Let me begin by stating that statistics are unfortunately no strength of mine.
I have conducted a questionnaire assessing loss of smell/taste in healthcare workers during the COVID-19 pandemic for my MSc thesis. I have surveyed 254 HCWs of which 44 reported loss of their smell and taste. The first question asked whether the participant had experienced a loss of smell or taste within the last 6 months (yes/no). The next question enquired whether the participant had experienced a burning or tingling sensation within the 6 months (yes/no).
I would subsequently like to compare the results of these two questions in order to establish if participants who reported a loss of smell and taste were statistically more likely to experience a burning and tingling sensation than those who did not report a loss of smell or taste.
I have already run a Chi-square test using spss however my assumptions have been violated so I want to ensure I am using the correct statistical test.
Any suggestions/guidance would be greatly appreciated. Thank you in advance.
Hello David,
I'm not sure why you conclude that a chi-square test of independence won't work for your analysis (was it a matter of one or more cells having too-low expected frequencies?). In addition to the chi-square, here are some other possible paths:
1. Fisher exact test.
2. Odds ratio (with CI), comparing odds of burning sensation reported for respondents with loss of smell/taste vs. those without loss of smell/taste.
3. Logistic regression, with IV of smell/taste status and DV of burning/tingling status (which should yield same result as #2).
Good luck with your work.
David Gillespie , since you are trying to evaluate the relationship between the two variables, I would suggest to calculate Phi, a measure of association between two binary variables:
https://en.wikipedia.org/wiki/Phi_coefficient
As all are qualitative traits so proceed with non-Parametric tests as very much suggested by David Morse.
Good luck
- Badges
- Science topic
- Similar topics
- Medicine
- Physiology
- Biosignals
- Biomedical Signal Processing
- Statistical Analysis