COR614.WP5
                                                       1/26/96

See Cor614a.asc for the log file from which the following
information was obtained.

Problem:
The 1992-93 respondent is asked if his/her reciprocal friend is
still alive.  If the response is "no", then year and state of death
are ascertained. Seventeen respondent's indicated that their
reciprocal friend was deceased.  Ten of these already had deceased
codes in our WLS23_2 data file.  All of the remaining seven friends
had completed both telephone and mail surveys in 1992-93.  The date
of the telephone interview for four of these friends was later than
the date of the interview for the respondent who indicated they
were deceased.  Another friend returned the mail questionnaire
after the respondent reported his/her death.  This only leaves two
possible deceased friends.  One of these was reported as deceased
in 1981, but this friend completed the telephone interview in 1992. 

The last one appears to be an accurate report.  The friend
completed the phone interview on October 9, 1992 and returned the
mail survey October 23, 1992.  The respondent, who reported the
death of this friend, completed the telephone interview on December
17, 1992.  R reported that the friend had died in 1992 and lived in
Minnesota (state code 24) when he died.  The friend is coded as
living in Minnesota at the time of interview.   In addition on the
mail survey, the friend indicated he was in poor health and had
been in the hospital 26 days in the past year. (The record for
this deceased friend was updated in the February 1996 release of
WLS23.)   

Solution:
Add code 3 for RM002RER for cases in which respondent reports 
friend is deceased, but friend was alive at time of interview. 
Add code 4 for the case in which respondent indicates friend is
alive, but other data indicate friend is deceased.