Hello
First, I am using a survey data for logistic regression analysis. I noticed that my predictor variable as well as some con-founders had some outliers and missing values. However, the missing values are less than 10% for all the predictors but my concern is if the missing values and outliers had a significant effect on the overall result of the model. I have run the logistic regression after adjusting for these con-founders and the results were non-significant.
Second, after fitting the model using the code - svy: logistic outcome x y z a b - I used the code estat gof to check the fitness of my model and the result was
F(9,41) = 0.55
Prob > F = 0.8326
Please can someone interpret the meaning of this. when are we supposed to say that the model fit very well? at what p-value? also, am expected to use the command with "svy" or just the way I have done it?
First, I am using a survey data for logistic regression analysis. I noticed that my predictor variable as well as some con-founders had some outliers and missing values. However, the missing values are less than 10% for all the predictors but my concern is if the missing values and outliers had a significant effect on the overall result of the model. I have run the logistic regression after adjusting for these con-founders and the results were non-significant.
Second, after fitting the model using the code - svy: logistic outcome x y z a b - I used the code estat gof to check the fitness of my model and the result was
F(9,41) = 0.55
Prob > F = 0.8326
Please can someone interpret the meaning of this. when are we supposed to say that the model fit very well? at what p-value? also, am expected to use the command with "svy" or just the way I have done it?
Comment