Hi Dean,
I just noticed this email, and see that you didn't get a reply. So,
if it's not too late, here goes. I'm guessing, because I haven't see
the article.
There are two kinds of confidence intervals in regression. Predicted
score CIs and regression parameter CIs.
First, let's imagine a (really simple) regression equation. It's a
straight line.
y = 1 + 0x
Let's imagine that the CIs are 1 in each direction, so the intercept
(constant) has CIs from 0 to 2, and the slope has CIs from -1 to +1.
So imagine you have the flat line which is the slope. We want to get
the CIs. So move it up one unit, and then make it slope from -1 to
+1, and imagine (or draw) the shape that is the highest point of the
line, all the way along the X axis. That's the upper CI for the
slope. It's curved, isn't it?
Now imagine we want to get the predicted score for each person on Y.
We know that each person's predicted score is 1 (because the slope is
0, it doesn't matter what they get on X). But there will be a
confidence interval around that. Say it's 0.5. Whatever a person's
score on X is, it their predicted score on Y will have CIs -0.5, 0.5.
Two straight lines.
Shameless plug (because I notice you're at Essex), I'll be teaching
the Summer School Course called "Applying Regression" this July in
Essex.
Jeremy
On 8 September 2011 05:26, Wybrow, Dean P <[log in to unmask]> wrote:
> Hello fellow postgrads,
>
>
>
> I have been searching for the answer for a problem with confidence intervals
> for a long time now, so I hope someone can help!
>
>
>
> The two papers below performed the same analysis, regressing irregular word
> reading onto reading nonword reading for controls, and plotting these
> regressions with 90% confidence intervals, before superimposing dyslexic
> participants to find outliers. However, one paper has straight line
> intervals, the other curved intervals. Why?
>
>
>
> I have scoured both textbooks and online resources, and the only
> descriptions of confidence intervals I have found are for confidence
> intervals of the regression line, which estimate an area in which there is a
> given probability that the regression line will fall. This is not what the
> studies below used, however. They used what may, I think, be termed
> ‘prediction intervals’ and describe an area in which a given percentage of
> the data will fall. Can anyone help with these distinctions, and/or provide
> a calculation for this type of interval? The curious thing is that SPSS/PASW
> will plot these confidence intervals for me, but offers no explanation of
> where they come from, nor can I edit the provided plot.
>
>
>
> Any help, no matter how small, will be greatly appreciated!
>
>
>
> References:
>
> Manis, F. R., Seidenberg, M. S., Doi, L. M., McBride-Chang, C. & Petersen,
> A. (1996). On the bases of two subtypes of developmental dyslexia.
> Cognition: 58, 157-195.
>
> Stanovich, K.E., Siegel, L.S. & Gottardo, A. (1997). Converging evidence for
> phonological and surface subtypes of dyslexia. Journal of Educational
> Psychology: 89, 114-127.
>
>
>
>
>
> Regards,
>
>
>
> Mr. Dean Wybrow
>
> Dean Wybrow BSc, MRes
> Department of Psychology
> University of Essex
> Colchester CO4 3SQ
>
> Tel: +44 (0)1206 - 872575
>
> www.essex.ac.uk/psychology/department/people/wybrow.html
>
> essex.academia.edu/DeanWybrow/About
>
>
>
> 'In all science, error precedes the truth, and it is better it should go
> first than last.' -Hugh Walpole
>
>
>
>
|