 Hello everyone. Welcome back to a new session on dentistry and more. So today we have a very interesting topic. So I suggest you to watch the full video, otherwise you won't get the concept. This is a very confusing topic which we might have seen during our post graduation and the under graduation. It looks very easy, but it is very confusing and it get changed between the types of errors most of the time. So when we study it, we get the concept, but after a while we get confused. So types of errors. So I have some common life examples. So the concept you can easily grasp it with comparing the types of errors with common life examples. So it won't get changed here after. So what's the video in full to understand what is type one and type two errors. Otherwise you won't get the concept if you're watching half or midway. So let's see what is types of errors. So if you're accepting a truth, it is perfect and if you are rejecting a truth that becomes type one error. And the second scenario is about a lie. If you're rejecting a lie, that is okay. But if you are accepting a lie, then becomes type two error. More examples you see you get the concept very clear. So this is a common scenario in our daily life. So we are rejecting a truth type one error. We are accepting a lie that is type two error. So only two things can happen. It's more than that. So type one and type two errors. Okay. So let's see the next example. Very convincing pictures for type one error and type two error. So if you have this picture in your mind, you won't get any confusion. This example, we have two scenarios. That is, we are telling a male person that is pregnant and we are telling a pregnant lady that she's not pregnant. So this is a type one error that is false assurance or false positive error. This is type two error that is false negative error. Actually she's pregnant, but we are giving a false negative information. This is we are giving a false positive information. Okay, so the positive or negative, we are mentioning about pregnancy. So we are giving a false positive pregnancy here. We are giving a false negative pregnancy here. Actually she is pregnant. So we are telling a false negative information. And here we are giving a false positive information. So this is type one error and this is type two error. So type one is always positive. That is false positive and type two is always negative. That is false negative. Okay. So this example will be the best example to get the concept because pictures will be easiest way to learn this type one or type two error. Now let's move on to a little more higher degree. That is, we are checking a new device. So next we have a new example to understand what is false positive and false negative error. So this is a new device to detect diabetes. We have a gold standard device. So we are trying to assess the properties of new device. Okay. So we have a few patients. So we do diabetes check of one patient by the new device and it is giving a positive result. Okay. So we don't know whether he is actually diabetic or not because it's a very new device. So on the second stage, we are checking it on a gold standard device. Okay. So where we get a negative risk. So that becomes type one type one error or false positive error. Because the first device or the new device has given a false assurance or the false positive or the false confirmation of the disease actually is not diabetic because we have a gold standard machine. We have H1, HB1, AC detection method. So we checked it. So we got the result as the patient is not diabetic. That is the type one error, false positive error. You are giving false assurance. The device is giving false assurance that is diabetes. The next scenario, we have a different patient and we get checked patient with the new device for diabetes and the device is giving a device is saying that he's healthy or he is not having diabetes. That is new devices given this information to make it confirm. We go to the gold standard in gold standard test. We get to know that this patient is diabetic. Actually, we got a false negative false negative information. Actually, the patient is diabetic, but this new device has given a false negative that is a false negative status of the patient that is actually is diabetic, but this new device is saying that this patient is not diabetic. So that is a type two error. So the common scenario is new device is saying that he's positive and the gold standard is saying that he's positive. That is true positive. The new device is saying that patient is not diabetic or negative and the gold standard is also saying that it is negative. That is true negative. So other two scenarios are the new device is saying positive new devices saying negative and the gold standard new device is sorry the new device is saying positive and the golden standard saying it is negative that is false positive error at the scenario is the new device is saying that the patient is diabetes that is false negative error and we get to know that the patient is actually diabetic it is a little bit confusing when we think of this positive negative and disease no disease contingency table we have to think that that there will be two types of examination okay so finally we have to compare the result with a gold standard so it gets confused most of the time we have to compare it with gold standard so whatever the result you get with your new test we have to compare it with gold standard so whatever the gold standard says that is ultimate truth so if you get positive and gold standard is saying negative that becomes false positive if you get negative and the gold standard is saying positive that becomes false negative error so these are the two types of errors which happens in a diagnosis testing okay so we have seen these three examples I hope you might have a clear picture about this now we go to the actual scenario that is hypothesis testing so before that you need to understand what is hypothesis hypothesis is nothing but an assumption or a suggestion that there is this much difference or this causes this is an assumption there is no concrete proof for it okay so we should know what is null hypothesis and what is alternate hypothesis null hypothesis states that there is no difference between two things alternate hypothesis says that there is a difference or this causes this null hypothesis says that this can't be the reason for this or this is not causing