Topic 6
Application of Data Science
in Global Economic Analysis
1
Outline
A. Panel Data – Economic Relevance
B. Fixed Effects Regression – Concept,
Application, Example
C. Setting Hypothesis
D. Data Collection
E. Data Analysis using Excel
F. Results Interpretation & Hypothesis
Testing
2
A. Panel Data – Economic Relevance
Panel Data (or Longitudinal Data): Data that contains
observations of multiple cross sections across time, a
combination of time-series and cross sectional data.
Most common form of economic data for multiple countries
(cross sections).
Example:
2 years x 3 countries panel = 6 country-year observations 3
B. Fixed Effects Regression
Fixed effects regression: Estimation technique that
controls for UNOBSERVED individual (cross
sectional) characteristics that do not vary with time
(fixed over time), but might impact independent
and/or dependent variables in the regression analysis.
It is usually employed in analyses involving panel
data.
Example: A multiple country panel data analysis
where each country might have individually different
but unobserved characteristics that do not vary with
time, which however might have an impact on the 4
observed variables in the analysis.
B. Fixed Effects Regression
GDP
Growth Country A
Country B
α +a
α
Value of Trade/GDP
• The fixed-effect “a” captures the unobservable differences
between Countries A and B not captured by Value of Trade, but
have a potential impact on either or both variables.
• An example of “a” is the political power of labour unions in a 5
country.
B. Fixed Effects Regression
Fixed effects regression example:
gdp_growth(i,t) = α + a(i) + β1 × unemp_rate(i,t) +
β2 × tariff_rate(i,t) + β3 × inflation_cpi(i,t) +
β4 × trade_gdp(i,t) + ε(i,t)
Dependent variable:
• gdp_growth(i,t) = GDP growth rate of country i in year t.
Independent (explanatory) variables:
• unemp_rate(i,t) = unemployment rate.
• tariff_rate(i,t) = mean weighted tariff rate..
• inflation_cpi(i,t) = inflation rate measured by CPI growth.
• trade_gdp(i,t) = trade value as a % of GDP. 6
B. Fixed Effects Regression
s
7
B. Fixed Effects Regression
Fixed effects regression example:
8
D. Data Collection: Illustration
Hypothesis:
H1: (adj)unemp_rate negatively affects (adj)gdp_growth (β1 < 0)
Explanation?
H2: (adj)tariff_rate negatively affects (adj)gdp_growth (β2 < 0)
Explanation?
H3: (adj)inflation_cpi negatively affects (adj)gdp_growth (β3 < 0)
Explanation?
H4: (adj)trade_gdp positively affects (adj)gdp_growth (β4 > 0)
Explanation?
Null Hypothesis: β = 0
9
D. Data Collection: Illustration
World Bank data website:
[Link]
world-development-indicators
Database: World Development Indicators (WDI)
Sample dataset: data_wdi
Country: 10 countries (for the sample dataset)
Series: GDP growth, Unemployment, Tariff, Inflation,
Trade (%GDP) (for the sample dataset)
Sample period: 2010-2019 (for the sample dataset)
10 countries x 10 years panel dataset (each variable is
mean adjusted using Excel) 10
D. Data Collection: Illustration
Customize the output layout:
11
E. Data Analysis using Excel
Step 1: Install Add-in “Analysis ToolPal” in Excel (go to “Options”)
12
E. Data Analysis using Excel
Step 2: Select “Data Analysis” tool under “Data” tab, and then
select “Regression”
13
E. Data Analysis using Excel
Step 3: Input the range for the dependant variable (Y) and the
independent variables (X). Check the box “Labels” if the first
row of your data spreadsheet contains the variables names.
14
F. Results Interpretation and
Hypothesis Testing
The overall model accuracy is measured by R-square
(R2):
0 ≤ R2 ≤ 1
It is the proportion of the variance that can be
explained by the model.
The larger is R2, the more accurate is your model
in fitting the data.
15
F. Results Interpretation and
Hypothesis Testing
t-statistic is the estimated coefficient divided by its standard
error. The smaller is the standard error (i.e., the larger is the
absolute value of the t-statistic), the more statistically
significant is the estimated coefficient.
p-value of the t-statistic: the probability that the Null
Hypothesis (β = 0) is wrongly rejected (i.e., Type-I error)
The smaller is the p-value, the higher is the confidence you
have in rejecting the Null Hypothesis (β = 0).
16
F. Results Interpretation and
Hypothesis Testing
Rules for Hypothesis Testing
17
F. Results Interpretation and
Hypothesis Testing
18