Statistics Help @ Talk Stats Forum - Regression Analysis
http://www.talkstats.com/
Linear regression, linear models, nonlinear regressionenWed, 25 May 2016 12:45:46 GMTvBulletin60http://www.talkstats.com/images/misc/rss.pngStatistics Help @ Talk Stats Forum - Regression Analysis
http://www.talkstats.com/
how to argue using wald-test/LR-test for logistic Regression
http://www.talkstats.com/showthread.php/65144-how-to-argue-using-wald-test-LR-test-for-logistic-Regression?goto=newpost
Wed, 25 May 2016 12:38:50 GMTHey there,

I'm writing my thesis and I did a logistic regression in a data set. At first, it didn't come into my mind that I should argue why I'm using the wald-test or lr-test to test the significance of the coefficients, since in all books I read, it said use these test to do it. Unfortunately I have to argue why I'm not using for example the f-test or the t-test. I know that I probably can't use them because of the dichotomous outcome variable I have but I can't find any proof for that.

Can someone help me with that maybe?
]]>Regression Analysisabcedhttp://www.talkstats.com/showthread.php/65144-how-to-argue-using-wald-test-LR-test-for-logistic-RegressionOLS regression with interdependence
http://www.talkstats.com/showthread.php/65141-OLS-regression-with-interdependence?goto=newpost
Wed, 25 May 2016 11:15:11 GMTHello everyone,
I have been writing my thesis and decided, together with my supervisor, to use OLS even though my actors are interdependent...Hello everyone,

I have been writing my thesis and decided, together with my supervisor, to use OLS even though my actors are interdependent (doing research on dyad level).

Could you explain what happens to my results by doing this compared to using QAP analysis for example?
]]>Regression AnalysisTiUResearchhttp://www.talkstats.com/showthread.php/65141-OLS-regression-with-interdependenceOptimal cut-off from ROC
http://www.talkstats.com/showthread.php/65135-Optimal-cut-off-from-ROC?goto=newpost
Wed, 25 May 2016 08:34:24 GMTHello,

as you know, ROC curve is not depend of the cut-off. But how can I get an "optimal" cut-off from ROC curve?

Example: I have a database which consist of 1 400 observations (clients) and 700 of them are "failed clients". When I conduct/perform a logistic regression I get coefficients/statistics/plots and so on. But when I want to depict a classification table which cut-off should I use? Which cut-off is optimal for decide which client is bad/good?

Is it possible to find out this information from ROC curve? (please find attached )