Review Question - QID 103976

This test is reliable, as it consistently produces the same result for any given patient.

In statistics, there are several standards that a test must meet to be considered a useful test. First, it must be precise, also known as reliable, which means that it consistently produces the same results. Additionally, it must produce the correct results, also known as accuracy or validity. Validity is further measured by sensitivity (the ability of the test to correctly identify those who have the disease) and specificity (the ability of the test to correctly identify those who do not have the disease).

Mackrides and Ryherd discuss the barriers that exist in screening children for developmental delay. One of these barriers is the relative lack of good screening tools. Several validated tools exist, but the authors note that abridged versions, while time efficient, often sacrifice both reliability and validity.

Cook et al. undertook a systematic review to evaluate the reliability as well as the validity of the available tests for Chlamydia trachomatis and Neisseria gonorrhoeae infections. Most notably, they found that the urine test results for C. trachomatis were valid compared to more invasive gold-standard testing; however, the sensitivity of the urine test for N. gonorrhoeae was too low to recommend the test.

Illustration A depicts the difference between reliability and validity.

Incorrect Answers:
Answers 1, 4: For a test to be valid, it must produce an accurate result. This test does not since its results do not match the gold standard test for Alzheimer's.
Answer 3: Although we know the test's results do not match the gold standard, we do not know why. It is possible there is a systematic error in the results, which would be bias, but it is also possible that the error is simply random.
Answer 5: This test is reliable.