DrVirgo
Forum Hero
Topics: 1096
Posts: 3,515
|
E. If we increase N (sample size) we increase the validity of a positive test result.... True?
**Sorry, but there no forum for Biostat/Epidemiology so I posted it here.
ANYONE?
Attached Files:
40.JPG (200 KB, 133 downloads)
Edited by DrVirgo on 07/23/08 - 04:00 AM
___________________
Our greatest glory is not in never falling, but in rising every time we fall.
|
| wasedf
Forum Junior
Topics: 5
Posts: 74
|
Dr. Virgo
I would go with B, which is basically the prevalence of HSV. This question is a bit deeper than it seems.
Validity deals with how accurate is something that it is supposed to measure. In this case, they are asking about the validity of a positive test result, so this may mean how true is a positive test of HSV from this study, in other words, how sensitive is the test.
Sensitivity of a test is dependent on the prevalence of the disease.
E- talks about increasing the number of participants, which is N. This will improve the power of a study. It would come in handy, if the question was asking about testing two kinds of tests for HSV and seeing which is superior. In that case, increasing the power of a study helps us to infer that the tests do acutually differ and are not a result of chance.
|
| WaqasQureshi
Forum Senior
Topics: 5
Posts: 178
|
it is not sensitivity that is dependent upon PREVALENCE. They are askin about the post test probability and hence positive predictive value that is directely related with prevalence.
|
| WaqasQureshi
Forum Senior
Topics: 5
Posts: 178
|
B
|
| suniti
Forum Newbie
Topics: 2
Posts: 19
|
b
as it tells how valid is the result which we have got
|
| DrVirgo
Forum Hero
Topics: 1096
Posts: 3,515
|
Now that I think about it, it makes sense... B sounds like the right answer... thanks everyone. 
___________________
Our greatest glory is not in never falling, but in rising every time we fall.
|
| WaqasQureshi
Forum Senior
Topics: 5
Posts: 178
|
|
| wasedf
Forum Junior
Topics: 5
Posts: 74
|
Waqas, Thanks for the correction. Yes, they are asking about the PPV which is directly related to the prevalence.
|
| DrVirgo
Forum Hero
Topics: 1096
Posts: 3,515
|
Thats right... I remember it now!!!
Prevalence is directly related to PPV.
However Prevalence has NO effect on Specificity or Sensitivity
I think the example given was something like this:
If we increase the actual proportion of students affected with HSV, and the test yields a + we are more likely to believe the result. (PPV).
For example: Two groups with 100 people… Group 1 has 10 people with HSV and Group 2 has 90 people with HSV. (prevalence). If you get a positive test result from both groups, which one are you more likely to believe? The test result from group 2 because the prevalence is higher.
___________________
Our greatest glory is not in never falling, but in rising every time we fall.
|
| paheli
It'sAllAboutGoodKarma
Topics: 197
Posts: 2,720
|
 Good explanation to yourself! and to others.
___________________
Prepare as if you're the worst, Perform as if you're the best! As you dream, so you manifest. So, DREAM BIG!! When you face hardship, remember, God never gives you more than you can handle. Keep your face to the sunshine and you cannot see the shadows. Internship, Here I come!!
|
|
| |
| | | | | | | | | | |
Similar forum topics