10,740 reputation
31944
bio website
location England
age
visits member for 1 year, 9 months
seen 15 hours ago

15h
comment Statistical publishing - english majors?
@EngrStudent: From the Academia SE help: "This site is for academics of all levels [...] as well as anyone in or interested in research-related or research-adjacent fields". There are proofreading & copy-editing tags. Note co-authors would generally be expected to have made a contribution to the content of the paper, not just its form.
17h
comment Statistical publishing - english majors?
Co-author or proof-reader!?
1d
revised How useful is Minitab in the real world?
added 298 characters in body
1d
reviewed Leave Open How useful is Minitab in the real world?
1d
answered How useful is Minitab in the real world?
2d
reviewed Close Error when using msmFit in R
2d
reviewed Close Comparing regression models, same and different response variables
Sep
12
comment What statistic to use
You're welcome. Scheffe's test doesn't rely on a previous ANOVA F-test to control the experiment-wise error rate either. Fisher's protected LSD does rely on a previous ANOVA F-test, (but doesn't truly control the experiment-wise error rate).
Sep
12
comment What statistic to use
Note that Tukey's test controls the experiment-wise error rate across all 3 comparisons, & that's not conditional on the result of the ANOVA F-test. See "Do we need a global test before post-hoc tests?".
Sep
10
comment Bias-variance tradeoff in the paired t-test
Comparing the model with subjects to the model without, the power of the t-test to detect departures from the null hypothesis that the coefficient for treatment is zero may be higher as well as the mean square error of predicted out-of-sample responses.
Sep
8
comment Bias-variance tradeoff in the paired t-test
@James: No rigorous ceteris paribus condition - after all you could transform the predictors to be orthogonal to each other - more to do with how an experiment typically covers the design space compared to observational data. But in general I'm agreeing with you: a model including 'subject' might over-fit compared to one not including it. Still, you couldn't say that's a problem for the test or for estimation of the treatment effect, only for prediction of new responses.
Sep
3
comment Interpretation of Zero-One inflated Beta Regression with R (GAMLSS)
Well, $\nu$ & $\tau$ are odds, not probabilities, modelled through a log link; $\mu$ they rather sloppily call a location parameter - it's the mean of the beta distribution left when you remove 0's & 1's - & it's modelled through a logit link. Given the complexity of the model you've chosen to fit, you can't necessarily expect the relationship between predictors & response to be easily summarized by saying $x$ increases $y$ - a predictor could lower $\mu$ while increasing $\tau$. Why not just plot it? [BTW, replace the incorrect p.d.f. with the correct one if you don't wish to confuse people.]
Sep
3
comment How to determine whether a set of data are qualitative or quantitative?
... quantitative variables, which, together with knowledge of the costs of different mis-classifications, you could use to formulate a decision rule. E.g. it might seem reasonable to suppose that a variable with 5 unique values is a Likert item, or one that tails off quite slowly toward higher values is a count & not a category code.
Sep
3
comment How to determine whether a set of data are qualitative or quantitative?
Yes, it's hard to imagine what you'd be wanting to do with data you knew so little about. Some sort of automated descriptive/graphical analysis? - even then it seems rather barmy to have to guess an appropriate scale of measure rather than stipulating it with meta-data. If for some reason you do have to guess, knowing something about the context - behavioural/attitudinal surveys, order lines from a transactional database, or whatever - might justify making some assumptions about typical differences in no. unique values, & the distribution across those values, between qualitative & ...
Sep
3
comment How to determine whether a set of data are qualitative or quantitative?
(1) That sounds like a bad idea, if it's more than a provisional identification to be reviewed by someone with domain knowledge. (2) Any such decision rule could perform very badly on average depending on the kind of data-sets to which it's commonly applied - you could sample some & look for patterns.
Sep
3
revised How to determine whether a set of data are qualitative or quantitative?
added 59 characters in body
Sep
3
answered How to determine whether a set of data are qualitative or quantitative?
Sep
2
reviewed Leave Open Transform long tail data set to bell curve
Sep
2
reviewed Close Power/sample size calculation for one group prospective study
Sep
2
reviewed Close Clustering using different distance measures