what or how do you find out if the null hypothesis is true or false?
-
It depends on how random the results of your experiments are, to be precise how significant your results are.
You understand the significance of a clinical trial/experiment after you do a statistic test, such as a t test to understand whether the results that you are obtaining are due to, for example, symptoms correlated/due to the action of a drug or whether these results are "random", aka uncorrelated to the drug that you are testing, but to inter-subject diversity.
This means that these results are due to other random events, that have nothing to do with the parameters that you are testing, they may be due to the sex, age, clinical history and other factors, e.g. IBD, other diseases, in the case of a medical trial.
In other cases we just say that results are significant if they exclude case, but are due to the experiments... E.g. standard deviation of the errors are a way to express the significance of a test.
The null hypothesis can be chosen depending on the confidence interval you choose and therefore the p-value (probability)... Usually the set confidence interval is of 95% meaning that the p-value is 0.05.
In this case if the value that you obtain from an Anova test for example, or another test is higher p>0.05 then you can say that the difference between the results/subjects is not significant, therefore the null hypothesis is true (probability that the results are due to randomness, to "case", is 95%, very high), whereas if your result is <0.05 then you can say that the null hypothesis can be rejected and the results are actually caused by the parameters that you are changing, in your case, by a drug you are testing...
You understand the significance of a clinical trial/experiment after you do a statistic test, such as a t test to understand whether the results that you are obtaining are due to, for example, symptoms correlated/due to the action of a drug or whether these results are "random", aka uncorrelated to the drug that you are testing, but to inter-subject diversity.
This means that these results are due to other random events, that have nothing to do with the parameters that you are testing, they may be due to the sex, age, clinical history and other factors, e.g. IBD, other diseases, in the case of a medical trial.
In other cases we just say that results are significant if they exclude case, but are due to the experiments... E.g. standard deviation of the errors are a way to express the significance of a test.
The null hypothesis can be chosen depending on the confidence interval you choose and therefore the p-value (probability)... Usually the set confidence interval is of 95% meaning that the p-value is 0.05.
In this case if the value that you obtain from an Anova test for example, or another test is higher p>0.05 then you can say that the difference between the results/subjects is not significant, therefore the null hypothesis is true (probability that the results are due to randomness, to "case", is 95%, very high), whereas if your result is <0.05 then you can say that the null hypothesis can be rejected and the results are actually caused by the parameters that you are changing, in your case, by a drug you are testing...
-
You don't know until you have done the experiment and analysed the results. The null hypothesis is the idea or proposal that you have before you start, the truth or falsehood of which you want to test. The null hypothesis might be that Eskimos living in New York earn more than Eskimos living in Chicago. You have to prove or disprove this by interviewing the Eskimos. The null hypothesis might be that tomatoes grown in a greenhouse are bigger than tomatoes grown in the open air. Again, you have to weigh all the tomatoes before you can say that the null hypothesis is true or false.