Analysing data from multiple select/ODK aggregate

Support
1

Hi,

maybe this is outdated, but since i stumbled upon that little problem, i wanted to share my solution.
working with excel 2010 i found my solution in using a straight forward excel formula.

=WENN(ZÄHLENWENN(C$1;""&$B2&"");1;0)
in English i think it is:
=IF(CountIF(C$1;""&$B2&"");1;0)

given you have your respondents listed in columns (starting at column "C"), your answers/questions in lines:

for every possible option add one line and put the options text in column "B"
so you got this:

A B Resp1 Resp2 Resp3
question_x original b1 c1 a1 a1 n c1 b1
question_x a1 1 0 0
question_x b1 1 0 1
question_x c1 1 1 1
question_x n 1 0 0

put the formular in the cells for every option and respondent and you get 0 and 1 instead of multiple (part)strings.

please forgive my poor english, hope that may help

··· Am Donnerstag, 4. April 2013 20:17:17 UTC+2 schrieb TLG: > Hi, > > I wondered if anyone had any tips on analysing data gathered using a 'select multiple' question, with data processed by ODK aggregate > > > At the moment, for each set of responses it produces a string separated by spaces in each cell, for example: > > 1 2 > 3 > 1 3 4 > 2 4 > > But the problem is, this makes things tricky for data analysis. Sometimes, I am interested just in everyone who answered '2' - but it isn't possible to filter the results to see just the people who answered '2' - because some people answered '1 2', some '2 4' etc. > > > I wondered what your tips were on dealing with this kind of data? > > Thanks! > > TG