When a research writes H0 hypothesis and when to writes an alternative hypothesis?
https://www.dummies.com/education/math/statistics/how-to-set-up-a-hypothesis-test-null-versus-alternative/
H0= implies there is no difference between groups under study
Normally in research u want to reject this hypothesis
HA= There is difference between the groups under study
In research you want to accept it to conclude that the differences between groups are statistically significant
Regards
if there are enough previews studies supporting the researcher's prediction of the results, he can set an alternative hypothesis beside 0H and here we have two tails test.
if the study has not been studied enough to predict the results, here, the researcher would better to set only 0H with one tail test.
I am very old fashioned on this question. Ho is thy your idea will NOT work, this always has to be your starting point. Why? Because what you find DOES work may not be what you predicted. The only time H1 is used is in drug company PhD-mills where a preacribed piece of work is done to get a predicted outcome. This is not research in my view, it is manipulating most knows to achieve a predicted unknown, or a known but to what power/ Real research is tinkering
Formulating a hypothesis is one of the first requirements in conducting research. Notably, there is no difference in using either a null hypothesis or an alternative hypothesis. It is simply a matter of choice and the researcher's personal previllage.However, in the case of alternative hypothesis, it is better to use the negative alternative hypothesis because ,as Best and Kahn (2006) state, the positive directional hypothesis is logically inconsistent. In using both the null or negative, directional hypotheses, the goal should be the rejection of the targeted prediction mirrored by the hypotheses under scrutiny.
Regarding Reza: My view is aligned. To reject Ho you must state is clearly. If you test for H1 and find you are right, you have no proof Ho is wrong. It just seems wrong because you found H1 right. Therefore Ho remains a possibility until proven wrong
Absolutely agree with Mr. Ebrall. The aim of the research should be rejecting or confirming the null hypothesis. In other words, if the literature on a certain topic says something is one way, but your hypothesis claims something different, then the view presented in the literature will be your null hypothesis (certainly stated clearly). When you eventually test it and get a statistically significant result, you thus reject it and come to the conclusion that your alternative hypothesis is true. So, technically you can have both hypotheses at the start of the research. But the one you are going to test is the null hypothesis.
To write the alternative and null hypotheses for an investigation, you need to identify the key variables in the study. The independent variable is manipulated by the researcher and the dependent variable is the outcome which is measured
The alternative hypothesis states that there is a relationship between the two variables being studied . It states that the results are not due to chance and that they are significant in terms of supporting the theory being investigated.
The null hypothesis states that there is no relationship between the two variables being studied . It states results are due to chance and are not significant in terms of supporting the idea being investigated.
Thank you Armine. I really like and value the 'intricate simplicity' of Meliana's answer. This perspective allows clarity on which to use.
Hi sir, it doesn't matter if all hypotheses (alternative) are accepted? is there any references to it? Thank you
Can we state te NULL HYPOTHESIS under the research question directly in a dissertation? Or should we state the research hypothesis first , and then state the null hypothesis in DATA ANALYSIS section?
- Badges
- Science topic
- Similar topics
- Statistics
- Survey