EDUCATIONAL VARIABLES 1964

                COMPUTER OPERATIONS REQUEST #275

     The purpose of this COR is to explain the procedures used
when coding college data from the 1964 follow-up. When R attended
more than three colleges, we needed to make a decision over 
which three colleges were to be selected for coding.  Our
analysis revealed that, in seventeen cases, we had a respondent
who attended four or more colleges.  

     Selection of the three colleges to be coded if R attended 4
or more colleges appears in the following order of priority:

1.   If R obtained a degree (BA or higher), the degree granting   
     institution was coded.

2.   If R was currently enrolled, current College was coded as   
     third college attended. This information was occasionally    
     obtained from the current occupation section of the       
     questionnaire rather than from the educational history.

3.   Schools attended for the longest periods of time were coded.

4.   Colleges which could be coded easily with exact campus       
     information were more likely to be selected. Parents tended  
     to give more complete information on Wisconsin colleges
     (i.e., distinguishing between different campuses; University 
     of Wisconsin-Madison) then they did on out-of-state colleges 
     (University of California-no campus specified).

5.   In several cases R attended summer school at many different  
     institutions. The "better known" or Wisconsin colleges       
     appear to have been coded.

6.   Foreign colleges were given the lowest priority. They were   
     only coded if the foreign school experience was more         
     extensive than the U.S. experience or if a degree was        
     obtained from the foreign school.

     In addition, it is worthwhile to note that when the
variables CLYDG6 (number of years in college up to completion of
the first degree) and EDYR64 (total years of schooling), were
created, they were based on the number of years attended for the
three colleges coded. Therefore, for those who attended 4 or more
colleges, these variables report fewer years of schooling than
actually occurred. This under-report is minor since the most
significant and lengthy components of R's educational experience
were those selected for coding.