- #1
jamesmartinn
- 1
- 0
Hello all, I've performed a simple logistic regression and could use some help in interpretations/minor calculations.
Five different doses of insecticide were applied under standardized conditions to samples of an insect species. The data were:
Dose: ---2.6---3.8---5.1---7.7---10.2---
Dead: ----7----16---20----48----54----
Total: ---60----60---59----57----60---
So I was asked to fit a logistic regression model that says the logit of chance of death increases linearly with the natural logarithm of dose (ml/G). So, I did a ln transformation of the dose variable, and then proceeded with the logistic regression.
Some important output from the analysis:
Beta_hat = 3.3364
95% Likelihood Ratio Confidence Interval for Beta = (2.6478, 4.0989)This is pretty much a multi-step type question... I've answered questions a) through f), but I am getting really confused on questions g) and h)
Here are the questions, verbatim.
g) Given an approximate 95% Likelihood Ratio Confidence Interval for B. Translate this interval into an interval for the effect on the odds of death of increasing the dose by 50% (ie., multiplying the dose by 1.5) and interpret. Hint: First translate the multiplying dose factor to the natural log scale.
For this question, I've been going with this interpretation.
e^Beta_hat = e^3.3364 = 28.12. Thus, for a one-unit increase in the log dose, there is a 28.12 multiplicative effect on the odds in favor of insect death. I'm just having trouble proceeding with the next parts...logarithms and all.
I need help converting the interval so I can say something like this. For a 50% increase in Dose (Not Log dose), there is a ___ multiplicative effect on the odds in favor of insect death, CI (___, ____)
h) Setup an approximate 95% CI for the rate of change of the chance of death per unit increase in log concentration at the median effective level.
Any help would be great. Like I said, its just these last two questions that are really confusing me.
Cheers
Five different doses of insecticide were applied under standardized conditions to samples of an insect species. The data were:
Dose: ---2.6---3.8---5.1---7.7---10.2---
Dead: ----7----16---20----48----54----
Total: ---60----60---59----57----60---
So I was asked to fit a logistic regression model that says the logit of chance of death increases linearly with the natural logarithm of dose (ml/G). So, I did a ln transformation of the dose variable, and then proceeded with the logistic regression.
Some important output from the analysis:
Beta_hat = 3.3364
95% Likelihood Ratio Confidence Interval for Beta = (2.6478, 4.0989)This is pretty much a multi-step type question... I've answered questions a) through f), but I am getting really confused on questions g) and h)
Here are the questions, verbatim.
g) Given an approximate 95% Likelihood Ratio Confidence Interval for B. Translate this interval into an interval for the effect on the odds of death of increasing the dose by 50% (ie., multiplying the dose by 1.5) and interpret. Hint: First translate the multiplying dose factor to the natural log scale.
For this question, I've been going with this interpretation.
e^Beta_hat = e^3.3364 = 28.12. Thus, for a one-unit increase in the log dose, there is a 28.12 multiplicative effect on the odds in favor of insect death. I'm just having trouble proceeding with the next parts...logarithms and all.
I need help converting the interval so I can say something like this. For a 50% increase in Dose (Not Log dose), there is a ___ multiplicative effect on the odds in favor of insect death, CI (___, ____)
h) Setup an approximate 95% CI for the rate of change of the chance of death per unit increase in log concentration at the median effective level.
Any help would be great. Like I said, its just these last two questions that are really confusing me.
Cheers
Last edited: