The Diet Study was our very first project and grew out of my own efforts to better my symptoms through changes in diet. We use a popular diet tracking app and symptom ratings as frequently as possible to look for patterns. We hope to expand this study in the near future to test the effect of specific diet types (e.g. the paleo-diet or ketogenic-diet). Currently we track what we 'normally' eat. More info about the study.
Diet Study Results
Any relationship or pattern could exist simply due to chance. We use statistics to quantify how likely a relationship is to be true. Using the values and table below you can roughly determine whether YOUR data is statistically significant, meaning, the relationship between the two variables is likely to be true. I have over 6 weeks of data in the analysis. From the values below I know that an R-value of +/- 0.30 would indicate a likely to be true pattern. In the Quick View Table my ID number is shown. If you follow the 2nd row (Pain) over to the Fat column you should see (0.33). Since 0.33 is greater than 0.30 we would say the effect is significant (less than -0.3 would also be).
How to interpret the effect is a different matter. You can use this Quick View Approach to scan your results and then go to the links below to see the actual data plotted. That should help with the interpretation. Note that the value you see when hovering over the "Best Fit Line" on a graph is called the R-squared value. It is not the same as the values in the Quick View Table. It is simply the value in the table squared.
Average Daily Symptom Severity
1 week - 0.75 or greater is significant 2 weeks - 0.53 or greater is significant 3 weeks - 0.43 or greater is significant
4 weeks - 0.37 or greater is significant 5 weeks - 0.33 or greater is significant 6 weeks - 0.30 or greater is significant