# Multiple numerical input

1 year 7 months ago #157601
I formulated a 'Multiple numerical input' questions, asking for the budget of organisations in (a) 2015 and (b) 2016. The survey is finished. Exporting the data I found out that Limesurvey presents the data of the respondents in a very strange, unusable way:

Limesurvey fills the number of decimals till 10, some examples:

Input of respondent
1000
Output of Limesurvey:

1000.0000000000

Input of respondent:
1.000

Output of Limesurvey:
1.0000000000

In some cases the original input of the respondents cannot be traced back (does he mean a budget of 23? 23000 or 23000000?) Does Limesurvey saves the original input of the respondents? And if so, how can I find this?
Looking forward to any help in this! Thanks in advance!

Please Log in or Create an account to join the conversation.

LimeSurvey Partners
1 year 7 months ago #157608 by LouisGac
Replied by LouisGac on topic Multiple numerical input
I don't understand your problem. Why is it a problem for you to have null decimals?

If you use numerical input, it means that the record in DB is numerical

A numerical entry in a db always uses a point as a decimal separator and nothing as thousand separators, and a have a specific number of decimals (10 in limesurvey case). See:
dev.mysql.com/doc/refman/5.7/en/precisio...characteristics.html

If what you want is to record the exact entry of participants, without converting it to a numerical value, use multiple short text question:
manual.limesurvey.org/Question_type_-_Multiple_short_text

Please Log in or Create an account to join the conversation.

1 year 7 months ago #157679
Dear LouisGac,

Thank you very much for your reply!
The problem is not to have null decimals, on the contrary. The values are supposed to have null decimals, but Limesurvey added these decimals, which was never my intention.

Concretely, this means that if a respondent typed 100 for this variable, that Limesurvey automatically adds 10 decimals, so that the output becomes 100.0000000000. Normally, this wouldn't be a problem, and I would simply delete the 10 decimals. However, if a respondent filled in a dot himself (which is not intended as a decimal but as a way to make the number more readable, 100.000 means one hundred thousand for example), Limesurvey does not add a dot with ten decimals, but simply adds zeros until there are 10 digits behind the dot.

The following problem occurs:

- a respondent types "100" (one hundred), the Limesurvey output becomes "100.0000000000"
- a respondent types "100.000" (one hundred thousand), the Limesurvey output becomes "100.0000000000".

As you can see, the output is identical, which makes it impossible for me to recognize whether the respondent originally typed 100 (one hundred) or 100.000 (one hundred thousand).

Therefore, I would very much like to know if there is a way to retrieve the original, unaltered input of the respondent? So the original respondents' input before Limesurvey automatically added all these unwanted decimals. I am aware that this problem could have been circumvented if some survey settings would have been different. However, the response period for the survey is already closed, so I simply have to deal with the data I've got.

I'm looking forward to hearing from you. It would be of great value to me if you have any thought or solutions for this problem.

Thank you very much!

Please Log in or Create an account to join the conversation.

1 year 7 months ago #157688 by LouisGac
Replied by LouisGac on topic Multiple numerical input
if your participants used decimal separators as a thousand separator: the data is lost.

You could have done different things to avoid this problem (like forcing integer values, displaying a warning message, use a text input to get the exact input of the user etc).

Now, you can't retrieve the data. Sorry for that.

Please Log in or Create an account to join the conversation.

1 year 7 months ago #157693
Thanks for your response! I was already afraid of that!

Please Log in or Create an account to join the conversation.