COR713.ASC Discussion of changes to 1970-basis mappings of occupational status scores: DUNCAN SEI, SIEGEL PRESTIGE, MSEI2, TSEI2, and OCCUPATIONAL EDUCATION AND INCOME SCORES. COR713.SAS contains the code which will MAP these scores when passed a 1970-basis Detailed Occupation code, Detailed Industry code, and Class-of-Worker code (Read the instructions in COR713.SAS to be sure that missing data is handled properly.) COR713l.SAS and COR713m.SAS contains the same code, only formatted as a SAS MACRO. This COR proceeds as follows: I. Introduction II. Definitions/Sources III. Checking of Duncan SEI/Siegel PRESTIGE scores A. Extended parenthetical comment about COR713h.XLS, and how to use it B. Problems with Non-Split Occupation lines C. Problems with Split Occupation lines D. FINAL LIST OF DUNCAN SEI/SIEGEL PRESTIGE SCORE CHANGES E. WHAT DIFFERENCE DOES IT MAKE? IV. Checking of MSEI2/TSEI2 scores A. FINAL LIST OF MSEI2/TSEI2 SCORE CHANGES B. WHAT DIFFERENCE DOES IT MAKE? V. Checking of 1970 Occupational Education and Occupational Income scores A. FINAL LIST OF OCCUPATIONAL EDUCATION AND OCCUPATIONAL INCOME SCORE CHANGES B. WHAT DIFFERENCE DOES IT MAKE? VI. References VII. List of Attached Documents (COR713a - COR713m) *************************************************************************** *** I . I N T R O D U C T I O N *** *************************************************************************** Through an annoying ability to focus on detail, I discovered that the SEI scores used in the WLS for "Miscellaneous and Not Specified Laborers" (1970 Detailed Occupation Codes 780 and 785) were exactly the same as those used for "Miscellaneous and Not Specified Opereratives" (1970 Detailed Occupation Codes 690, 692, 694, and 695). Mentioning this observation to Bob in the middle of a completely different conversation, I was soon immersed in the search for the source of the error, the true scores, and the scope of the problem for not only the WLS data, but other data sources to which we may have appended the bad scores. After concluding that, indeed, only the WLS data are affected by this error, I agreed to research the source of ALL SEI/PRESTIGE scores used in COR521c and COR612c, and to create the last list of these scores to use as THE DEFINITIVE LIST. COR713.SAS (and the COR713l.SAS and COR713m.SAS macro code) is this definitive list. COR713.SAS can be used with other datasets that use 1970-basis Occupation/Industry codes as well-- but be sure to follow instructions for missing data (see notes in Section IIIc below), and Allocated Occupation categories. *************************************************************************** *** I I . D E F I N I T I O N S / S O U R C E S *** *************************************************************************** Class-of-Worker --------------- Distinguishes workers by where their pay comes from. The categories used by the WLS include: 1 - Salaried Worker 2 - Government Worker *3 - Self-Employed Worker, Business Incorporated 4 - Self-Employed Worker, Business Not Incorporated 5 - Unpaid Family Worker 6 - Family Worker (Not Further Specified) * This category is considered "Salaried" by the Census, but is often changed to "Self-Employed" in the WLS. Allocation Code (Occupation) ---------------------------- One of 12 1970-Basis Occupation codes, corresponding to the 12 Major Occupation cateogries used by the Census to randomly assign a case with missing data to a Major Occupation category. The WLS uses Allocation codes for responses too vague for the Detailed Classification System (generally, Aspirations variables in Wave 1). Split Occupation Lines ---------------------- Some Detailed Occupation categories are quite large and heterogeneous; for example, "Managers and Administrators n.e.c." groups the small grocery store owner with the CEO of IBM. Many Census documents break such large categories down into groupings of workers based on (1)their Detailed Industry designation, and/or their (2)Class-of-Worker designation. Occupation lines which are broken down in this way in order to assign status scores (SEI, PRESTIGE, etc.) are referred to as "Split". Major Group (Occupation) ------------------------ Occupations which are somewhat similar in their characteristics and/or types of duties are lumped together by the Census into 12 Major Groups in 1970: (1)Professional, Technical and Kindred Workers; (2)Managers and Administrators, Except