If the residuals are normally distributed and exhibit homoscedasticity,
linearity typically holds, simplifying the analysis process. Multicollinearity refers to when your
predictor variables are highly correlated with each other.

Absence of Multicollinearity: In multiple linear regression, where several predictor variables are
involved, multicollinearity can obscure the distinct impact of each predictor. This phenomenon
occurs when predictors are highly correlated with one another, complicating the attribution of
variance in the outcome variable. Multicollinearity assessment can be approached through two
methods:
Correlation Coefficients: By constructing a correlation matrix among predictors, coefficients
nearing or exceeding .80 indicate strong correlations, suggesting multicollinearity.

Variance Inflation Factor (VIF): VIF values offer a quantitative measure of multicollinearity,
with values below 5.00 indicating minimal concern, and those exceeding 10.00 signaling
significant multicollinearity.

Fig 1: A scatter plot showing the relationship between GDP and Unemployment

The graph illustrates a negative linear correlation between GDP and unemployment, indicating
that as GDP rises, unemployment generally falls. The trend line suggests that unemployment will
continue to decrease as GDP grows. The R-squared value of 0.402 indicates that 40.2% of the
variation in GDP is explained by changes in unemployment.
Fig 2: A scatter plot showing the relationship between GDP and Exchange rate
The graph illustrates a negative linear correlation between GDP and exchange rate, indicating that
as GDP rises, exchange rate generally falls. The trend line suggests that exchange rate will continue
to decrease as GDP grows. The R-squared value of 0.326 indicates that 32.6% of the variation in
GDP is explained by changes in exchange rate.

Conclusion

The purpose of multiple regression is to explore a set of predictors (independent variables) and
either find their best combination to predict the value of a criterion (dependent variable) or to
evaluate the independent contributions of each predictor. Although referred to as dependent and
independent, these variables are in fact in relation, and explanations of causality should be avoided.
If and only if the analysis is significant (P<0.05), squared multiple correlation coefficient and other
coefficients can be explained.
 REFERENCES

Aiken, L. S., & West, S. G. (1991). Multiple regression: Testing and interpreting interactions.

Thousand Oaks, CA: Sage.

Cohen, J. (1994). The Earth is round (P<.05). The American Psychologist, 49(912), 997–1003.

Field, A. (2013). Discovering statistics using SPSS (4th ed.). London: Sage.

Fornell, C., Johnson, M. D., Anderson, E. W., Cha, J., & Johnson, B. E. (1996). The American

customer satisfaction index: Nature, purpose, and findings. Journal of Marketing, 60(4), 7–18.

Green, S. B. (1991). How many subjects does it take to do a regression analysis? Multivariate

Behavioral Research, 26(3), 499–510.

Hill, C., Griffiths, W., & Lim, G. C. (2008). Principles of econometrics (3rd ed.). Hoboken, NJ:
Wiley.

Kelley, K., & Maxwell, S. E. (2003). Sample size for multiple regression: Obtaining regression

coefficients that are accurate, not simply significant. Psychological Methods, 8(3), 305–321.