A real world example of the need for labled missing values.

[Art Kendall]

https://www.washingtonpost.com/politics/2021/12/10/wisconsin-voter-registration-conspiracy-debunked/

[Art Kendall]

How to have labeled missing values on date variables.
New file.
data list list /MyDate (adate10).
Begin data.
12/25/2021
7/4/2021
5/5/2000
12/25/2021
7/4/2021
5/5/2000
12/25/2021
7/4/2021
5/5/2000
10/15/1582
10/16/1582
12/25/2021
7/4/2021
5/5/2000
10/15/1582
10/16/1582
10/17/1582
12/25/2021
7/4/2021
5/5/2000
10/15/1582
10/16/1582
10/10/1800
end data.
display dictionary.
missing values MyDate ('10/15/1582' THRU '10/20/1582', '10/10/1800').
Value labels MyDate
'10/15/1582' 'unknown'
'10/16/1582' 'refused to answer'
'10/10/1800' 'forgot to ask'.
display dictionary.
frequencies MyDate.

[jkpeck]

Note this requirement, quoting the CSR,
values must be expressed in the same date format as the defined date format for the variable.

[parisec]

Thanks Art. This is a nice bit of code to have handy!

Carol

[Art Kendall]

For a long time I was mistaken that dates could not have missing values.
I had tried tried day 0 as missing.  I later found out that we can have many labelled missing values as long as we start them at day 1.


missing values MyDate(10/14/1582).

I was going to test the syntax above for day 0.  But for some reason SPSS 27 is coming up with a Java error and SPSS 24 will not start. The timing of this is not as bad as it could be since I have several SPSS projects that may be starting, I have none at the moment.

[Art Kendall]

I finally downloaded SPSS 28 and updated the example syntax to force error messages for Day 0.

TMissing_Values_for_Dates.sps

[parisec]

I was also under the same impression, that you could not label dates as missing, which is why i was really happy to see this code!