COR 701 Written: March, 1977 Revised: July, 1999 SPECIAL PROBLEMS IN THE CLEANING OF GRADUATE'S EDUCATION DATA FROM THE 1975 SURVEY A. The dates for completing high school and starting college were inconsistent in a few cases. Some graduates apparently took university courses while still in high school while others didn't officially graduate from high school until a number of years after 1957 and had attended college before t then. B. Physicians frequently included internships or residencies in their reports of regular education. Other MDs only had problems with internships/residencies and first job: generally, a physician's first job is an internship or residency. In both cases changes were made to make the information consistent. C. Problems with currently enrolled graduates: 1) Question 5, regarding last date of college attendance, was sometimes incorrectly answered by those who were currently enrolled. For these graduates, last date of attendance should be the date of the interview. Therefore responses with different dates were corrected. 2) There were some cases of a "yes" to Question 10, ("Are you currently enrolled?"), being inconsistent with a "no" to Question 8, ("Have you been enrolled since your last degree?"). The difference in the wording of the two questions may have been the cause. Most of these cases were resolved in favor of the response to question 10, but some were changed in favor of the response to question 8. 3) Dates and years of schooling are probably fuzzier for currently enrolled people (with some reporting years completed that were not) as well as for those cases where the interview occurred at a time of transition. D. Coding of open-ended questions: 1) The coding scheme for name of highest degree (Question 7a) is one for degree titles, but often a specific title was not obtained as a response. It appears that such responses were either given a specific code or relegated to a residual category. We feel that stricter standards in giving such responses specific codes would result in a more accurate coding procedure and allow more control over these allocations. As it stands, one is not certain that response given a certain code was in fact the corresponding degree. The residual categories are mixed-they contain both specific degrees of low frequency and nonspecific responses. The rationale behind the treatment of low frequency degrees doesn't seem sound, given the large number of nonrealized codes (i.e., code values in the code scheme that have zero frequency) and the high specificity of some other codes IDs that were originally 00 on EDDGNM (name of highest degree) were allocated to other categories. The new categories were generally the level-specific residual ones. Exceptions are R.N. and D.C. for which new codes were created. 2) The coding scheme for major fields of study and vocational training is more problematic. This is due to its origination, the artificial method of coding responses that didn't appear as codes and the lack of a second stage revision of the scheme itself (see Memo #5 pp. 1-3). Some resultant problems: a) Codes with fairly specific names (or names that give a stereotypic image) often are in fact general categories and contain a variety of things. b) Code descriptions are often institution specific because they came from catalogues for particular schools (e.g. UW-Madison, MATC). c) The other (000) category is large. E. Question 7a is awkward for graduates with two degrees at the same level; sometimes they reported the second in Question 8 but most did not. This problem is probably quite infrequent. F. Timing problems arose most frequently among those with a year or less of college. These graduates apparently forgot that the school year incorporates two calendar years. G. Two cases had June 1957 as the date when they completed eleventh grade. Possibly these were not seniors in 1957. We decided to leave them in. H. One should be aware that the three "measures" of schooling (i.e., dates, number of years, and degree received) are not perfectly associated. Changes were made only on the most flagrant inconsistencies. I. The interview schedule presumes that if one has received a graduate or professional degree, then one has also received a baccalaureate degree. This is not always true (e.g. dentists frequently don't get a bachelor's degree). Cases for which this was true were given special codes as follows: 996 for CMDTBA, "Month Last Attended for Baccalaureate Degree"; 996 for EDMJBA, "Major (Baccalaureate)"; and 96 for EDBANM, "Degree Received (Baccalaureate)." J. Some majors (e.g., medical technology) are offered both in college and vocational school and some schools may be considered either college or vocational school (e.g.,Madison Area Technical College). There were a few individuals who took similar studies, but classified themselves differently in this regard. The problem was not widespread and with one exception no reclassification was done for these cases. K. There were several individuals who had double majors for a single degree. They were coded under the first major given. Information on the second is lost. L. The century month allocation procedure was different for some date variables in the education section than elsewhere in the questionnaire, because of special regularities in college attendance. See COR #256. M. The answers to questions 14(b-i) and 15(b-i) were merged in accordance with COR #229a. Generally, this was done because the questions were felt to have elicited responses in the order of importance of the programs. Thus the information from Question 14 was listed prior to that from Question 15. Exceptions to this are stated in COR #229a. N. The (Apprenticeship/Program/Other) distinction for variable, VTAPPR, "Apprenticeship Program Experience" was made post hoc. See COR #256. O. A general comment must be made in regard to the quality of data in the Vocational-Technical section. There was more missing data and more various responses to the questions asked than in the other educational section. We feel this reflects the nature of the schooling rather than poor interviewing. One of the questions gave the graduates unusual difficulty and perhaps should be dropped from future questionnaires: The respondents had a lot of trouble deciding whether a school was commercial or private. A measure of the intensity of the training (e.g. full-time/ part-time or how many days per week) would be valuable and might be included in future questionnaires.