7ssmn100 Quantitative Methods. Assignment 2 Table Templates Research Paper

Part A. Correlation Table.
* significant at 95%; ** significant at 99%
There is a significant positive correlation between FDI and export, the significance level is 95 or 99%. In 1980, the correlation is weak, but close to moderate. In later years, the moderate correlation is observed. This happens because the FDI has had more influence on export recently.
The correlation between FDI and population is weak, and it is significant only in 1995 and 2010. In 1980 and 2007, the FDI and population does not correlate and the correlation coefficient is 0.
As for the relation between the population and export, it is significant in 1995, 2007, and 2010. In 2007 and 2010, the moderate correlation is observed.
It should be noted, that the relations between the variables was also influenced by other factors in 1980, for example political situation.

Part A1. Correlation Table for Middle Income Countries

The correlation for the middle income countries shows the same values as for all set of data. The values of correlation coefficients and the significance are the same. It shows that the middle income countries determine the character of FDI, export and population relation for the whole set of data.

Part A2. Correlation Table for Less Developed Countries

The correlation profile of the less developed counties is quite different. It should be noted that all the correlations are significant. The FDI and export shows high strong positive correlation, which indicates that FDI values are mostly determined by export. The same pattern is observed for the FDI and population relation, and the correlation is also strong, only in 2010 it is moderate. The export and population are also related, and the strong positive correlation is observed. The year 2010 presents slight difference comparing to other periods, and the correlations is less. Again, in 2010 the other factors have impact on FDI, export and population.

Part B. Total sample

Dependent variable = ln(Exports)
The character of the Export vs. FDI regression line is the same for all years. The determination coefficients (R2) are close for all years, and it is about 0.3. Thus, FDI variable explains only about 30% of the export variable. For all the studied cases, the intercept value is insignificant, which is apparent from the low values of the t-ratio. On the contrary, the FDI coefficients are in all cases significant (t-ratios are about 5).
When the second variable is introduced, the model becomes more reliable, and the determination coefficients are about 0.88 for 1980, 1995, and 2007. In 2010, the R2 is 0.67. This corresponds well with the results of the A section, namely the influence of other factors on export. The intercept coefficient is significant in 1980 and 1995, and it can be used for the export value assessment. The FDI and population coefficients are significant for all years, which is seen from the high t-values.

Part C. Division by country

Dependent variable = ln(Exports)
The regression model shows high determination coefficients for both groups of countries (0.8-0.9). Only in 2010, the determination coefficients are lower, namely 0.72 and 0.64 for the middle and less developed countries, respectively. The intercept coefficients in 1980, 1995, and 2007 are significant. The coefficients for FDI and population are also significant, and thus for 1980, 1995, and 2007 the regression model is Export = I + C1·FDI + C2·Population.
The FDI coefficients for middle and less developed countries are close, that indicates the equal proportion of FDI in export. As for the population variable, the values are close in 1980 and 1995. However, in 2007 and 2010, the proportion of population in export value increased. This proves that for the export growth in less developed countries is influenced by the population growth.
In 2010, the tendency of discrepancies from the other years is observed. Apparently, other the other factors apart from the social and economic, influence the process.

Part D. Including Dcountry

Dependent variable = ln(Exports)
The introduction of the dummy variable does not have a significant influence on the model validity, which is indicated by the determination coefficients. They remain high for 1980, 1995, and 2007, and is lower for 2010, the same as was observed in B(ii).
The value of the significant intercept coefficient changes to the negative value, possibly because the dummy variable takes its role. The values of the FDI increases in the model with dcountry variable, and the population variable value decreases. However, since the determination coefficients remained within the same value, the model with a dummy variable can be used for export estimation. This is explained by the big volume of the data set, and the statistical method smooth the influence of the dummy variable.