|
Hi all,
I'm using SPSS 16 for Mac for both school and professional work, and I'm constantly frustrated by the lousy quality of the graphic output. It's boxy, the colors are bland, no touches that have become standard in other apps (e.g. shadows) -- in short, definitely not something I would want to use for professional presentation. Plus, the Java platform of the app makes it somewhere between frustrating and impossible to manipulate the output. (I know the speed has improved in SPSS 17 for Mac, but I'd be curious whether the quality of the graphics has as well.) Can anyone recommend an alternative way to produce good-looking charts and graphs from SPSS data? I still want to be able to run statistics off the same dataset (i.e not have to transfer to another app like Excel). I do have a Bootcamp partition running WinXP under Parallels so I'd entertain Windows alternatives (but prefer to stay on the Mac side if possible). Any tips? Should I just buckle down and try to learn R? Should I buy Stata? Is there another option? Best, Max Freund P.S. In the meantime, I'll give bonus points to anyone who can tell me how to get SPSS to print or export an Output file correctly. I want charts to display with variable labels, which is how it's set in my application preferences and how they appear in the Output window on my screen. But when I export or print the Output, the graphs appear with the variable names instead of labels. WTF???? And is there a way to bypass the Java printing so I can use MacOS's native PDF conversion abilities? Is any of this fixed in SPSS 17???? Does SPSS have any plans to build native MacOS software in the future? -- |
|
I am not familiar with how graphs come out on a MAC. However, lot depends on what you mean by "professional". Is the goal to be eye-catching or to communicate something about the data? "Bland" (e.g., pastel) colors are frequently more communicative of content than highly contrasting colors that can induce optical illusions. I have recommended for years that SPSS look to things like Color Brewer to choose the default colors. http://statecancerprofiles.cancer.gov/colorscheme.html http://www.gcn.com/Articles/2002/09/11/NSF-studies-ways-to-show-complex-data.aspx Bill Cleveland has summarized research showing that in general people are best able to make accurate reading of the meaning of graphic layout in this order comparison of horizontal lines with a common start (best), horizontal lines with different starts, vertical lines . . . areas . . . volumes (worst). Shadows are deceptive because they cause the graph elements to seem to be volumes. Art Kendall Social Research Consultants Max Freund wrote: Hi all,
Art Kendall
Social Research Consultants |
|
From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of Art Kendall
[>>>Peck,
Jon] About the same as on a PC.
[>>>Peck,
Jon] We do use some schemes from Color Brewer in the SPSSINC MODIFY TABLES
customstylefunctions.py module.
[>>>Peck,
Jon] It is very easy to make misleading graphs. The SPSS graphics language,
GPL, focuses on chart structure rather than on fancy effects, but our Viz
engine is capable of delivering a lot of graphic effects. VizDesigner provides
more access to the style specifications than does plain GPL. Regards, Jon Peck
Hi all, I'm using SPSS 16 for Mac for both school and
professional work, and I'm constantly frustrated by the lousy quality of the
graphic output. It's boxy, the colors are bland, no touches that have become
standard in other apps (e.g. shadows) -- in short, definitely not something I
would want to use for professional presentation. Plus, the Java platform of the app makes it somewhere
between frustrating and impossible to manipulate the output. (I know the speed
has improved in SPSS 17 for Mac, but I'd be curious whether the quality of the
graphics has as well.) Can anyone recommend an alternative way to produce
good-looking charts and graphs from SPSS data? I still want to be able to run
statistics off the same dataset (i.e not have to transfer to another app like
Excel). I do have a Bootcamp partition running WinXP under Parallels so I'd
entertain Windows alternatives (but prefer to stay on the Mac side if
possible). Any tips? Should I just buckle down and try to learn
R? Should I buy Stata? Is there another option? Best, Max Freund |
| Free forum by Nabble | Edit this page |
