GAMLj-Logistic Mixed Effects Regression PSE/Threshold values
Posted: Wed May 18, 2022 2:20 pm
hello,
I'm trying to calculate the point of subjective equality (PSE)/threshold value (i.e., the x axis value for the p = 0.5) of the logistic mixed effects function fits. to do this, I take the inverse of the logistic regression function p(which should be equated to 0.5) = 1/1+e^-(ß0+ß1X), and isolate X which brings X = -ß0/ß1. however, when I calculate X with this formula and check the result from the graphs, the obtained values from the formula and the graph do not seem to match. I suspect that this discrepancy is a result of implementation differences between the standard logistic regression versus the logistic mixed effects regression.
any suggestions for such a calculation for logistic mixed effects?
thanks so much in advance, tutku
I'm trying to calculate the point of subjective equality (PSE)/threshold value (i.e., the x axis value for the p = 0.5) of the logistic mixed effects function fits. to do this, I take the inverse of the logistic regression function p(which should be equated to 0.5) = 1/1+e^-(ß0+ß1X), and isolate X which brings X = -ß0/ß1. however, when I calculate X with this formula and check the result from the graphs, the obtained values from the formula and the graph do not seem to match. I suspect that this discrepancy is a result of implementation differences between the standard logistic regression versus the logistic mixed effects regression.
any suggestions for such a calculation for logistic mixed effects?
thanks so much in advance, tutku