National open university of nigeria introduction to econometrics I eco 355

10. How to we determine power?
Unfortunately, we don't know "beta", the exact value of the power of a test. We do know, 
however, that the power of a test is effected by:
(i). Alpha Level
: Reducing the value of alpha also reduces the power. So if we wish to be 
less likely to make a type I error (conclude there is an effect when there isn't) we are also 
less likely to see an effect when there is one.
(ii). One-Tailed Tests
: One tailed tests are more powerful. They make it easier to reject 
null hypotheses.

---

111 
(iii). Sample Size
: Larger samples are better, period. Tests based on larger samples are 
more powerful and are less likely to lead to mistaken conclusions, including both Type I 
and Type II errors.
4.0. CONCLUSION 
The unit concludes that hypothesis testing is an act in statistics whereby an analyst tests 
an assumption regarding a population parameter. The methodology employed by the 
analyst depends on the nature of the data used and the reason for the analysis. Hypothesis 
testing is used to infer the result of a hypothesis performed on sample data from a larger 
population. 
5.0 
SUMMARY 
We then summaries that 
Procedure for deciding if a null hypothesis should be accepted or 
rejected in favor of an alternate hypothesis. A statistic is computed from a survey or test 
result and is analyzed to determine if it falls within a preset acceptance region. If it does, 
the null hypothesis is accepted otherwise rejected. 
 
6.0.
 
TUTOR MARKED ASSESSMENT EXERCISE 
List the stages of hypothesis testing in statistical analysis 
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. 

---

112