Hypothesis Testing

HYPOTHESIS TESTING DR. SEEMA MUMTAZ

HYPOTHESIS • It is a supposition, which one develops for any usual or unusual happening • Is the testing in the analysis of bio-medical data usually relates to deciding whether or not the parameters (mean, proportion, relative risk etc. )in a study population, which can be estimated only by observing the sample., is equal to the values given in by the hypothesis. • There are two types of hypothesis; NULL & ALTERNATE .

HYPOTHESIS TESTING (Example) • A medical officer wants to use the best possible management toll for dehydration in diarrhea. He is informed about two possible management techniques: • ORS (Oral rehydration salt)Local Herbal Solution He wants to know which one is better, so the question that’s comes in his mind is: • In this population is their a difference in over all improvement after three days of treatment between the ORS and the Herbal solution????

HYPOTHESIS TESTING (Example) • There could be only two answer to this question, YES OR NO • There is no difference between such improvement (NULL HYPOTHESIS) • There is a difference between the improvement achieved by a three day treatment with the ORS & That of the herbal solutions.(ALTERNATE HYPOTHESIS) these statements are called null & alternate hypothesis and corresponds to YES OR NO.

HYPOTHESIS TESTING • Hypothesis testing procedures are always subject to some errors.  α (Alpha) error.  β (Beta) error. You will never know whether the hypothesis is true or false, if not tested or not in the real sense.

TYPE OF ERROS • TYPE ONE ERROR When we mistakenly reject the null hypothesis when indeed the null was true, then the type wrong decision is known as type one or α error. • TYPE TWO ERROR When we mistakenly accept our null hypothesis when in fact it is false then we commit type two or β error.

P - VALUE • P Stands for probability • P value is the probability of committing type one or α error in a research study. • Alternatively this is the risk of erroneously rejecting the null hypothesis when, in fact, it was true. • Also expressed as the level of significance • Smaller the P-value, more significant are the re