Hypothesis testing is trying to find statistical evidence to back up the rejection of a null hypothesis so that a person would begin to accept the alternative hypothesis. The researchers must determine the significance level for their test that is represented by alpha. An alpha level of 0.05 is very common. This means that there is a chance of the null hypothesis occurring only 5 times out of every 100 times (Grove & Cipher, 2017, p. 179). If the alpha is set lower and you can still show that the null hypothesis would not occur, having a lower alpha would give a greater significance to your test. However, the lower the alpha, the greater the risk for a type II error to occur. This is where the null hypothesis is false but the research showed that the null hypothesis should be accepted. Setting a low alpha may also make it harder for research to support rejecting the null hypothesis. Setting a larger alpha will make type II errors less likely to occur but type I errors are more likely (Jbstatistics, 2013). A type one error is there the null hypothesis is true but the research showed that it should be rejected (Grove & Cipher, 2017). It seems that you would want to use a lower alpha if rejecting the null hypothesis could result in damage to someone or something. You might accept a higher level of alpha (0.1) when testing a hypothesis that would not result in damage to someone or something.
The alpha level always depends on the nature of an experiment. Alpha level is a statistical figure used in determining the possibility of rejecting a null hypothesis when the null hypothesis is true which is usually computed by subtracting confidence level from 1 (Hox, 2017). Most researchers use an alpha level of 0.05 because they find it convenient to use. However, the alpha level can be fixed at a higher or lower depending on the factors that surround an experiment.
The alpha level can be as lower as 0.01 in some experiments. For example, in an experiment where research deals with an issue that is a matter of life and death the alpha level will be low to avoid risking any lives. In research about cancer where an individual seeks to determine whether growth in an individual is cancerous or not so that the growth can be removed an individual would fix an alpha level of 0.01 because the researcher would not want to remove the growth if it is not cancerous and risk a life. Also, the researcher would not want to risk not removing the growth and later it is found to be cancerous.
A higher alpha level can be used in issues that are not critical where no lives depend on the decisions to be made out of the experiments (Pett, 2015). In such situations, the alpha can be raised to 0.1. For example in an experiment where one is trying to find out whether an online brain dominance test is a true reflection of one’s normal functioning. The alpha level in this experiment can be raised since the experiment does not revolve around any critical issue but is instead just a fun activity.