It would be really helpful if someone could share some information regarding this and also some resources which could help me understand the above things much better.
Thank you
That means that both data groups are of same number of observations, while unpaired not equal numbers. Regards.
Hi Krishnapriya,
to give an example for paired t-test: If you want to compare the blood sugar levels of a group of people between morning levels and evening levels of the exactly same Individuals the t-test should be paired because the samples have a direct relation because they are from the same persons. If you would do the comparison with blood samples of two different groups of persons the t-test to apply should be unpaired.
the null hypothesis is the assumption that there is no difference in the compared groups. Taken the null hypothesis as the working hypothesis in place you can test if the null hypothesis fits for this you get a p value to show how good the null hypothesis fits to your comparison/observation.
If the p value would be 1 it means that the null hypothesis fits to 100%.
For a p value below 0.05 (null hypothesis fits to less than 5%) it means that the null hypothesis fits not well and that there is place for bringing in another hypothesis, that there is a "significant" difference between the compared groups.
I hope the above explanation helps
Best,
Murat
Check out these sites, which has clearly stated the answer of your questions.
https://www.bmj.com/about-bmj/resources-readers/publications/statistics-square-one/7-t-tests
https://www.statisticssolutions.com/free-resources/directory-of-statistical-analyses/paired-sample-t-test/
3. Null hypothesis is an assumption without any relevant evidence.
2. The level of significance is generally taken 1%(99,% purity or accept), 5%(95%purity or accept),
We many take any 1,2,..50%, etc.
Generally, level of significance referd to level of rejection region.
1) the paired t- test is used for the comparison of test results before and after some process.
Ex. 1) class test and after some special coaching the test results in the same portion, hear same number of students appeared in the test, so that this is called paired t - test.
Anything in statistics is not 100%, so that we must choose the level of significance.
Medhat Elsahookie , "That means that both data groups are of same number of observations, while unpaired not equal numbers."
This is not the distinction between the paired t-test and unpaired t-test.
Answering this question would require quite a long introduction to null hypothesis significance testing. For an accessible introduction to the paired t-test, I might recommend the Handbook of Biological Statistics. ( http://www.biostathandbook.com/pairedttest.html ). For an accessible introduction to hypothesis testing, I might recommend the chapter from the same ( http://www.biostathandbook.com/hypothesistesting.html ).
Murat Eravci your explanation may be somewhat misleading. since you cannot "test if the null hypothesis fits" with a (frequentistic) t-test. What you get is the probability of your observed or more extreme test statistic, IF the null hypothesis is true. You do not test the fit of the null or how likely it is.
- Badges
- Science topic
- Similar topics
- Biological Science
- Molecular Cell Biology
- Culture Cells
- Cell Line