I see that the format that we receive the ‘start’ and ‘end’ metadata has changed (see screenshot).
Unfortunately, it seems to have become impossible to isolate the time component of ‘start’ and ‘end’ in Excel. If you try and use ‘text to columns’ for example, the output doesn’t make much sense.
Stata also seems to only import the date component and not time.
Could you please help me figure out how I can isolate the time component or get Excel to recognize the time part?
K.
Answering my own question, in the hope that it will help someone someday!
You need to use the advanced download options and select ‘store date and number responses as text’
Excel and Stata can handle it better, I think and it can be useful if you want to analyze how long it took enumerators to collect data.
Hello,
I would like to propose that the KoBo export is changed back to the previous (text) format for start and end and other time including (long) formats, as it was in the past.
We don’t want to get anything as text now, e.g. simple dates.
@Josh
Hi @wroos, the purpose of the export option “Store date and number responses as text” is to continue with the previous export behaviour if that’s desired.
Hi @Josh,
this was well understood, thanks.
Nevertheless, we would prefer that in the new date/numeric export version no data are lost/cut. So, it might be better to keep (only) any date-time variables as text here, e.g. start and end.
Hi @wroos, ok I see. So are you saying that it’s still useful to have the numeric types in the export but be able to specify that the dates remain as text?
Dear @Josh,
Yes, we would like - as standard - to have ONLY those (few) values as text which cannot be converted by KoBo without loss of content or format-structure, like the long date-time-timezone combinations as start.or end, or the select_multiple combined text var.
An objective is to reduce reformatting work in Excel, SPSS etc., esp. after XLS download.
But we might also need to assure that those exported data can be reimported to KoBo.
