Caution! This message was sent from outside the University of Manitoba.
Hi all
I have a project where I compare results of statistical analysis of free to use, easy to use (menu driven) statistical analysis programs.
Free To Use Statistical Software: Comparing Statistical Analyses<https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4105959>
I did comparison of a little results from EpiData, comparing frequency and means to results from SAS.
I'm happy to say EpiData results and SAS results are the same.
I usually also compare results from correlation and regression. The help file in the downloaded program says regression is, at present, only in Analysis classic? The commands.htm file on the epidata website also lists correlation. I'm not clear if correlation is available in the all in one install? I can't seem to get a correlate command to run.
correlate :=PctBelowPov =BirthRate =Services
results in
ERROR: Identifier "correlate" not found!
Thanks
Gene
Caution! This message was sent from outside the University of Manitoba.
Hi all
I hope this is something simple.
I'm reading in this data set, the one I mentioned before.
ShortDataSet.csv<https://drive.google.com/file/d/11WIET3s4eMHsB6JQ6hOQBx0SKquGRKKw/view?usp=…>
I can get a frequency for AreaRegion, using both the drop down menu system and by typing in "freq AreaRegion" (without the quotes)
and I can get a means in both the drop down menu system and by typing in "means population
I can't seem to get means by population, in the line at the bottom of the analysis system where we can type in commands.
I went to the list of commands
http://www.epidata.dk/commands.htm
where it says this:
means variable1 [/by=variable2] [/T] [/q]
basic descriptive statistics for numeric variable1, optionally stratified by variable2
Options:
/by=variable2 : Stratify by this variable
I typed in " means population /by=AreaRegion" (without the quotes)
and got this
.means population /by AreaRegion;
ERROR: Line 1: Syntax error at pos 18: /
Expected: "-", "!", "(", ")", ":=", ";", "@", "[", "else", "Identifier", "in", "to"
I also tried a couple of variations, like "means population [/by=AreaRegion]"
and means population (/by=AreaRegion)
and got similar error lines.
Any help with this?
By the way, on the commands.htm web page above, in the first column, it looks like you can click on the command and it's supposed to go to another web page, for example,
means<http://www.epidata.dk/means.htm>
is supposed to lead to http://www.epidata.dk/means.htm
but that results in not found.
And again, I greatly appreciate all the work that folks put into EpiData and all the accompanying web pages, help pages, documentation, etc. It's a lot of work and your work is appreciated.
Thanks
Gene
Caution! This message was sent from outside the University of Manitoba.
Hi
I think I mentioned a couple of places on the epi data website where the email list website needs to be updated.
Also, when you open epi data analysis, and click on help and about, it says the epi data email list is here
http://lists.umanitoba.ca/mailman/listinfo/epidata-list
but that clicks onto 'page not found'.
Thanks
Gene
Gene Shackman
Albany NY 12203
Applied Sociologist
Caution! This message was sent from outside the University of Manitoba.
Hi all
First, thanks to everyone involved in EpiData. It's a lot of work and I appreciate it. I list EpiData in my list of free to use stat software
https://sites.google.com/site/gsocialchange/statsoft
I have a data set that's probably difficult.
ShortDataSet.csv<https://drive.google.com/file/d/11WIET3s4eMHsB6JQ6hOQBx0SKquGRKKw/view?usp=…>
I can do some analysis, such as frequency, and a scatter graph, but I'm having problems with means.
Using the menu drop downs, I tried to get a means for population and got this outcome
.means population !vla !l;
ERROR: Unknown option/variable: vla
Actually, I get the same error if I try to get means for pctbelowpov, and gdppercap too.
I -can- use population in a scatter plot, for example, population by gdppercap, or population by pctbelowpov.
I can also use "describe" on a set of variables, for example, pctbelowpov to population.
Just can't get means to work.
Any help is greatly appreciated.
Thanks
Gene
Gene Shackman
Albany NY
Applied Sociologist
https://sites.google.com/site/gsocialchange/global-social-change-research-p…
