I am looking for code in order to do a simple a two-proportion test for mutually exclusive groups. To my knowledge, a chi-square test tells you me that the two categories are related (not indepedent) and does not provide what I desire. I need to test whether the proportions of the groups are equal applying the the principles in this https://onlinecourses.science.psu.edu/stat414/node/268 or this online calculator http://www.socscistatistics.com/tests/ztest/Default2.aspx
Example question: Is the proportion of URM students who succeed (58.33%) different than the proportion of non-URM students who also succeed (77.06%).
Table of URM by Success_category |
URM | Success_category |
Frequency(n)
Percent
Row Pct
Col Pct | ABCs | DFWs | Total |
0(non-URM) | 84(n)
36.68
77.06%
54.55 | 25(n)
10.92
22.94
33.33 | 109
47.60 |
1(URM) | 70(n)
30.57
58.33%
45.45 | 50(n)
21.83
41.67
66.67 | 120
52.40 |
Total | 154(n)
67.25 | 75(n)
32.75 | 229
100.00 |
Statistics for Table of URM by Success_category |
Statistic | DF | Value | Prob |
Chi-Square | 1 | 9.0987 | 0.0026 |
Likelihood Ratio Chi-Square | 1 | 9.2409 | 0.0024 |
Continuity Adj. Chi-Square | 1 | 8.2681 | 0.0040 |
Mantel-Haenszel Chi-Square | 1 | 9.0589 | 0.0026 |
Phi Coefficient | | 0.1993 | |
Contingency Coefficient | | 0.1955 | |
Cramer's V | | 0.1993 | |
