Re: Problems with display

[Gaj Vidmar]

Just don’t do a histogram with a “normalising” curve! Just don’t! Ever!

Trust a highly experienced and qualified professional statistician, or, if 
you don’t accept arguement by authority, read any (text)book BY A QUALIFIED 
STATISTICIAN (or better yet, a giant; see hint) on data visualisation, 
statistical graphics, Exploratory Data Analysis (major hint here), ...

And, BTW, it’s NOT a “NORMALISING” curve! It’s just a superimposed best-fit 
normal (Gauss) distribution to your observed data (by the criterion of least 
squares; by maximum likelihood the SD would have been negligibly different, 
namely with N instead of N-1 in denominator). This is all introductory 
stuff – even high school or pre-college or whatever one would call it in the 
US ...

Again, use statistics right, or don’t use it/them at all, rather than do it 
wrongly.

Best regards,
Gaj Vidmar
(from somewhere in Europe, someone with nearly 100 SCI/SSCI co-authored 
paper, with degrees in social science and stats and more, but still fooling 
myself students can be made enlightened)

P.S. Occasionally, very rarely, the feature in question is admissible or 
even useful, but before you get to that (if ever), the good people behind 
PSPP will have long fixed it.

Or perhaps aband it upon my advice (or I should rather say rearrange such 
things), if the whole graphical output is redesigned as it should have been 
long ago (or even in the first place, but what’s done’s done), making it 
“grammar based” (the good people will know what I’m talking about) in a 
feasible and managable way actually helpful to the user. That would be a 
“giant leap” forward for PSPP, rather than trying to mimic SPSS in vain and 
wasting the highly limited human resources so that even the functionality of 
SPSS V6 seems far from sight, let alone of the present V20 (some 20 years 
ahaed in all ways). But that’s a different topic, and these seemingly harsh 
words are actually meant as motivation and high praise for the good people, 
i.e., the PSPP developers (who will hopefully realise that).

---
"Bill Chadkirk" <address@hidden> je napisal v sporočilo 
news:address@hidden ...
Hi

As part of an MPhil study on the sociology of religion I’m using PSPPIRE to 
analyse data on attendance at religious meetings, on a windows 7 laptop. 
When I try do display a histogram with a normalising curve the x-axis labels 
are compacted and overwritten making them impossible to read. Is there a 
solution?

Bill Chadkirk


_______________________________________________
Pspp-users mailing list
address@hidden
https://lists.gnu.org/mailman/listinfo/pspp-users