Recall that when it comes to the experimental method, the null hypothesis states that there no cause-and-effect relationship between two variables, whereas an alternative hypothesis states that there might be some cause-and-effect relationship between these variables. When there is an actual cause-and-effect relationship between two variables, you can expect a significant result from an experiment.
While you may carefully perform experiments to establish whether or not these two variables are related, you still need to be careful to prevent an error that might come from the results of the experiment. The two major errors that are possible when conducting experiments are rejecting the null hypothesis when it is actually true, and failing to reject the null hypothesis when it is actually wrong.
When two variables are not actually related in terms of a cause-and-effect relationship, it is still possible to see a significant result from an experiment. An example of such errors could be when a doctor orders a lab to conduct blood work for a specific medical condition for a specific patient.
Since the test is unlikely to be perfect, there exists a chance that the person interpreting the results rejects the null hypothesis when they shouldn’t. In a case like this, the null hypothesis is that the patient does not have the condition. This means that the test is erroneously interpreted as the patient having the condition. This is also known as a Type I error, or a false positive. It is possible that the error could lead to a misdiagnosis and treating for a condition that the patient simply doesn’t have.
On the flip side, it is possible that the blood test is not rejected when, in fact, it should be. It could lead to a lack of a diagnosis where the patient actually has a condition but winds up not being treated for it due to such an error. An error such as this is referred to as a Type II error, or a false negative.
It is very helpful for a researcher to know if they run the risk of committing either type of error. It can have a major impact on whether or not the results are being communicated accurately.
Suppose that a researcher conducted a study that showed a result as being significant. This is a clear indicator that the variables are related, when in fact, they really aren’t. This an example of a Type I error. Such an error would lead the researcher to believe that the sleeping habits of people may in fact be influenced by how tall they are.
Suppose that a researcher conducted an experiment to test this relationship and discovered that the results were not significant. This would be a possible indicator that a Type II error, or a false negative, occurred. This could lead to the conclusion that larger homes don’t actually take more building materials to construct. This is something that could turn out to be quite costly to a builder.
Source: This work is adapted from Sophia author Dan Laub.