Reputation
15,510
Next tag badge:
84/100 score
23/20 answers
Badges
5 26 57
Impact
~315k people reached

6h
comment Can I use Cox Regression here?
(+1) for the first paragraph, but if the linearity assumption's dubious then a polynomial or spline basis will generally be a better way to allow for a curvilinear relationship - see What is the benefit of breaking up a continuous predictor variable?. (There's plenty about splines in RMS.)
6h
comment The notation of hypothesis testing – are many authors “wrong” or am I missing something?
Both your examples are conditional on the truth of $H_0$ so $\Pr(H_0)$ doesn't come into it. You could write $\Pr_{H_0}(\mathrm{result})$ instead.
1d
comment Model selection, issues of judgement
See Should covariates that are not statistically significant be 'kept in' when creating a model?
2d
comment modeling sales based on different quantitative and qualitative variables
Welcome to Cross Validated!. It's hard to see exactly what your question is. See our help page & please try to focus on something specific that you need help with.
2d
comment Question about number of observation in Generalized ESD
The number of observations after you've removed one of them is denoted by $n-1$. The number of observations is therefore $n$.
2d
revised Different output for cox regression in R vs SPSS
fixed typo
2d
awarded  generalized-linear-model
2d
comment G-test vs Pearson's chi-squared test
See Deviance vs Pearson goodness-of-fit.
2d
revised G-test vs Pearson's chi-squared test
fixed typos
2d
comment how to calculate R-squared in glm?
Could you explain what you mean in a little more detail? (And presumably that should be "confusion matrix").
2d
comment how to calculate R-squared in glm?
See e.g. Compare classifiers based on AUROC or accuracy? & Measuring accuracy of a logistic regression-based model.
2d
comment how to calculate R-squared in glm?
Despite their awkwardness the pseudo-R^2s are at least likelihood-based proper scoring rules. When your logistic regression model is not being developed for a specific classification task there's no sense in assessing its performance as a classifier using an arbitrary cut-off - thus ignoring that an predicted probability of "success" slightly below the cut-off is much less discrepant with an observed "success" than one far below. And AUC measures pure discrimination without any regard to calibration.
2d
comment how to calculate R-squared in glm?
See Which pseudo-R2 measure is the one to report for logistic regression (Cox & Snell or Nagelkerke)?.
2d
revised Goodness-of-fit test in Logistic regression; which 'fit' do we want to test?
added 78 characters in body
2d
comment Are the $\text{P}$ and $\Pr$ operators equivalent?
See also http://math.stackexchange.com/q/875310/59351
2d
reviewed Approve Bayesian data analysis with R
2d
reviewed Close Error in gbm.step: In cor(y_i, u_i) : the standard deviation is zero
2d
reviewed Close R ggplot2: overlaying multiple geom_ribbon objects in a single plot using a nested loop
2d
comment Is rejecting the hypothesis using p-value equivalent to hypothesis not belonging to the confidence interval?
@JonasBerge: That seems curious. Were you using the same test statistic for confidence intervals & p-values e.g. Wald's?
2d
revised Is rejecting the hypothesis using p-value equivalent to hypothesis not belonging to the confidence interval?
fixed typo