Nigeria Economic Growth

EFECTS OF FOREIGN DIRECT INVESTMENT ON ECONOMIC GROWTH IN NIGERIA (1970-2011)

CHAPTER 3

DATA ANALYSIS

The Descriptive statistics analysis

There is a 42 year analysis on the GDP of Nigeria has been studied with a mean per capita GDP of almost $509 having a SD of $343 which shows that there is a great variation in the per capita GDP of Nigeria since the last 42 years. Though the kurtosis and skewness are following a normal distribution since both the values are close to 1. The FDI manufacturing has a mean investment ratio with the level of GDP is 1.0717 with a very high standard deviations of 1.47 again showing that the variation exists. Now in FDI, the kurtosis and skewness is are a little also very close 0 which shows that the data is normally distributed therefore its high SDs cannot imply to remove this variable. The similar case is with the FDI telecom the SDs are very high but its skewness and kurtosis are close to normal distribution.

The variable of openness in the trade is an important variable with a mean of almost 61% openness and most importantly with a low standard deviation of only 3% which makes this very much reliable. The similar case is with the human capital that its mean is 41% investment on secondary and university enrolment with a SD of 3.75 which quite acceptable for such a large mean. The political risk factor also seems to be significant with a less SD of 0.065 which is showing an almost same political risk throughout the 42 year time period. The rate of inflation is another significant variable with a very acceptable SD of 3% along with the mean of 19% inflation.

Infrastructure development and Returns on investment are little ambiguous in the decision making of being significant or not because they have SDs and therefore we cannot conclude their significance without analyzing much of their significance. All those variables which are found to be a little deviating from the mean will now be checked by comparing their means through t-test statistics. The descriptive statistics does not give ideal relation for the significance of the variables so therefore t-test is used and if they found significant they will be included in the model otherwise will be eliminated because including insignificant variables in the model can lead to unreliable results.

Descriptive Statistics

GDP per capita

Foreign direct investment on manufacturing

Foreign direct investment on Telecom

Openness in the economy

Human capital

Political Risk

Government size (consumption as a ratio of GDP)

Inflation rate

Returns on investment

Infrastructure development

N

Valid

42

42

42

42

42

42

42

42

42

42

Missing

18

18

18

18

18

18

18

18

18

18

Mean

509.90

1.0717

1.0134

69.6429

41.3095

.6233

.8355

19.1667

4.8155E8

2.9098E6

Mode

246a

-.50a

-.50a

69.00

40.00a

.59a

.85

15.00a

-1.99E7a

680165.00a

Std. Deviation

343.024

1.47958

1.26919

3.12988

3.75784

.06577

.07924

3.96273

9.35458E8

1.42189E6

Variance

117665.552

2.189

1.611

9.796

14.121

.004

.006

15.703

8.751E17

2.022E12

Skewness

1.388

2.077

1.903

.138

-.210

-.248

.508

.112

3.474

-.020

Std. Error of Skewness

.365

.365

.365

.365

.365

.365

.365

.365

.365

.365

Kurtosis

1.023

3.637

2.884

-1.157

-1.247

-1.038

-1.025

-1.400

13.593

-1.478

Std. Error of Kurtosis

.717

.717

.717

.717

.717

.717

.717

.717

.717

.717

Range

1296

6.33

5.37

10.00

12.00

.23

.25

12.00

4.88E9

4265796.00

Percentiles

25

258.25

.2258

.3064

67.0000

38.0000

.5700

.7675

15.7500

-1.5684E6

1.5638E6

50

375.50

.4504

.4504

69.0000

42.0000

.6350

.8150

19.0000

1.5955E8

3.1005E6

75

644.00

1.2760

1.2760

72.0000

45.0000

.6800

.9050

23.0000

8.7395E8

4.2328E6

a. Multiple modes exist. The smallest value is shown

Analysis of Correlation

Here we will focus on the correlation between two variables. Calculations of two-dimensional criteria such relationships are based on the formation of binary values, which are formed from a consideration of dependent samples. If for example we take the data on the level of cholesterol for the first two moments of time from the study of hypertension, in this case we should expect a fairly ...