Just worked through a statistical analysis where I thought chi-square would be the best tool. There were 66 pilots surveyed to get the information shown in the table below. After running the test I realized that one of the rules of chi square is that the inputs have to be independent, which these are not because each pilot was able to answer for both columns. Is this correct? Or is there a better test that I should be using to correctly show that there is or is not a relationship between napping on duty, and the likelihood of falling asleep?

Thanks!

Hypothesis:

Null hypothesis H0: variable Likelihood of Falling Asleep and variable Use of napping during duty are independent.

Alternative hypothesis HA: variable Likelihood of Falling Asleep and variable Use of napping during duty are not independent.

Variables: Ai: Likelihood of Falling Asleep. Bi: Use of napping during duty. Use Chi-square test for independence. State the hypothesis: Null hypothesis HO: variable Likelihood of Falling Asleep and variable Use of napping during duty are independent. Alternative hypothesis HA: variable Likelihood of Falling Asleep and variable Use of napping during duty are not independent. Contingency table results: Rows: Yes/No + Columns: None Likelihood of Falling Asleep Use of Naps on Duty Total Yes 49 39 88 No 17 27 44 Total 66 66 132 Chi-Square test: Statistic DF Value P-value Chi-square 13.4090909 0.0648 Because the P value is greater than .05, we accept the null hypothesis that the feeling of likelihood of falling asleep in the GCS is independent of the use of naps on while on duty