Skip to main content
Search form
  • 00:01

    SPEAKER 1: All right, so let's talk a little bitabout diagnostic testing and some of the aspects of that.Some research focuses on developing new teststo diagnose diseases.So how does one know if a test is good or not?Unfortunately, no laboratory testis perfect and doesn't work 100% of the time.So we can imagine a test is sampling

  • 00:21

    SPEAKER 1 [continued]: from a giant pool of people with and without a certain disease.So let us use the flu as an example disease.A true positive is when we samplea person with the flu and the test says they have the flu.So that would be in this section here.This is all the people with the disease.

  • 00:42

    SPEAKER 1 [continued]: This is a true positive, the test comes back positive,and the people with the disease where the test comes backnegative is a false negative.Likewise, when you have true negative--these are healthy people without the flu--we have a true negative and then falsepositives when disease is detected and it's not, in fact,

  • 01:03

    SPEAKER 1 [continued]: there.This is also called a type I error.And then, again, a false negativeis when a test incorrectly says somebody doesn't have the flu,but they actually do have the flu.And this is called a type II error.And here's another way to think about this.It's a similar thing, except pregnancy is not a disease.Certainly, we don't consider it that.

  • 01:25

    SPEAKER 1 [continued]: But let's say this doctor is tellingthis man that he's pregnant.All right, unless some miracles happened,that's going to be a type I error or a false positive.He's not, in fact, pregnant, and the doctoris saying that he is.Now here in this situation, the woman is clearly pregnant.And even this woman is doing an ultrasound,

  • 01:46

    SPEAKER 1 [continued]: but she's telling the patient that she is not pregnantwhen she is.And so this is a type II error or a false negative.And then we also have the terms sensitivity and specificity.So if someone has a sensitive nose,that would mean that they could detect even the faintest,

  • 02:07

    SPEAKER 1 [continued]: smallest odors.The sensitivity of a test refers to the ability of the testto correctly identify those patients infectedwith a particular disease.If a test is 100% sensitive, it properlyidentifies all the patients with a disease.If it's, let's say, 70% sensitive,it identifies 70% of patients with the disease

  • 02:29

    SPEAKER 1 [continued]: and misses 30% who truly have the condition.So these will be false negatives.And then we also have specificity.So for example, if you're looking for a specific person--your aunt, for example--you wouldn't pick up strangers that vaguely resembleyour aunt, OK?Specificity of the test refers to the ability of the test

  • 02:51

    SPEAKER 1 [continued]: to correctly identify those patients not affectedby a certain disease.So if a test is 100% specific, it correctlyidentifies all the patients who do not have the infection.It has no false positives.Ideally, a test would have both high sensitivityand specificity.

  • 03:11

    SPEAKER 1 [continued]: But unfortunately, this is rarely possible.In HIV testing, for instance, often aninitially highly sensitive test is used.And those who test positive are thengiven a second test that is more specific to seeif the first result was a false positive or not.


Presentation of sensitivity and specificity in the result section of a scientific paper is discussed.

Looks like you do not have access to this content.

The Results: Sensitivity and Specificity

Presentation of sensitivity and specificity in the result section of a scientific paper is discussed.

Copy and paste the following HTML into your website