Farm; (3)Sales Workers; (4)Clerical and Kindred Workers; (5)Craftsmen and Kindred Workers; (6)Operatives, Except Transport; (7)Transport Equipment Operatives; (8)Laborers, Except Farm; (9)Farmers and Farm Managers; (10)Farm Laborers and Farm Foremen; (11)Service Workers, Except Private Household; (12)Private Household Workers. Major Group (Industry) ---------------------- Industries which are somewhat similar in their characteristics and/or products are lumped together by the Census into 12 Major Groups in 1970: (1)Agriculture, Forestry, and Fisheries; (2)Mining; (3)Construction; (4)Manufacturing; (5)Transportation, Communications, and Other Public Utilities; (6)Wholesale and Retail Trade; (7)Finance, Insurance, and Real Estate; (8)Business and Repair Services; (9)Personal Services; (10)Entertainment and Recreation Services; (11)Professional and Related Services; (12)Public Administration. FSD --- Featherman, Sobel, and Dickens (1975). Appendix B lists the Duncan SEI and Siegel PRESTIGE scores, mapped to the 1970 Detailed Occupation Classification system. S&F --- Stevens and Featherman (1981) paper creating the MSEI2 and TSEI2 scales. *************************************************************************** *** I I I . D U N C A N S E I / S I E G E L P R E S T I G E *** *************************************************************************** Duncan created his SEI scale based on the 1950 Census Detailed Occupation categories (reproduced as COR713a.ASC). This 1950 system was very similar to the one used in 1960, and Duncan created a "mapping", showing which of his 1950-basis SEI scores were to be used with the 1960 Census Detailed Occupation categories (COR713d.ASC). He also created a list of SEI scores to be used with the Major Occupation categories--both the Census-defined 12-category system, and his own 17-category system (COR713b.ASC and COR713c.ASC). Siegel used various occupational prestige surveys to produce PRESTIGE scores for all Occupations in the 1960 Census Detailed Occupation categories system (COR713e.ASC). This table includes PRESTIGE scores for the Major Occupation categories as defined by the Census's 12-group system. To use these scores for the 1970 Census Detailed Occupation categories, FSD used Tech. Paper 26 to describe each 1970-basis Occupation in terms of its 1960-basis components. The percentages were used as weights, and multiplied by the status score for each 1960-basis Occupation, thus producing a status score for the 1970-basis Occupation category. For example, the 1970 Occupation category "1=Accountants" is comprised of four different 1960 Occupation categories, with the following percentages of the MALE EXPERIENCED CIVILIAN LABOR FORCE: 0 = Accountants and auditors 95.729% 174 = Statisticians and actuaries .232% 301 = Agents, n.e.c. .582% 370 = Clerical and kindred workers, n.e.c. 3.457% To arrive at the SEI and PRESTIGE scores for the 1970 Occupation category of "1=Accountants", just take the weighted averages of each SEI and PRESTIGE score associated with the 1960 categories: % of 1960 1970 1960 1960 Occ. Occ. SEI PRESTIGE ---- ------ ---- -------- 0 95.729 78 56.7 174 .232 81 55.4 301 .582 68 42.5 370 3.457 44 36.2 ---- -------- WEIGHTED AVG: 76.8 55.9 Multiplying the SEI and PRESTIGE scores by 10, we get a 3-digit score for each 1970 Occupation category (SEI=768 and PRESTIGE=559). The SEI and PRESTIGE scores that FSD calculated in this way for the 1970- basis Detailed Occupation categories are included in COR713g.ASC. Now, to check that the SEI and PRESTIGE scores are correct for use in the WLS, one must pay attention to two things. First, note that FSD did NOT include scores for 1970 Detailed Occupations that are split by Industry and/or Class-of-Worker, even though Duncan's original lists contained these Splits, and Siegel's list contains 1960 Occupation lines split by Industry (but not Class-of-Worker). Second, note that all of these calculations and printings of the SEI/PRESTIGE scores for 1970-basis Occupations were done in the days before computers, word processors, spreadsheets, and statistical packages. The lists had to be typed by hand, and many errors could have been introduced. For example, FSD's original list was "reproduced" in Hauser and Featherman's 1975 book "The Process of Stratification: Trends and Analysis", but NUMEROUS typographical errors were introduced, compared to the original FSD list. So how does one know whether FSD also had typographical errors in *their* original list! To handle the first problem (split scores missing), we followed the solution used by Hauser in his 1975 memo to Mossel and Hauser (COR315--see MEMO133.ASC). 