In this scenario of probability testing we must decide if a certain drug is considered safe for trial on humans. The null hypothesis is the drug is considered safe. The alternative hypothesis is that the drug is considered unsafe.

The P value does not conclude that the null hypothesis is true, meaning in this case that the drug the pharmaceutical company wants to test on humans is safe. The P value would be the measurement of how much evidence there is against the safety of the drug (Welkowitz, Cohen, & Lea, 2012).

The P value set by the researcher, indicates how strongly the evidence is against the null hypothesis ("P-value," 2009). So in this case, the P value would measure how much evidence there is against the safety of the drug being studied. If the P value is set at 0.05 then any number below 0.05 would reject the null hypothesis. A more confining P value that is set at 0.01 means that any number below 0.01 would reject the null hypothesis (Welkowitz, et. al., 2012).

If a Type I error was identified, the study drug would be found unsafe when it was actually safe. An error of this type could have an impact because this drug may have been a useful treatment for depression not helped by existing medication (Welkowitz, et. al., 2012).
If a Type II error was identified, the study drug would be found safe when it actually was unsafe. In this study, a Type II error would be worse than a type I error because the drug was found to be safe when it was actually unsafe which could dangerously harm people if administered in the study.

In this example when human trials are in question it would be better to use 0.05 as the rejecting value because if 0.01 is used, it could fail to reject the hypothesis because as the null hypothesis area becomes larger, the alpha becomes smaller and could fail to reject the null hypothesis (Welkowitz, et. al., 2012). If the results fail to reject the null hypothesis, it could cause a Type II error which would mean that the drug was found safe for human trial but actually it was unsafe.

If the null hypothesis was not rejected and the drug was considered safe for human trials then I would recommend human trials.

References - Retrieved from Explorable.com (Jun 23, 2009). P-Value. Retrieved Dec 16, 2014 from Explorable.com: https://explorable.com/p-value

Welkowitz, J., Cohen, B. H., & Lea, R. B. (2012). Statistics for the behavioral sciences (7th ed.). Hoboken, New Jersey: John Wiley & Sons.