1970 Integration: ASCII files for Individuals and for Households

Dear Doug: Let me try to reconstruct some history.

First, the three sets of hh data for 1945, 1960, and 1970 were created in 1970-1971 for purposes of doing my dissertation. I coded by hand all three then available village censuses to get comparative data to use in what is now the first chapter of my Migration and Adaptation book (1977). It seemed logical to use the 1970 jefe ID number as a way to identify the cases across all three censuses because the 1970 number was freshly minted at that time. It had just replaced the 1960 number, and, of course, no 1945 number existed to get in the way. All of the analysis was done on the Berkeley mainframe using a litle program called TABLE.

I am not sure that a computerized file of the Fortran coding sheets for 1945, 1960, and 1970 still exist. Certainly, they formerly existed, but the raw data are on Hollerith cards still kept in my lab at SMU. The data at one time existed on a big computer tape on the mainframe, and those files were downloaded to diskettes (5.25 format) back in the late 1980s. But, I am not sure that they ever were passed over to 3.5 format diskettes. So, if you have someone willing and able to do the relatively small amount of typing, then go for it! You do have photocopies of all three hh files, right? The 1945 file has 246 lines of data; the 1960 file has 320 lines of data; and the 1970 file has 360 lines of data. By the way, note that the first coding sheet for 1960 census has an error on line 6, cols. 26 and 27. The correcty entry should be a "0" in col. 26 and a "2" in col. 27. This will make it agree with the first record "001" for the same two cols.

The fichero numbering system only started after the 1960 village census. Therefore, there is no parallel ID code number for 1945. The numbers in the individual entries in the fichero for those persons living in the village in 1945 correspond to the number assigned to the household in which they were living in 1945. It is not a personal ID number.

Best regards, and congratulations on all your excellent integrating work so far. Van

P.S. I have no recollection of what a TPF extension would involve on any file with a filename = Tzin45-80.

1980 Analysis of Earlier Data

First response: I canĄt think of any place where we ever tried to match up the 1970 ID code numbers (as used in the 1945, 1960, and 1970 hh data files) with the 1980 ID code numbers (as used in the 1980 hh data files and in the various personal data files).

Second respons: I have just encountered three interesting computer printouts called FREQ45, FREQ60, and FREQ70. I will bring these with me for us to look over. They were all created based on the 1980 fichero and census data. the underlying SPSS file name is CONT3M.

Let me emphasize that FREQ45, FREQ60, and FREQ70 are hard copy printouts. They all were created from a data file called CONT3M. I have just had son John copy all of my old 5.25 disks to 3.5 format disks. And these ten disks I have, in turn, copied to a 100 MB ZIP disk. Further research in the printout reveals that CONT3M was a subfile created on 06 Dec 1984 from TZ45, TZ60, and TZ70, respectively. These three large data files had 4075 records which corresponds to what we had in the master fichero after the 1980 census. So, these printouts deal with individual data for the first three census dates. We do have some files that look like TZ60 and TZ70; both have .DAT extensions. I canĄt locate a parallel TZ45.DAT, but there is a TZ80.DAT.

These are all files now transferred from the 5.25 disks to the 3.5 disks and also to the ZIP disk (and to my hard disk here). Do you have these files there in La Jolla or in Irvine? I am pretty sure that you do. Anyway, I will be able to check my email messages again on Thursday morning about 8:30 am my time. So, if you need them, I can probably figure out how to attach them to a message and send them along to you.

I am going to send by EXpress Mail the ZIP disk plus photocopies of the actual hard copy printouts of the FREQ45, FREQ60, and FREQ70. Then, I think that you will be able to "recreate" the printouts by seeing what are the underlying files and reconstruct the SPSS command structure.