1960-basis Occupation lines split by Industry and/or Class-of-Worker are assumed to DIRECTLY map to the same Split in the 1970-basis Classification System. Thus, Occupation Split lines are not to be handled in the same way as other Occupation categories--that is, they are not to be "mapped" using Tech. Paper 26, as FSD did for the Non-Split Occupation categories. To handle the second problem, I calculated the mappings myself, using a spreadsheet program. I had already typed Tech. Paper 26 into an Excel spreadsheet for another project, so I cleaned it up, added "total" lines for each 1970 and 1960 Detailed Occupation, added the Industry and Class- of-Worker Splits where appropriate, and finally added the original Duncan SEI and Siegel PRESTIGE scores from the original sources into the spreadsheet, so that I could calculate the scores in the same way FSD did. This spreadsheet has been saved as COR713h.XLS. ****************************************************************************** *** A. Extended parenthetical comment about COR713h.XLS, and how to use it *** ****************************************************************************** *** *** *** The Excel spreadsheet of Tech. Paper 26 is found in COR713h.XLS. *** *** To sort it by 1960-basis Occupations, highlight the ENTIRE spreadsheet *** *** (A11..Z3196), and click on the DATA/SORT menu option, sorting on *** *** COLUMN A. The worksheet will sort the 1960 Detailed Occupation *** *** categories in order (column C). Starting at row 2510, the *1970* *** *** Detailed Occupations are listed in order; these are the "total" rows *** *** for each of the 1970-basis Occupations (notice that "total" rows for *** *** Splits within the 1970-basis Occupations are also included, although *** *** we will not use these scores.) *** *** *** *** To sort the worksheet by 1970-basis Occupations, highlight the ENTIRE *** *** spreadsheet (A11..Z3196), and click on the DATA/SORT menu option. *** *** Sort the worksheet by COLUMN B. The worksheet will now be sorted *** *** according to 1970-basis Detailed Occupations (column F). Starting at *** *** row 2689, the *1960* Detailed Occupations will be listed in order; *** *** these are the "total" rows for each of the 1960-basis Occupations. *** *** *** *** The Duncan SEI scores are located in columns U, V, and W. Column U *** *** contains the 1960-basis Duncan scores, as listed in Duncan's Oct. 1961 *** *** mimeo list (reproduced as COR713d.ASC.) Column V (labeled *** *** COR713 1970 DUNCAN MALE) is the Duncan SEI score as constructed using *** *** this Excel Worksheet. Column W contains the 1970-basis Duncan SEI *** *** (Male-Based) scores as listed in Appendix B of FSD (COR713g.ASC). *** *** *** *** Similarly, the Siegel PRESTIGE scores are located in columns X, Y, and *** *** Z. Column X contains the 1960-basis Siegel scores, as listed in Table *** *** 5 of Siegel's 1971 dissertation (reproduced as COR713e.ASC.) Column *** *** Y (labeled COR713 1970 SIEGEL MALE) is the Siegel PRESTIGE score as *** *** constructed using this Excel worksheet. Column Z contains the 1970- *** *** basis Siegel PRESTIGE (male-based) scores as listed in Appendix B of *** *** FSD 1975 (COR713g.ASC). *** *** *** *** It is very important to note that the SEI/PRESTIGE scores used for *** *** 1970-basis Occupation lines that are split by Industry and/or Class-of-*** *** Worker are NOT FOUND IN THE EXCEL WORKSHEET (COR713h.XLS)!!!!! The *** *** reason for this is that Tech. Paper 26 does NOT divide the large 1970 *** *** Detailed Occupations split by Industry and/or Class-of-Worker in the *** *** same way that Duncan and Siegel split their 1960-basis scores. For *** *** example, in 1970 Occupation 690, the Detailed 1960 Industry categories *** *** are noted, but are grouped together into "minor" Industry groupings. *** *** That is, Duncan and Siegel both considered (for example) "Sawmills" *** *** and "Misc. Wood Products" as separate Industry categories with *** *** different SEI/PRESTIGE scores in their 1960 Classification, but Tech. *** *** Paper 26 lumps them together under the heading "Lumber and Wood *** *** Products, except Furniture." *** ****************************************************************************** B. Problems