statistics question

[indigoblue22]

Sorry if this is a dumb question, but I can't figure it out from FA or high yield behavioral science...

how exactly are predictive values influenced by diseaes prevalence? I know that a higher prevalence means a higher positive predictive value, but how exactly do you calculate the PPV based on the number of cases in the population? I've tried drawing 2x2 boxes but still can't get these Qbank questions right...

Thanks for any help!

---

Members don't see this ad.

 

[BrooklynDO]

Sorry if this is a dumb question, but I can't figure it out from FA or high yield behavioral science...

how exactly are predictive values influenced by diseaes prevalence? I know that a higher prevalence means a higher positive predictive value, but how exactly do you calculate the PPV based on the number of cases in the population? I've tried drawing 2x2 boxes but still can't get these Qbank questions right...

Thanks for any help!

umm take this with a grain of salt because stats are not something im decent with...but this was my understanding....
its not based on any calculated prevalence (not based on duration X incidence= prevalence)

a PPV is basically "how many of those who came up positive are actually positive" so take the true positives and divide by all positives... this means devide by true positives and false negatives

the same goes for negative predictive value... true negatives devided by all negatives (meaning devided by the true negatives plus false positives)
oh and if im wrong someone correct me...this is just my understanding of a subject ive only seen in kaplan at some point

 

[nuclearrabbit77]

Sorry if this is a dumb question, but I can't figure it out from FA orhigh yield behavioral science...

how exactly are predictive values influenced by diseaes prevalence? I know that a higher prevalence means a higher positive predictive value, but how exactly do you calculate the PPV based on the number of cases in the population? I've tried drawing 2x2 boxes but still can't get these Qbank questions right...

Thanks for any help!

predictive values are going to be dependent on the number of people who have disease in the first place. if 99% of the population has a disease, even if the test has mediocre sensitivity/specificity, the positive predictive value of a positive result on the test will still be high, because so many people have the disease and so few of population doesn't. the best way to reinforce this concept is to make a 2x2 and in one scenario create a disease with a high prevalence and one scenario with a low prevalence and see how it affects the calculation of ppv.

hope this helps.