Confidence interval for expected value of variable in linear model R

Question

I have uploaded a dataset and fit a linear model to it where yis the predictor, x is the response, and d is the dataset.

Unclear. Confidence interval on the mean value of Y across the data set? What does "predictive interval for just y" mean? By definition, the expected mean value of Y will be occur at E(x), so you might try: `predict(linmod, newdata=data.frame(x=mean(d$x)), ...)` — Ben Bolker, Dec 01 '17 at 00:36
I apologize for the confusion. I am trying to find the confidence interval for the expected value of y and the predictive interval for y (not the predictive interval for the expected value of y). Thank you for your help. — uytt, Dec 01 '17 at 00:39

score 1 · Answer 1 · answered Dec 01 '17 at 00:57

You can generate a confidence interval for the point estimate like this:

CI <- confint(linmod, 'x', level=0.95)

You can generate a prediction interval for each of the observed points in d like this:

PI<-predict(linmod, interval="prediction", newdata = list(d$x), level=0.95)

1 Answers1