with Non-Split Occupation lines Four main types of problems with the SEI/PRESTIGE scores of the Occupation lines NOT SPLIT by Industry and/or Class-of-Worker were found: (1) Individual errors/typos in scores; (2) Assignment of SEI/PRESTIGE scores for 8 new 1970 Occupations not included in the Tech. Paper 26 mappings; (3) Assignment of SEI/PRESTIGE scores for the 12 new 1970 Allocation codes (corresponding to Major Occupation groups); (4) Bizarre Occupation codes included in the FSD paper which do not occur at all in the 1970 Detailed Occupation Classification System. (1) INDIVIDUAL ERRORS/TYPOS IN SCORES 5 Occupation lines (which are not split by Industry/Class- of-Worker) have differences between the SEI/PRESTIGE score in the FSD paper, and those calculated in the Excel worksheet of Tech.Paper 26. (Only listing the differences), these five are: 1970 Old New Old New Occ. SEI SEI PRESTIGE PRESTIGE ----- ----- ----- -------- -------- 180 518 519 260 423 424 563 211 210 672 283 282 764 086 076 185 175 Clearly, only Occupation line 764 (Vehicle Washers and Equipment Cleaners) has a substantial difference between the scores. The other differences are probably differences in rounding. In the case of the PRESTIGE scores, they are the result of using the average for Misc. Operatives (29.0) or Managers (50.3) rather than the score for an individual Industry Split within that category. These changes have been made, and are updated in COR713.SAS. (2) ASSIGNMENT OF SEI/PRESTIGE SCORES FOR 8 NEW 1970 OCCUPATIONS NOT INCLUDED IN THE TECH. PAPER 26 MAPPINGS Eight 1970 Occupations are not included in Tech. Paper 26 because they "... had very few people employed in them at the time of the 1960 census, and for which no cards were found in the 100,000 sample card deck." (Tech. Paper 26, page 4). These 8 Occupations are: 1970 1970 Occ. Title ---- ------------------------------------------------- 052 Life and physical scientists, marine scientists 073 Physicians, dentists, & related practitioners, health practitioners, n.e.c. 092 Social scientists, political scientists 094 Social scientists, sociologists 131 Teachers, college and university, home economics teachers 156 Engineering and science technicians, mathematical technicians 311 Clerical assistants, social welfare 475 Mechanics and repairmen, data processing machine repairmen Because we have no information in Tech. Paper 26 about which 1960 categories supplied workers into these new 1970 categories, it would seem that the best thing to do would be to actually code job titles from these categories above into the 1960 Classification. Each of the categories above would have been coded into the following 1960 categories: 1960 1960 Occ. Title ---- ------------------------------------------------- 134 Natural scientists, n.e.c., Geologists and geophysicists 193 Therapists and healers, n.e.c. 175 Social scientists, Miscellaneous social scientists 175 Social scientists, Miscellaneous social scientists 053 Professors and instructors, Social sciences, n.e.c. 191 Technicians, other engineering and physical sciences 171 Social and welfare workers, except group 473 Mechanics and repairmen, Office machine Assuming that the new 1970 categories all came 100% from the 1960 categories listed above, we can proceed as if they were included in Tech. Paper 26. That is, we can assign the 1960- basis SEI and PRESTIGE scores to the new 1970 categories, multiplying by 10 to get three digits. The new scores would then be: 1970 1960 New New FSD FSD Occ. Occ. SEI PRESTIGE SEI PRESTIGE ---- ---- --- -------- --- -------- 052 134 800 672 800 672 073 193 580 367 580 367 092 175 810 656 810 714 094 175 810 656 810 714 131 053 840 783 720 532 156 191 620 470 530 468 311 171 640 524 440 314 475 473 360 339 270 408 Unfortunately, these scores are quite different from those found in FSD--especially for the "131=Home Economics Professors", and the "156=Welfare Clericals". The FSD scores seem to be based on some sort of personal judgement. For example, the '714' PRESTIGE score for Sociologists is the same as that for Psychologists, even though in 1960 Sociologists fell into the category of "Miscellaneous Social Scientists" with a PRESTIGE score of 656, and furthermore, "Sociologist" was one of the specific job titles used in the 1964 Hodge-Siegel-Rossi prestige study, so this 656 score was computed specifically using the prestige rating for