The "R-Word"

We fail to reject the null hypothesis at the 0.05 level of significance because the p-value is greater than alpha. Therefore, we have insufficient evidence to say that there is an association between whether or not a student has a personal relationship with someone with an intellectual disability and how frequent they use the r-word.

We reject the null hypothesis at the 0.10 level of significance because the p-value is less than alpha. Therefore, we have sufficient evidence to say that there is an association between whether or not a student uses the term as a medical term and how frequent they use the r-word.

     We first tested our data with a chi-square test. In this test, we tested all five frequency choices available. However, every test we performed did not meet the assumptions of having an expected cell count of greater than five. We then decided to group the frequency choices together in order to get valid results. We grouped everyday and a few times a week together and once every few months, never, and not sure together. We placed not sure in the latter group because we felt that if a person was not sure of how often they said the r-word, they most likely didn’t say it as often as the former group. Using this new grouping, we performed another chi-square test, this time with much more valid results. We did three of these chi-square tests. We tested the association between whether a student had a personal relationship with someone with a disability and the frequency they use the r-word, the association between whether the student uses the r-word as a medical term and the frequency the student uses the word, and the association between whether the responder was a teacher or a student and the frequency they used the word. We failed to reject the first null hypothesis, thus we found that there is no association between whether a person knows someone with a disability and how often they use the word. This was surprising to us as we assumed that students who personally knew someone with a disability would use the word less than those who didn’t. We were able to reject the second null hypothesis but only at the 0.10 level of significance. The association of use as a medical term and frequency of use in students did not surprise us. We originally thought that if a student uses the r-word as a medical term, they are less likely to use it casually. Our last test failed to meet the assumptions as its expected cell counts were less than 5. In fact, none of the other tests involving the data gathered from teachers was valid because their expected cell counts were less than 5. This was because of the small sample of teachers and the fact that all of the teachers filled out the survey in a similar way.

We were ultimately surprised by our results because we thought the strong association would be between whether a student personally knew someone with a disability and how often they use the r-word. However, our tests proved our original thinking wrong.