Re: MW & t-test

From: Eric Bohlman <ebohlman@xxxxxxxxxx>
Date: 11 Jul 2005 23:42:10 GMT

dag.tidemalm@xxxxxx (Dag Tidemalm) wrote in
news:1eee739e.0507110129.54a57d16@xxxxxxxxxxxxxxxxxx:

> How would you interpret these results for same comparison?
> Mann-Whitney p=0.04; t-test p=0.18
> The two groups compared have size n=60 and n=73, respectively.
> Significant or not?

Plot your data. Do you see anything out of the ordinary, like either group
being substantially bimodal? Do histograms of both groups have the same
overall shape (MW is sensitive to shape differences)?

.

Follow-Ups:
- Re: MW & t-test
  - From: Dag Tidemalm

References:
- MW & t-test
  - From: Dag Tidemalm

Prev by Date: Re: Jittering does redeem stepwise regression
Next by Date: Re: Comparison of Logistic regression and Cox regression
Previous by thread: MW & t-test
Next by thread: Re: MW & t-test
Index(es):
- Date
- Thread

Relevant Pages

Re: Unintentionally Sympathetic Villains
... PLOT, and that's _The Magic Flute._ Yeah, I guess the broken ... essentially a child custody dispute and he's badmouthing the ex the ... that Sarastro is Pamina's father. ... you can interpret it that way if you like. ...
(rec.arts.sf.written)
logistic regression: plot of deviance residuals against estimated prob
... I am working on a logistic regression. ... I am asked to plot the deviance residuals against the estimated probabilities and interpret that plot. ...
(sci.stat.math)