Inferential analysis, non parametric and parametric testss Flashcards
What are non parametric tests?
These are the test that use nominal or ordinal data, it includes 5 tests.
Spearman's Rho Man Whitney U Chi squared Wilcoxon Sign test
How to see if your results are significant or not?
After you have chosen and completed your test you will end up with a calculated/observed value. You will also have your critical value.
To check if the calculated value is significant you will need to check it against the critical values in the tables for that test.
To read the critical value table you need 3 things:
1)You need to know if your hypothesis is 1 or 2 tailed.
2) The N value/ degrees of freedom value.
3) The level of significance.
Using this information you can get the appropriate critical value. In some test the calculated value needs to be larger then the critical value to be significant in some tests it needs to be lower.
In Chi squared and Spearman’s rho the calculated needs to be larger then or equal to the critical value. In all the parametric tests (related t-test, unrelated t-test and Pearson’s r) the calculated value also has to be equal to or larger then the critical value.
In Sign Test, Mann Whitney U and Wilcoxon test the calculated needs to be lesser then or equal to the critical value.
Things to remember when using inferential (statistical tests)
1) Justify your choice of statistical test, refer to whether it is a test of difference, correlation or an association, what design and level or measurement has been used.
2) calculate the results of your statistical test.
3) Compare the results to your critical value to see if they are significant or not?
4) Mention whether the hypothesis was 1 or 2 tailed and significance level used.
5) State if you have accepted or rejected the experimental hypothesis and the null hypothesis.
What are parametric tests?
These are the 3 tests that use interval data:
Related t-test
Unrelated t-test
Pearson’s r
