I have a problem.I want to find if there is a statististical significant difference between two groups which I have put through a practical test.The outcome was 1 or 0 (success/fail).Group 1 was subject to extra training material before the "exam". Group 2 was not.Each group had one attempt in three different scenarios: easy - medium - hard.Lets take the medium scenario as example:The number of participants in each group is 10Group 1 scored: 7 successes, 3 failures --&gt; success rate 0.70 Group 2 scored: 2 successes, 8 failures --&gt; success rate 0.22Now what I have done, is try to use the Z-test for two population proportions in an attempt to reject the null hypothesis, buuuuut, I think my sample size is to small for the z-test?Im not sure what test to use at this point, can anyone give me a clue?Thanks for reading

Question

I have a problem.I want to find if there is a statististical significant difference between two groups which I have put through a practical test.The outcome was 1 or 0 (success/fail).Group 1 was subject to extra training material before the &quot;exam&quot;. Group 2 was not.Each group had one attempt in three different scenarios: easy - medium - hard.Lets take the medium scenario as example:The number of participants in each group is 10Group 1 scored: 7 successes, 3 failures --&amp;gt; success rate 0.70 Group 2 scored: 2 successes, 8 failures --&amp;gt; success rate 0.22Now what I have done, is try to use the Z-test for two population proportions in an attempt to reject the null hypothesis, buuuuut, I think my sample size is to small for the z-test?Im not sure what test to use at this point, can anyone give me a clue?Thanks for reading

Mollie Roberts · Accepted Answer

Comment continued: Here is output from Minitab statistical software for comparing two proportions. There is a warning message for the approximate normal test. The result for Fisher&#039;s exact test is also shown (with larger P-value).Please look at textbook, class notes, or online for a suitable explanation of Fisher&#039;s test. (Because you have shown no work of your own, I have no way to judge your background or what level of explanation to provide.)Test and CI for Two Proportions Sample  X   N  Sample p1       7  10  0.7000002       2  10  0.200000Difference = p (1) - p (2)Estimate for difference:  0.595% lower bound for difference:  0.183606Test for difference = 0 (vs &amp;gt; 0):  Z = 2.60  P-Value = 0.005* NOTE * The normal approximation may be inaccurate for small samples.Fisher’s exact test: P-Value = 0.035In case it helps, here is a computation of the Fisher P-value from R statistical software:sum(dhyper(0:2, 10, 10, 9))[1] 0.03488926You have an urn with 10 red balls and 10 white balls, you withdraw 9 balls from the urn without replacement. What is the probability that two or fewer of the 9 balls drawn are red ones? Sum three terms of the appropriate hypergeometric PDF (or PMF). [Computations by hand are a little tedious, but elementary.] Perhaps see this Q &amp;amp; A or Wikipedia on Fisher&#039;s test.

I have a problem. I want to find if there is a statististical significant difference between two g

Answered question

Answer & Explanation

New Questions in College Statistics