Having estimated model, variables considered are Real Gross Domestic Product (dependent variable), Government Expenditure on Health and Government Expenditure on Education (Independent variables) it covers the period of years: 1980-2008
GDP = + 7.1724
+ 0.6196GEE + 0.1539GEH
T* =
(9.8554) (1.4409) (0.3715)
S.E
= (0.7277)
(0.4300) (0.4142)
t0.025 = 2.056
F (2, 26) = 76.203
F0.05
= 3.37
R2 = 0.854
DW = 2.149
ANALYSIS OF RESULTS
(a) T-test: It is used to
test for the statistical significance of the individual estimated parameters.
The calculated t-value for the regression coefficients of GEE and GEH are 1.4409
and 0.3715 respectively. Since the calculated t-values of GEE and GEH are less
than the tabulated t-value of 2.056 at 5% level of significance; we conclude
that the regression coefficients are statistically insignificant.
(b) Standard Error test: It is used to test
for statistical reliability of the coefficient estimates.
S(b1) = 0.4300
S(b2) = 0.4142
b1/2 = 0.3099
b2/2 = 0.0769
Since S(b1) >
b1/2, and S(b2/2) > b2/2
, we conclude that the coefficient estimate of b1 and b2
are is statistically insignificant.
(c) F-Test: This is used to test for the
joint influence of the explanatory variables on the dependent variable. The F-calculated value is 76.203 while the F-tabulated value is 3.37 at 5% level of significance. Since the F-calculated value is greater than the F-tabulated value, we
conclude that the entire regression plane is statistically significant. This
means that the joint influence of the explanatory variables (GEE and GEH) on
the dependent variable (GDP) is statistically significant. This result can as
well be confirmed from the F- probability which is statistically significant.
(d) Coefficient of Determination (R2): It is used to measure
the proportion of variations in the dependent variable, which is explained by
the explanatory variables. The computed coefficient of determination (R2= 0.8542) shows that 85.42% of
the total variations in the dependent variable (LGDP) is influenced by the
variation in the explanatory variables namely Government Expenditure on Health
(GEH) and Government Expenditure on Education (GEE) while 14.58% of the total
variation in the dependent variable is attributable to the influence of other
factors not included in the regression model.
(e) Durbin Watson statistics: It is used to
test for the presence of positive first order serial correlation. The computed
DW is 2.1497. At 5% level of
significance with two explanatory variables and 29 observations, the tabulated
DW for dL and du are 1.270 and 1.563 respectively. The value of DW is greater than
the lower limit. Therefore, we conclude that there is no evidence of positive
first order serial correlation. i.e. no autocorrelation in the model.
TEST OF HYPOTHESIS
F-test is employed in testing
the hypothesis. Using 5% level of significance at 2/26 degrees of freedom, the
F-tabulated value is 3.37 while
calculated F-value is 76.203. Since
the calculated F-value is greater than the tabulated F-value, we reject H0.
Thus, Hi is accepted on the proposition that Human Capital Development had
significant impact on Economic growth in Nigeria within the period under study
i.e. 1980-2008.
IMPLICATION OF THE RESULT
The regression result shows that
there existed a positive relationship between dependent variable (RGDP) and the
explanatory variables (GEH and GEE). It is estimated from the result that a
unit increase in government expenditure on health and government expenditure on
education, on the average, will lead to increase by 0.15 and 0.62 units in GDP
respectively. However, holding the explanatory variables constant, GDP increase
by 7.17 units. It is obviously seen that the sign borne by parameters estimate
meet the prior expectations.
The
entire repression plan is statistical significant. Invariably, the joint
information of the explanatory variables has a significant influence on the
dependent variable. Also, the coefficient of multiple determinations shows a
valid goodness of fit. However, there is presence of autocorrelation. This could
be attributed to the absence or omission of some variable which are captured in
the stochastic variables but not included in the regression model.