2025 – PAGE 389 – STATISTICS

POSITIVE PREDICTIVE VALUE = TP/(TP+FP)

The positive predictive value [TP/(TP+FP)] looks at all of the POSITIVE results from a test and predicts how likely it is that a positive result actually means a patient has the disease. For example, what are the chances that a patient with a positive flu test actually has the flu?

NEGATIVE PREDICTIVE VALUE = TN/(TN+FN)

The negative predictive value [TN/(TN+FN)] looks at all of the NEGATIVE results from a test and predicts how likely it is that a negative result actually means a patient does not have the disease. For example, what are the chances that a patient with a negative flu test in fact does not have the flu?

NULL HYPOTHESIS

The null hypothesis is stated at the beginning of a research study. It assumes that the outcome being studied is the result of pure chance. In other words, it assumes that there is no variable (environmental exposure, medication, etc.) that could result in an increase in this outcome. For example, if studying the association between smoking and lung cancer, the null hypothesis would state that there is no increase in lung cancer in patients who smoke.

P VALUE

The P value represents the chance that the null hypothesis was rejected in error. In other words, what is the chance that the null hypothesis was accidentally rejected? Or what is the chance that the difference in outcomes of a test or exposure was simply due to chance?

SIGNIFICANT RESULTS

Results of a study are considered significant if the P value is < 0.05. They are considered “highly significant” if the P value is < 0.01.

TYPE I ERROR

Type I errors basically mean that a study claimed there was a significant difference when there in fact was not one.

MNEMONIC: type “i” error. i made the error and said this is the greatest test on earth when I shouldn’t have. Or, i claimed fame for discovering an association between patients born in December and a higher incidence of cancer. It turns out “i” just read the data wrong and made a type “i” error.

TYPE II ERROR

Type II errors basically mean that a study claimed there was NO significant difference in the results when there in fact was one.

MNEMONIC: TWO rhymes with YOU. In a type TWO error, YOU made the mistake of claiming that my test or medicine doesn’t work, when in fact it does! Think of it as a competition between two drug companies.

PREVALENCE

Prevalence refers to the proportion of a population found to have a condition at a given point in time.