swathi
Forum Junior
Topics: 5
Posts: 20
|
A PHYSICIAN WISHES TO STUDY WHETHER MODERATE ALCOHOL CONSUMPTION IS ASSOCIATED WITH HEART DISEASE. IF, IN REALITY,MODERATE ALCOHOL CONSUMPTION LEADS TO A RELATIVE RISK OF HEART DISEASE OF 0.60,THE PHYSICIAN WANTS TO HAVE A 95%CHANCE OF DETECTING AN EFFECT THIS LARGE IN THE PLANNED STUDY .THIS STATEMENT IS AN ILLUSTRATION OF SPECIFYING A.ALPHA ERROR B.BETA ERROR C.A NULL HYPOTHESIS D.CRITERIONODDS E.STATISTICAL POWER
|
| swathi
Forum Junior
Topics: 5
Posts: 20
|
EXPLAIN WITH ANS
|
| kingsofke
Forum Guru
Topics: 24
Posts: 715
|
This an example of setting up alpha error (which is always done in a study and is now standarised to 95% confidence ,you will read it in almost any study as p value of .05)
beta error and statistical power are the most easy to remember. They are related only to number of people (for usmle purpose, this statement will definitely not be true). So when somebody is doing something with the no of persons in his study he is manipulating them And that is not the case in his study.
Alpha Error is a means of saying that if i say that null hypothesis is wrong (or what i originally meant to do/say is right) then what are the chances that i had made a mistake (ie the null hypothesis is actually right and not what i thought). You fix up the alpha error by p value. A p value of .o5 actually is a way of saying that there is 5% chance that null hypothesis is right or 5% chance of what i thought was wrong and 95% chance of what i thought is right. So there is 5% or 1 in 20 chance of having been wrong in making a conclusion.
p.s. Beta error is quite the opposite. its about accepting null hypothesis when it is wrong. Saying that i am wrong when you are actually not. and it cannot be accessed on p value (only one thing to remeber about it that the only modifying factor will be the n.
___________________
In Life there are neither rewards nor punishments, just consequences.
|
| robin082006
Forum Hero
Topics: 471
Posts: 5,125
|
Good explanation Kingsofte.
In summary
Alpha error: we reject the null hypothesis but in fact it is true.
Beta error: we fail to reject the null hypothesis but in fact it is false.
Power = 1-beta=probability of rejecting the null hypothesis when in fact it is false.
___________________
The Key to Succeed is Patience.
|
| doc179
Forum Guru
Topics: 67
Posts: 1,217
|
ok... what is the answer?
|
| robin082006
Forum Hero
Topics: 471
Posts: 5,125
|
E, statistical Power
___________________
The Key to Succeed is Patience.
|
|
| |
| | | | | | |