Sociologists! By assigning '714' rather than '656' to Sociologists, FSD seem to be saying that Sociologists should be at least as prestigious as Psychologists (and more prestigious than Economists, whose PRESTIGE score is '568'.) This kind of judgement call seems unwarranted, of course. The only discrepency that makes sense, and which we have decided to keep, is the much lower scores for welfare clericals. These people were split from a 1960 *Professional* category of "171=Social Workers" to a *Clerical* category in 1970. As such, they have been given the same SEI and PRESTIGE scores as 1970 category "325=File Clerks". The final list of scores for these 8 new 1970 categories, then, is: 1970 New New Occ. SEI PRESTIGE ----- --- -------- 052 800 672 073 580 367 092 810 656 094 810 656 131 840 783 156 620 470 311 440 314 475 360 339 These changes are made in COR713.SAS. (3) ASSIGNMENT OF SEI/PRESTIGE SCORES FOR THE 12 NEW 1970 ALLOCATION CODES (CORRESPONDING TO MAJOR OCCUPATION GROUPS) Twelve codes in the 1970 Census Detailed Occupation Classification System have been assigned as "Allocation" categories." Allocation categories correspond to the 12 Major Occupation Groups. The Census uses these 12 codes to randomly place individuals who have nonresponse for Occupation (the Detailed Industry Classification System also has a set of Allocation categories in 1970.) In the 1960 Occupation Classification System, nonresponse on Occupation was coded '995'. The WLS has used these allocation codes for another purpose. For Aspirations questions, which often have vague responses, or even broad categories of response as in the original 1957 survey given to the graduates, the allocation category corresponding to a Major Occupational group has been assigned if that is all the information we have, rather than coding the item as nonresponse. Thus, the WLS does have these allocation codes in the data. Only Wave 1 data contain any allocation codes (variables affected are: OCZ/OCZU, INZ/INZU, INX1/INX1U, OCR10, OCR10U, INSPHH, INWYAL, XOCZ, XOCR10, XINR10, OCNXYR1.) For use in the WLS, then, the SEI and PRESTIGE scores for Allocation codes should be the weighted average of all SEI/ PRESTIGE scores within the Major category corresponding to the Allocation code. For example, for 1970 Allocation code "196=Professional, Technical, and Kindred Workers", the SEI should be the weighted average of the SEIs for Occupation lines 001-195. Duncan provided the Mean SEI scores for the 1960 Major Occupation groups, which correspond to 11 of the 12 1970 Major Occupation Groups (Transport Equipment Operatives were included with all other Operatives in the 1960 Classification.) This table is duplicated in COR713c.ASC. These scores have been multiplied by 10 (to get 3 digits), then taken as-is to get scores for the 1970 Occupation Allocation categories. 1970 Allocation Code "726=Transport Equipment Operatives (Allocated)" gets the same Duncan SEI as "696=Operatives, Except Transport (Allocated)". The Siegel PRESTIGE scores for the 1970 Allocated Occupation categories come directly from Siegel's 1971 dissertation. Each 1960 Major Occupation category in Siegel's Table 5 has a PRESTIGE score for the entire Group. These scores were multiplied by 10 for use in the WLS. As above, "726= Transport Equipment Operatives (Allocated)" were assigned the same PRESTIGE score as "696=Operatives, Except Transport (Allocated)". These changes had previously been made to COR521C and COR612C; thus, no changes were requires for the SEI/PRESTIGE scores of the 1970 Allocation categories. (4) BIZARRE OCCUPATION CODES INCLUDED IN THE FSD PAPER WHICH DO NOT OCCUR AT ALL IN THE 1970 DETAILED OCCUPATION CLASSIFICATION SYSTEM At the end of FSD's list of Duncan SEI and Siegel PRESTIGE scores for the 1970 Occupational Classification System is an additional list of 15 1970 Occupation codes, with corresponding SEI/PRESTIGE scores. Eight of these 15 are discussed in (2) above--they are the 8 1970 Occupation categories that are not included in Tech. Paper 26. Code 280 refers to a generic "Salesmen n.e.c." category, which would combine 1970 categories 281-285. I am treating this code as a Split category, and provide details below (it DOES appear in the WLS data). 