Statistics – True positives

[Victoria Villanueva]

Hey – can anyone explain why question 23 on exam 1 is C and not D? I always remembered specificity as “yes means yes, no means maybe” and sensitivity as “no means no, yes means maybe”. Maybe I’m just not understanding how we determine true positives?
Thanks!

[Jonathan Zins]

Hi Victoria,
First, great work putting in the hours now and staying ahead of things, you’re going to do GREAT in the Spring!
Second, thank you for catching this error. However, the correct answer should be B, not C or D.

Let’s dive into this. A sensitivity of 98% means that the test will correctly identify 98% of people truly at high fall risk, but it will fail to identify 2% who are truly at fall risk.
A specificity of 21.2% means that the test will correctly identify 21.2% of people not at fall risk, but will also identify 78.8% as having fall risk when they do not

In other words, the chance of a false positive is very high with the test in option C.

The answer should be “B”. Your sensitivity is still very high, and the specificity is also pretty high. Because the specificity is high, there is less chance of a “false alarm”. In other words, if a person does screen positive, it is more likely they are actually positive and not a false positive.

Here is another way to look at it: Option b (Sn 97.5%, Sp 81.9%) provides the highest chance of obtaining a true positive because it combines very high sensitivity with moderate specificity. Sensitivity reflects the test’s ability to correctly identify individuals who have the condition — so a high sensitivity means most true cases will be detected. While specificity helps reduce false positives, the relatively high specificity in this option ensures that the number of false positives remains reasonably low. Together, these values increase the Positive Predictive Value (PPV), meaning a positive test result is more likely to be a true positive.

Hope this helps!
Jonathan

[Victoria Villanueva]

Hey!

Yes this helps a lot. I took the time to revisit the charts regarding the math related to sensitivity and specificity which also was helpful to visualize. I realize that high sensitivity tests cast a wide net, and will capture every true positive. However Sensitive tests will not be able to tell you anything regarding the false positives. Therefore a sensitive test is a good way to ensure every patient who is a faller will be identified as a high fall risk. Then, if a negative result comes up on a very sensitive test, you can be confident that it is a true negative since false negative rate is very low in a highly sensitive test. Then SNout makes sense to me.

That makes sense right?? I’m trying to think about sensitivity and specificity in a way that I can recall at any time or apply to any situation.

Thanks again!

[Jonathan Zins]

Hey Victoria,
Yes, that makes perfect sense! You are thinking about it correctly. I find “SnOUT” and SpIN” the easiest ways to remember this. Keep up the good work!

[Jonathan Zins]

Clinically, start with a sensitive test to screen broadly, then follow it with a specific test to confirm the diagnosis or risk 🙂

[Author]

Posts