Research Methods - Coursework Example

The value of the statistic is too small and the probability is too high to reject the null hypothesis. Thus, from the F-test we conclude that all coefficients could be jointly equal to zero. (ii) H0:?2=0 against H0:?2?0 using a significance level of 0.05 Since the alternative hypothesis is that of non-equality but no direction (greater or less than) is specified, the test will be two tailed. The computed t- statistic is equal to -2.66937 which is greater in terms of absolute value than the two-tailed 5% critical value of 2.

018 (given the number of observations and variables, the degrees of freedom are 42), we reject the null hypothesis that the coefficient of 1990 GDP per-capita is not statistically significantly different from zero. (iii) H0:?3=0 against H0:?3>0 using a significance level of 0.05 Since here the alternative hypothesis is of the greater than type, the test will be right tailed. For the given number of observations and variables, the critical one sided 5% t value is 1.682. Our computed t-value is 2.

598522 which is greater than the critical value. Therefore, the null hypothesis is rejected at 5% level of confidence. This implies that we have statistical evidence of secondary enrollment having a positive impact on GDP growth. (iv) H0:?7=0 against H0:?7>0 using a significance level of 0.1 Again, the alternative is of the greater than type, implying a right tailed test. The critical 1% t-value is 2.418. From the table above, we find that the computed t-statistic is 1.50471. Since this is smaller than the critical value, we fail to reject the null hypothesis.

Therefore, we fail to find any evidence that credit ratio has any statistically significant impact on GDP growth. Therefore, we find a contradiction between our conclusions in (i) and (ii). While in (i) we fail to reject the notion that all the coefficients on the predictor variables are jointly zero, we reject the hypothesis that the coefficient on the first explanatory variable, the 1990 percapita GDP is zero. But if this is true then (i) should have rejected the null in favour of the alternative which requires atleast one of the coefficients to be non-zero.

Typically, such contradictions arise because of the violation of one or more of the basic assumptions underlying OLS estimation. Particularly, if there are outliers that distort the estimates, then such contradictory results can emerge. 3. Advice for choosing between alternative spending From the fitted model in the previous part we have found that secondary enrolment has a positive impact on GDP growth as does the private credit ratio. Infact the coefficients are quite close though that of private credit ratio is slightly lower.

However, only the former is statistically significant. This implies that there is no evidence of increases in private credit ratio having any impacts on the GDP growth. Therefore I would recommend investing the sum of money on policy measures that will increase the country’s rate of enrolment in secondary education. 4. Diagnostics for evaluating the validity