1970 Code 796 corresponds to an allocation category, detailed in part (3) above, although the SEI and PRESTIGE scores corresponding to 796 are not at all close to Duncan's and Siegel's scores for this Major category. The rest of the codes are a mystery. They do not correspond to any 1970 Occupation code--neither in Census documents nor in the WLS data. These codes (775, 782, 795, 805, 882) have already been removed from COR521C; no further action is needed. C. Problems with Split Occupation lines Duncan and Siegel both assigned different scores for large Occupation categories that could be differentiated by the Detailed Industry categories of their incumbents; Duncan also used Class-of-Worker to differentiate among some workers. The 1960 Detailed Occupations that Duncan "Split" by Industry and/or Class-of-Worker are: ****************************************************************************** * 1960 1970 Occupation CLASS- Code Code Title IND. OF-WRKR ---- ---- --------------------------------- ---- ------- 260 215 Inspectors, Public Administration X 270 222 Officials and Administrators, n.e.c., X Public Administration 290 245 Managers, Officials, and Proprietors, X X n.e.c. 394 280 Salesmen and Sales Clerks, n.e.c. X 430 441 Foremen, n.e.c. X 450 452 Inspectors, n.e.c. X 685 640 Mine Operatives and Laborers, n.e.c. X 775 690-695 Operatives and Kindred Workers, n.e.c. X 853 964 Policemen and Detectives X 973 780-785 Laborers, n.e.c. X ****************************************************************************** * and those that Siegel split by 1960 Detailed Industry are: ****************************************************************************** * 1960 1970 Occupation Code Code Title INDUSTRY ---- ---- --------------------------------- -------- 260 215 Inspectors, Public Administration X 270 222 Officials and Administrators, n.e.c., X Public Administration 290 245 Managers, Officials, and Proprietors, X n.e.c. 394 280 Salesmen and Sales Clerks, n.e.c. X 430 441 Foremen, n.e.c. X 685 640 Mine Operatives and Laborers, n.e.c. X 775 690-695 Operatives and Kindred Workers, n.e.c. X ****************************************************************************** * To assign the Duncan SEI and Siegel PRESTIGE scores to these Occupation lines, first a mapping between the 1960-basis Detailed Industry categories and the 1970-basis Detailed Industry categories had to be done. This mapping is provided in COR713i.XLS. Because we are using 3-digit SEI/PRESTIGE scores, both the Duncan and Siegel 1960-basis scores had to be multiplied by 10. Unfortunately, when Occupation lines are split by Industry and/ or Class-of-Worker, every possible Industry/Class-of-Worker is not always accounted for. This is a problem, because most of the time, real data often contains these Occupation/Industry/ Class-of-Worker combinations, if only because of missing data. For example, Duncan split the category "Inspectors, Public Administration" by (1)Federal and Postal Service, (2)State, and (3)Local public administration industries. No other categories are given. What to do if someone has missing data for Industry? In his mappings to 1960 Occupation/Industry categories (reproduced in COR713d.ASC), Duncan says that for this occupation, "all other industries and industry not specified" should receive the default SEI of 63, which is the Occupation-line total in his original list (COR713a.ASC). He also says that "Officials and Administrators" and "Mine Operatives" should be treated this way, also. Therefore, in the cases where Duncan provides default values for an Occupation line split by Industry and/or Class-of-Worker, this default will be used when there is no specific "All other industries" split. Furthermore, this definition will also be used for the Siegel PRESTIGE scores, and later in the Occupational Education and Occupational Income scores. The same problem occurs for the occupations split by Class-of- Worker. Here, Duncan was also specific, "When class of worker is unknown, code class of worker as Pr (Private wage and salary)." (Duncan 1961c, reproduced as COR713d.ASC.) This is what we will do for 245=Managers. But the Private/Self-Employed distinction doesn't work well for the 964=Police and Detectives, because this category distinguished between Government workers and others. In this case, we will use the weighted average of the entire category when Class-of-Worker is unknown for any reason. Six main types of problems occured in the SEI/PRESTIGE scores for these "Split" Occupation lines: (1) Wrong Duncan scores used for 1970 Occupations 780 and 785 splits; (2) the default scores used