Order of how yes no settings appear in tables
Hi Jamie
A little while ago when you answered my question about changing the Boolean fields to 0-1, you mentioned there was a way in both EpiData Analysis and EpiInfo to reverse the order that two by two tables were shown, that would help in a situation where 0 was no and yes was 1 (so the 1 or yes answers appeared first in the table). I have looked around Epi Info but couldn't find a setting like this. Could you let me know how to do it? And also how to do it in Epidata analysis? I
Many thanks
Christine
Christine Roseveare Analyst Regional Public Health 570 9194 027 495 9671
Christine Roseveare 20/09/2010 5:21 p.m. >>>
Hi Jamie
Could you please explain why you changed the booleans to 0-1 before loading it into manager? Sorry if this is a silly question. I am asking because someone recently gave me an outbreak file where they had used 0 and 1 (created in Excel 0 = no and 1 = yes). The problem was that when I came to analyse it (in this case using Epi Info) it meant that the two by two tables were in the wrong order. Because no was 0 it came out first in the table. I have always used the boolean field in epidata for our outbreak questionnaires and it has worked very well. They come out in the right order in both Epidata Analysis and Epin Info, and also make intuitive sense to the data entry person.
thank you
Christine
Christine Roseveare Analyst Regional Public Health 570 9194 027 495 9671
epidata-list@lists.umanitoba.ca 13/09/2010 2:34 p.m. >>>
I successfully loaded a moderately-sized .rec file (changing all booleans to 0-1 before loading into Manager--it should prompt to do that for us). Entry handled the existing and new data as expected.
I would like to have the autojump to next field when all digits are filled. I assume that this is coming.
I successfully loaded a fictitious file with just under 500,000 records and 12 fields (41.8 MB .epx from an 8 MB .rec). Given the capacity of removable memory today, size really isn't a problem. I cannot imagine ever having a real file that large. On my old Windows laptop, this file loaded/saved in 20-30 seconds, but GO TO any record in the file was instantaneous.
Jamie
In EpiData, the default for tables is to put the higher values first. So if you code yes=1 and no=0 then the top left corner of a 2x2 table will be for yes,yes. If you code yes=1 and no=2, then add /sw to the tables command. This switches the order:
tab case exposure /sw
There is lots of help for the tab (tables) command. Just type tab and then press <F1>
In Epi Info for DOS, you have to recode variables to make the "No" value higher than the "Yes" value. e.g. for yes=1 and no=0 case = 2-case exposure = 2-exposure tables exposure case
For Odds Ratios, there is no need to recode. For Relative Risk, you must have the fields in the right order and have yes, yes in the top left. Notice that EpiData has better syntax in that the outcome field is first, followed by one or more exposures.
I haven't used the windows versions of EpiInfo for years, so I don't know whether there is a way to reorder cells without recoding.
Jamie
On 2010-10-12, at 9:31 PM, epidata-list@lists.umanitoba.ca wrote:
A little while ago when you answered my question about changing the Boolean fields to 0-1, you mentioned there was a way in both EpiData Analysis and EpiInfo to reverse the order that two by two tables were shown, that would help in a situation where 0 was no and yes was 1 (so the 1 or yes answers appeared first in the table). I have looked around Epi Info but couldn't find a setting like this. Could you let me know how to do it? And also how to do it in Epidata analysis? I
Many thanks
Christine
participants (1)
-
epidata-list＠lists.umanitoba.ca