National open university of nigeria introduction to econometrics I eco 355

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
88 percent. Therefore, we do not reject the hypothesis that the error terms are normally 
distributed. But keep in that the sample size of 55 observations may not be enough. 
We leave it to the reader to establish confidence intervals for the two regression 
coefficients as well as to obtain the normal probability plot and do mean and individual 
predictions. 
4.0 
CONCLUSION 
We conclude in this unit that we usually apply normality tests to the results of processes 
that, under the null, generate random variables that are only 
asymptotically
or 
nearly
normal (with the 'asymptotically' part dependent on some quantity which we cannot make 
large); In the era of cheap memory, big data, and fast processors, normality tests should 
always
reject the null of normal distribution for large (though not insanely large) samples. 
And so, perversely, normality tests should only be used for small samples, when they 
presumably have lower power and less control over type I rate. 
 
5.0 
SUMMARY 
𝑅𝑒𝑠𝑖𝑑𝑢𝑎𝑙𝑠
Series: Residuals 
Sample 1.55 
Observation 55 
Mean 
×
4
Median
Maximum 
Minimum
Std.dev
Skewness 
Kurtosis
Jarque-Bera 
Probability 
𝑁𝑢
𝑚𝑏
𝑒𝑟
𝑜𝑓
𝑂
𝑏𝑠𝑒
𝑟𝑣
𝑎𝑡
𝑖𝑜𝑛
𝑠
 

---

137 
The unit discusses at length the normality test analysis such as the histogram of residuals, 
probability plot and Jraque Bera test of normality. Therefore, we can say that you must 
have learn through the test of normality at length. 
 
6.0 
TUTOR MARKED ASSESSMENT EXERCISE 
Discuss the differences between Normal probability plot and Jarque-Bera test of 
normality 
 
7.0 
REFERENCES/FURTHER READINGS 
Gujarat, D. N. (2007) Basic Econometrics, 4
th
Edition, tata Mcgraw – Hill publishing 
company limited, New Delhi. 
Hall, S. G., & Asterion, D. (2011) Applied Econometrics, 2
nd
Edition, Palgrave 
Macmillian, New York city, USA.