for missing data were sometimes incorrect; (3) the WLS does not use the Split scores for 1970 Occupations 215 and 222 (these are split by Federal, State, and Local Government workers); and (4) COR521c and COR612c did not correctly assign Public Administration Managers (1970 Ind=907-947) as "Other Industries" (Prestige=503). (1) WRONG DUNCAN SCORES USED FOR 1970 OCCUPATIONS 780 and 785 SPLITS The original problem that prompted this investigation is that the Duncan SEI scores used for the 780 and 785 (Misc. and NS Laborers) split lines are the exact same scores used for the 690, 692, 694, and 695 split lines (Misc. and NS Operatives). Laborers have much lower SEI scores than Operatives, so this error mistakingly inflates the scores of laborers. This error was introduced in COR315, and has been present throughout the use of the Duncan SEI scores in the WLS. PRESTIGE scores are not affected because Siegel did not split 780 and 785 by Industry. The proper scores to use for the 780/785 split lines come from Duncan's list of 1960-basis SEI scores, reproduced as COR713d.ASC. (2) DEFAULT SCORES USED FOR MISSING DATA SOMETIMES INCORRECT "Missing data" has two different meanings when assigning SEI/PRESTIGE scores to 1970-basis Occupation lines split by Industry and/or Class-of-Worker. In the first type, Occupation, Industry, and Class-of-Worker have all been *measured*, but an individual observation might have a missing data value for a particular variable. For example, a respondent replied "Don't Know" to the question about Industry placement. The second type of missing data for these Split lines occurs when Industry and/or Class-of-Worker is not even measured for an entire Occupation variable. This occurs in the WLS for several questions, usually questions that ascertain an Occupational Aspiration. The first type of missing data (individual instance) should be handled by assigning the "default" SEI/PRESTIGE score for the Split Occupation category, as originally provided by Duncan and Siegel. This is equivalent to assigning these cases the SEI/PRESTIGE score for "All Other Industries". If there is no "All Other Industries" category, then the default for the entire Occupation line shall be used, as detailed above (and see COR713d.ASC). The same is true for missing Class-of-Worker data. The second type of missing data (Industry and/or Class-of- Worker not measured) should NOT be handled as above. Cases such as this should not automatically receive the SEI/PRESTIGE score for "All Other Industries", because this generally is the lowest possible score for the Split Occupation line. Rather, an average score for the Split Occupation line should be used. *This* score is computed by FSD, and is the score in the Excel worksheet (COR713h.XLS). This line is the weighted average of all SEI/PRESTIGE scores within the Split Occupation category. If the Occupation line is split by Class-of-Worker (Duncan SEI only), then a WEIGHTED AVERAGE of the SEI score should be used, also. In the case where a 1970 Detailed Industry score is present, but a Class-of-Worker score is not, this presents a problem for 1970-basis Occupation '245=Managers and Administrators n.e.c.', because Duncan did NOT provide the default scores to be used in such a case. The question then becomes which Census (1950? 1960? 1970?) should we use to weight the two SEI scores within the Industry Split of '245=Managers and Administrators n.e.c.'? We decided to use the 1950 Census (males)--this is the Census upon which the Split scores were originally based, and because we are keeping those scores as-is for the Split Occupation lines, we will use those weights in computing the default scores that Duncan did not provide. Occupation line '964=Policemen and Detectives' was also split by Class-of- Worker (Government vs. Private); however, this is the only split, and so the default score for the entire line shall be used per Duncan 1961c (reproduced as COR713d.ASC). For the 12 1970-basis Allocation Occupation codes, the scores are found in COR713h.XLS--the reproduction of the FSD scores. For the weighted averages of Industry splits within 245=Managers (COW not measured), see COR713j.XLS. In COR713.SAS, the proper handling of missing data is ensured by passing a '-9' code for variables that are entirely missing Industry (OCASSK, OCEXSK, SPOCASP1) or Class-of-Worker (OCASSK, OCEXSK, OCZ, XOCZ, OCNXYR1, SPOCASP1). All other missing data is handled by passing the usual missing data scores ('997', '998', or '999' for Industry; '7', '8', or '9' for Class-of- Worker) before linking to modules in COR713.SAS. More detailed instructions are included in COR713.SAS. (3) THE WLS DOES NOT USE THE SPLIT SCORES FOR 1970 OCCUPATIONS 215 and 222. 