I read a published paper with multiple graphs as attached, all of them have a caption that says something in the lines of "Predicted estimates (± 95% confidence interval) for xxx. The dashed line indicates the reference level: ‘
Hello Zaara,
If a graph gives estimated point values, with 95% CI, then you are, in fact, getting information about the expected amount of error associated with the estimated/predicted scores. However, you're getting approximately 1.96*SE as the bars, rather than 1.0 * SE.
Without knowing more about the specific study, I've no way of knowing what the import of the 10% reference line in the graph is. Is that some bound that author/s declared as salient in some way?
Good luck with your work.
The graph is a result of glmm from the study. I don't want to give out more details just for the sake of not "outing" anyone. My concern is, that the "dot" should have error bars if it is a predicted value, however, if it is a reference point, it should be at zero, because it is result of a glmm from Jtools in R, which gives log odds on the Y axis. So it can only be one of the two things according to my understanding, which I am trying to get some clarification on.
- Badges
- Science topic
- Similar topics
- Geoscience
- Asia
- India