The key variables, iduser and idswl have been removed from the public and private data, respectively. The new public key will be idpub and the new private id will be idpriv. Individuals who receive private data by request will be issued a randomly generated id that will not be linkable to any other past or future WLS data. These changes are necessary to increase the security of all WLS data.
Imputed data values for missing data on economic Variables are available for graduate and sib data. For graduates imputed values are available for the 1975, 1992, and 2004 waves of data. For siblings imputed data are available for the 1994 and 2005 wave.
For each variable, five separate imputations are available. The first 6-7characters in the imputed variable names match the first 6-7 characters in the original analysis variables. The imputed variable names then add an additional two characters. The first is the letter "I" and the second is a digit between one and five indicating the imputation number.
For more information on how the imputations were created and how to use the data please see cor1014 in Appendix L (see Appendices).
An update to the attractiveness data is now available. Due to some minor errors in processing, the removal of cases with less than 11 ratings, and the addition of 4500+ new cases, all variables from the previous release (variables starting with YRBK) have been archived.
More cases have been added to the RBMI data set.
The following variables have been updated
statkdna: Status code for return of Oragene Kit for graduate sample
statcdna: Status code for consent to use Oragene Kit for graduate sample
statmdsq: Grad Medicare Part D SAQ response status for graduate sample
deatyr: Year of graduate's death.
deatmn: Month of graduate's death.
livgrad: Is the graduate alive as of current release as far as we know?
cmbrdx: Graduate's century month of birthdate
brdxdy: Graduate's year of birth.
xdeatyr: Year of Sib-Respondent's death.
xdeatmn: Month of Sib-Respondent's death.
xlivsib: Is the Sib-Respondent alive as of current release as far as we know?
xcmbrdx: Century Month of Sib-Respondent's Birthdate
xbrdxdy: Year of Sib-Respondent's Birth
The following variables have been added
xstatkdna: Status code for return of Oragene Kit for sibling sample
xstatcdna: Status code for consent to use Oragene Kit for sibling sample
dupid: ID that links a graduate record to its duplicate (as a selected sibling of another record)
xsibstat: Selected sibling status
xduprtp: Respondent type of the sibling record's duplicate (either as the graduate or the sibling of another record)
xdupid: ID that links a sibling record to its duplicate (either as the graduate or the sibling of another record); see codebook for an example
In the previous releases, negative (missing) values in the constituent variables were inadvertently included in the calculation of the Total Score for Psychological Well-Being for the 2004 graduate sample (GN114REC). In the new release, GN114REC is now coded -2 (INAP) if at least one of the constituent variables is negative (missing or INAP). Twenty-seven respondents are affected by this change because they did not give a valid answer to at least one of the eighteen items that constitute the scale.
Two 2004 female graduates were given the male version of mail form. As a result, the question about prostate problem (IX152RER) that should have been coded -2 (INAP) had a value of 2 (no problem) or -3 (refused). The variable now is coded to -2 for these two respondents. This change also affects IX152ARE (age first diagnosed), IX153RER (how much the problem interfere with what R like to do), and IX117RED (the number of illness items answered).
In previous release, the summary scores for psychological distress/depression for the graduate and sibling respondents (IU001REC and DU001REC) were calculated by summing across all constituent items, with four of the items reversed coded. Items with missing values were given a score of zero. The summary scores were then recoded to -2 if any of the items was missing. In current release, the summary scores have been re-calculated using the same rule for constructing the 1993 graduate version of the variable (MU001REC). Based on the rule, a summary score was calculated by summing all items with valid responses (0-7 days) if the respondent gave valid responses to at least ten questions or if the respondent gave nonzero responses (1-7 days) to at least three items and left all the remaining items blank. In the later case, it was assumed that the respondent only answered the questions that applied to them, and therefore each of the blank items was given a score of zero. The changes also affect IU002RE and DU002RE (Number of clinical depression questions answered). See Documentation of Scales for details.