Confidence intervals for phi (interaction) parameter

[sixtystrat]

I am doing a 2-species model and I am coming up with a phi value of 1.0254 with a SE of 0.0560. However, the 95% CIs that PRESENCE gives me is 1.0194 - 1.2629, which doesn't make sense to me. The standard error would lead me to believe there is no interaction whereas the CI would indicate a positive interaction. What gives?

[jhines]

The confidence interval for phi is computed using the beta estimates for phi, so it's not the simple (symmetrical) phi +/- 1.96 * std.err. To compute the 95% confidence interval, take the beta estimate for phi, add/subtract 1.96 times it's std.err., then transform those values (using exp function).

Jim

[sixtystrat]

Thanks Jim. The CIs make sense doing it that way.
Joe