|
|
One-sided confidence intervals |
|
Message/Author |
|
|
Dear Dr. Muthén, I would like to ask if there is any possibility to obtain one-sided 95% confidence intervals for analyses using the Mplus 5.2 software. Specifically, we hope to obtain one-sided CIs for coefficients from straightforward linear regression and ordinal logistic regression, as we are testing one-tailed hypotheses. One way we thought of might be to take the range of the two-sided 95% confidence interval from the Mplus output, then determine the lower (or, depending our hypothesis, the upper) bound of a two-sided 90% CI, and then add (or substract) the range of the two-sided 95% to (or from) this bound. However, if I understand correctly, Mplus only gives two-sided 95% and 99% CIs for regression parameters, not 90% CIs. Do you perhaps have any other suggestions on how to obtain one-sided CIs for regression parameters? Thank you very much for your help. |
|
|
Mplus does not provide one-sided confidence intervals. You would need to compute those yourself. |
|
Back to top |
|
|