The
research design to be adopted in this study is the quantitative research
design, this is because it places emphasis on the statistical data and this
data are used to test hypothesis. However, it also shows the methodology to be
applied in other to ascertain the impact of exchange rate on economic growth in
Nigeria.
Model
Specification
One of the important steps the
researcher has to take in studying the relationship between the variables is to
express this relationship in mathematical form
Therefore
GDP = (EX, MS) in the implicit form
GDP = b0+b1Ex+b2Ms+ut
EXC = Exchange
rate
Ms = Money
supply
B0 = constant
B1 = regression
coefficient
Ut = error
term
Data
Required and Sources
The data required is the secondary data. They are time
series data on GDP at current factor.
They were collected from the under
listed sources
(1)
CBN statistical
bulletin for planned year
(2)
CBN economic and
financial review
(3)
National planning
office
(4)
Federal bureaux
of statistics publication
(5)
CBN annual report
and statement of account
Addition of some relevant tests and journals were also
consulted for this work.
Method
of Evaluation/ Data Analysis
The method of evaluation in this study is the multiple
regression method applying the ordinary least square (OLS) method. This method
will be used to evaluate the working hypothesis, the test statistics and the
econometrics tests to be conducted.
Method of Testing the Hypothesis
(i)
R2 coefficient of determination is used to measure the percentage of
variation in the dependent variable attributable to the independent variable,
this is known as the goodness of fit.
Decision Rule
An R2 of one (1) means a
perfect fit, r2 of zero means that there is no relationship between
the regress and the regrssor. Also if 0<r2<1, it means there
is relationship between the regress and the regressor. And the closer r2
is to one (1) the better the fit.
(ii) Student t-test shall be used to test for
the statistical significant of the individual regression coefficient.
Decision Rule
If t*<t0.025, accept the null hypothesis H0 and if
otherwise reject it:
(iii) F-statistics measures the overall
significant of the entire regression plane. It will be used to test the joint
influence of the explanatory variable on the dependent variable.
Decision Rule
If F*<f0.05, accept the null hypothesis and if
otherwise reject it.
(iv) The standard error test shall be used to
test or measure the reliability of the parameter estimate.
Decision Rule
If the standard errors is smaller than half the
numerical value of the parameter estimate, i.e S (b0) < b0/2
or S (b1) <b1/2, it means that the estimate is statistically.
Also if S (b0) >b0/2 or S (b1)>b1/2,
it means that the estimate is not statistically significant.
(v) Durbin-Watson d-test shall be used to
measure the presence of autocorrelation with a first order scheme.
Decision Rule
If d*<dL, reject the null
hypothesis H0 and concluded that the forecasting power of the model performance
is poor, and if d*>dL, accept the null.