Reputation
14,394
Next tag badge:
87/100 score
19/20 answers
Badges
4 23 51
Newest
 Constituent
Impact
~263k people reached

May
24
reviewed Approve Issues with derivative of a copula density
May
23
revised Can I combine 10 variables into one variable before performing logistic regression on 18 total variables?
edited tags
May
23
comment Can I combine 10 variables into one variable before performing logistic regression on 18 total variables?
Welcome to Cross Validated! Data reduction is clearly a sensible approach to model fitting here - you've only 43 cases! I think you need to add a little more context about what you're trying to achieve & clarify what you're asking - how to combine them? how to decide whether to? what are alternatives? & so on.
May
23
answered Using KS test to evaluate differences
May
23
reviewed Approve How to test predictive power of ARIMA model
May
23
comment KS, RMSE tests in R
If you've doubts about particular things, or parts of the software manuals you don't understand, focus on those. Give some context: whatever "streamflow" data is, there could well be some feature of how it's been collected that would affect statistical inference. Also please expand abbreviations on first use; & if you include code you might as well make it runnable.
May
23
comment KS, RMSE tests in R
Welcome to Cross Validated! Could you please expand a little on what you're trying to do, & make it clear what the statistical question is? The Kolmogorov-Smirnov part is straightforward; but you haven't said what you want to calculate the root means square error of, & refer to a simulation that doesn't appear in the code - rather you seem to be calculating the theoretical quantiles (of your generalized extreme value [?] distribution) corresponding to the empirical distribution function.
May
22
revised Reduce subsetting of the dataset?
deleted 1 character in body
May
22
reviewed Close Comparing logistic regression models with different predictors
May
22
revised Aside from Durbin-Watson, what hypothesis tests can produce inconclusive results?
deleted 19 characters in body
May
21
comment Aside from Durbin-Watson, what hypothesis tests can produce inconclusive results?
The first question's a bit vague ("What other statistical tests [...]?"), but I don't think you could clarify it without answering the second ("Is there a general explanation [...]?") yourself - overall I think it's all right as it stands.
May
21
revised Aside from Durbin-Watson, what hypothesis tests can produce inconclusive results?
added 344 characters in body
May
21
revised Aside from Durbin-Watson, what hypothesis tests can produce inconclusive results?
added 182 characters in body
May
21
revised Aside from Durbin-Watson, what hypothesis tests can produce inconclusive results?
added 492 characters in body
May
21
answered Aside from Durbin-Watson, what hypothesis tests can produce inconclusive results?
May
21
comment Restoration of gender balance after wars - what principle underlies this?
@wildetudor: Hard to see how effects on birth sex ratio of the magnitude discussed in that paper - half a percentage point or so - could make much contribution to restoring a "dramatic" imbalance in population sex ratio within five years. To continue with the Germany example: with a population of about 70 million, there should have been about 5 million births in the five post-war years. Allowing for a large shift in birth sex ratio from 1.04 to 1.11 gives only 70 thousand additional male births, which wouldn't go far towards plugging any significant gap.
May
20
comment Restoration of gender balance after wars - what principle underlies this?
Mightn't the fast restoration of sex ratio balance in post-war Germany (if there was such) be largely due to POWs returning home?
May
20
reviewed Close Finding AIC in a Holt-Winters Forecast
May
20
reviewed Close What does Boston housing data be projected?
May
20
reviewed Close How do I obtain the BOOTSTRP slope and intercept