this okay so same scenario we apply here this is hypothesis is true and this is hypothesis is false and we are taking a decision here okay so actually the null hypothesis is true okay and we accept a true null hypothesis that is very good conclusion okay and what if we reject a true null hypothesis is type one error or false positive error that is null hypothesis is true but we are rejecting it null hypothesis is true means there is no difference between the two groups but we are saying that there is a difference that is a type one error or false positive error there is no difference we are rejecting the null hypothesis and says that there is a difference that is type one error the second scenario the null hypothesis is false actually hypothesis is false means there is a difference between the groups and we reject the false null hypothesis is fine it is very good good conclusion but what if we accept a false null hypothesis okay null hypothesis is false means there is a difference so we should reject it but we are accepting it and we are saying that there is no difference so these are the two things which can happen in hypothesis testing either to reject a true null hypothesis or to accept a false null hypothesis rejecting a null hypothesis means we are saying that there is a difference accepting a false hypothesis says that there is no difference because false hypothesis hypothesis is false means there is a difference to learn what is type one error and what is type two error so type one error is also known as or also represented by alpha symbol or alpha okay so it is incorrect rejection of a true null hypothesis there is no difference but we are saying that there is a difference that is rejection of a null hypothesis that is null hypothesis null means no difference actually no difference but we reject it and we are saying that there is a difference okay so always we keep 5 percentage of alpha that is we are accepting a fact that 5 percentage error can happen so we can say that out of 105 percentage type one error is permissible okay on 100 observations 5 times we can reject a true null hypothesis that is what it meant okay one example for type one error so let's say a null hypothesis that there is no wolf present okay so a type one error or a false falsity would be there is a wolf when actually there is no wolf okay so the actual condition was that there was no wolf present but the shepherd wrongly indicated that there is a wolf so that is a type one error or false positive error we are saying that there is a wolf when actually there is no wolf or there is no difference and we are rejecting the null hypothesis saying that there is a difference okay that is a false positive error when we talk about null hypothesis so that is type one and let's move on to type 2 error so type 2 error is also known as a beta error okay that is failure to reject a false null hypothesis null hypothesis actually false we are supposed to reject it but if we fail to reject a false null hypothesis that is type 2 error we should accept a true null hypothesis and we should reject a false null hypothesis if the opposite happens that becomes type one error and type 2 error okay so we should accept a true null hypothesis and we should reject a false null hypothesis if we accept if we reject a true null hypothesis that becomes type one error and we accept a false null hypothesis that becomes type 2 error so type 2 error is also known as beta error so sorry it happens when I one accept a null hypothesis that is actually false okay so let's see one example for beta error or type 2 error so the hypothesis is adding fluoride to toothpaste protects cavities also our null hypothesis is adding fluoride to toothpaste has no effect on cavities so null hypothesis is tested against our data so the type 2 error occurs when failing to detect an effect that is adding fluoride to toothpaste protect against cavities present so the null hypothesis is false okay that is adding fluoride is actually effective against cavities but the experimental data is such that null hypothesis cannot be rejected so we get a data and we are not in a position to reject a null hypothesis that is there is actually a difference but what happens with the data available we can't reject a false null hypothesis so there is type 1 and type 2 error false positive and false negative error I'll just start with our first example that is truth and lie we are rejecting a truth and we are accepting a lie that becomes type 1 and type 2 error next we seen two different example that is we are false assurance to a male person that is type 1 error and giving a false negative assurance that she's not pregnant that becomes type 2 error next we have seen a new device comparing with the gold standard so the new device says it is positive but gold golden standard saying it is negative so it becomes false positive error and if actually the new device is giving negative result and the golden standard giving positive that becomes type 2 error or false negative error so we'll come to the the core of the our topic that is hypothesis testing if we accept a true hypothesis that is true positive if we are rejecting a false hypothesis that is true negative but what if we are rejecting a true hypothesis that is we are saying that there is a difference when actually there is no difference that is type 1 error or false positive error and we are accepting a false hypothesis there is actually a difference and we are saying that there is no difference it is type 2 error so type 1 error the example the shepherd saying there is a wolf but actually there is no wolf and type 2 error the example we have seen is that there is a type 2 error we are failing to detect effect that is flow rates toothpaste actually protects against cavities but with the data we are saying that it doesn't protect the cavities the flow rates so that is type 1 and type 2 error okay so that's a little tricky part of research selecting a type 1 and type 2 error this type 1 and type 2 error should be included in our study while calculating the sample size type 1 is alpha and type 2 is beta type 1 we keep always 5 percentage that is out of 105 chances to reject a true hypothesis and we keep type 2 error almost 20 percentage that is accepting a false hypothesis failure to accept a false hypothesis so that's all about types of errors i'll come up with a new session and then to stream more thank you for watching