RE: Differences in the /IMPORTCASE=... subcommand between SPSS and PSPP

[Charles Johnson]

CJT


----------------------------------------
> Date: Sun, 20 Sep 2015 18:25:00 +0200
> From: address@hidden
> To: address@hidden
> CC: address@hidden
>
> I suppose if that is the spss behaviour, then that is what PSPP should do 
> also.
>
>

   I asked the question because I do not know if it was a feature of PSPP.

Unfortunately, the documentation is contradictory.
 
* On the one hand, on the subcommand IMPORTCASES (SPSS 21)

Percent n
Read approximately
the first n percent of cases. The value of n must be a positive integer less 
than 100.
The percentage of cases actually selected only approximates the specified
percentage. The more cases there are in the data file, the closer
the percentage of cases selected is to the specified percentage.

http://www-01.ibm.com/support/knowledgecenter/SSLVMB_21.0.0/com.ibm.spss.statistics.help/syn_get_data_importcases.htm?lang=es

   * Moreover, on the import wizard (the same command): (SPSS 21)

Wizard step 3
[...] or a random sample of a specified percentage. Since the random sampling
routine makes an independent pseudo-random decision for each case,
the percentage of cases selected can only approximate the specified
percentage. The more cases there are in the data file, the closer
the percentage of cases selected is to the specified percentage.

http://www-01.ibm.com/support/knowledgecenter/SSLVMB_21.0.0/com.ibm.spss.statistics.help/idh_text_wizard_step3_fixed.htm?lang=es

* The cherry on the cake: (SPSS 22 & +)

"The IMPORTCASES subcommand is obsolete and is
ignored. To obtain a random sample of cases, use the SAMPLE command."

http://www-01.ibm.com/support/knowledgecenter/SSLVMB_22.0.0/com.ibm.spss.statistics.reference/spss/base/syn_get_data_importcases.htm

Cheers