False Positive

This type of error involves the rejection of the null hypothesis when it is actually true. It is also known as a Type I error or α error. The null hypothesis, a statement of equality, always asserts that there is no relationship between the independent variable and the dependent variable. It is this assertion that forms the basis of inferential statistics, which can be used to determine our level of confidence that our findings (based on data from a random sample) represent true patterns in the population. A statistical test's level of significance (α, or alpha) [Page 658]indicates the risk of a false-positive finding. A level of significance of 0.05 is normally accepted in social science; this indicates that we accept a 5 percent chance of making a false-positive error. At more demanding significance levels, such as 0.01 or 0.001, the probability of making false-positive conclusions are lower.

A false positive involves the rejection of the null hypothesis when it is actually true, and is also known as a Type I error.

For example, consider a study that observes a 2.5-point difference in mean body mass index between groups A and B. A researcher may use a t-test to examine the statistical significance of that difference in mean scores. If the resulting test has a significance value of less than 0.05, the researcher would conclude that the difference between the groups is real. However, if the significance value of the t-test was greater than 0.05, the likelihood of a false-positive finding would be too great and the researcher would conclude that the groups do not differ in body mass index score (even though their results indicated a 2.5-point difference).