I got my original file from the Social Science Data Center in Israel. However, the first simple check is the number of observations: my file has 1,113,420 observations; but the very same file (Census-1995, Israel) at IPUMS reports 556,365 observations only (half the original!!). Why is that? Is IPUMS just a “sample” of the original data??
Why in Israel Census-1995 IPUMS has 556,365 observations, while my original file includes 1,113,420 obs?
Yes, the IPUMS-International sample is a 1:2 sample of the original data, as noted on the IPUMS sample description page for the 1995 microdata for Israel (see sample design statement): https://international.ipums.org/international/sample_designs/sample_designs_il.shtml
Privacy and statistical confidentiality concerns may be greater for IPUMS disseminated samples than for a national source of dissemination, such as the Social Science Data Center. Generally, 10% is the maximum for samples disseminated by IPUMS. For more details on IPUMS-International privacy and statistical disclosure control procedures, please see: http://www.hist.umn.edu/~rmccaa/ipums_psd20100713.pdf and similar publications listed on my personal home-page. Please email me to discuss this further.