1970 Occupation '215=Inspectors, exc. construction, Public Administration' and '222=Officials and Administrators, Public Administration n.e.c.' were NOT included in Hauser's list of Occupation lines split by Industry and/or Class-of- Worker in COR315. The decision was made to not distinguish among Federal (and Postal), State, and Local government workers in these Occupations because the WLS did not collect the data to do so. Taking a look at the WLS data, however, reveals that enough detail WAS obtained to distinuish among these types of government workers, and so we should include these distinctions when assigning SEI/PRESTIGE scores to the WLS data. These two lines do not have a default "All Other Industries" category, and therefore do not provide a category for missing Industry cases, nor cases with Industries other than 907-937 (in particular, 947). In this case, the recommendation of Duncan 1961c (reproduced as COR713d.ASC) shall be used, and the default for the entire category will be assigned in such cases (DUNCANSEI=630 for Occupation=215 and DUNCANSEI=660 for Occupation=222). These Splits have been added to COR713.SAS. (4) COR521C AND COR612C DID NOT CORRECTLY ASSIGN PUBLIC ADMINISTRATION MANAGERS (1970 Occ=245, 1970 Ind=907-947) AS "OTHER INDUSTRIES". Although it may be argued that Managers in Public Administration industries should be coded as 222 rather than 245, the WLS nevertheless contains many cases where a 245 Manager has a 907-947 Public Administration Industry. These cases should be included under both Duncan's and Siegel's "All Other Industries" designations. The Industries 907-947 were left out of COR315 entirely, and so anyone with these Industries was assigned the default score for the entire 245 category (Salaried), rather than the "Other Industries, Salaried" score. The Duncan SEI scores were assigned correctly merely by accident, because the default score is the same as the "Other Industries" score (620) for salaried workers. If there were any self-employed Public Administration (907-947) workers, they would have INCORRECTLY been assigned 620, and should have been assigned 490. The Siegel scores are not assigned correctly. The default PRESTIGE for Salaried Managers is 508 and that for "All Other Industries" (Salaried Managers) is 503. This error is corrected in COR713.sas. (5) DEFAULT FOR 640=MINE OPERATIVES NOT ASSIGNED PROPERLY. In Duncan 1961c (reproduced as COR713d.ASC), it clearly states that for the category "Mine Operatives and Laborers, n.e.c.", "All other industries...and industry not specified" shall receive the default SEI of 10. This default score is the same as that for the 1950 Category in Duncan 1961a (reproduced as COR713a.ASC). In COR521C and COR612C, the default score used is 7, not 10. In addition, the default score for this category in Siegel 1971 (reproduced as COR713e.ASC) is 263, not 264 as appears in COR521C and COR612C. These changes are included in COR713.SAS. (6) SCORES FOR 1970-BASIS INDUSTRY ALLOCATION CODES. Finally, the 1970 Detailed Industry categories contain 14 Allocation codes, similar to the 12 Allocated Occupation codes discussed above. The WLS does use some of these codes, and again, the WLS uses them to indicate a response in the MAJOR INDUSTRY category, rather than a hotdeck allocation of data for missing responses. Therefore, we should be sure that each of the 14 Allocated Industry codes is included in any Occupation line split by Industry. To assign a score for these split lines, the following decision rules apply: a. If the entire major category is the split, for example all Construction Industries (067-077) are one split line for Occupation "441=Foremen n.e.c.", then the Industry Allocation code (in this example, 078 for Construction Industries) shall receive the same score as for the entire split line. b. If the Major Industry category corresponding to the split Occupation line is itself split (as, for example, Retail Trade (607-698) within Managers n.e.c. (245), AND a default score for that Major Industry category was provided by Duncan or Siegel, then use the default provided by Duncan 1961a (reproduced as COR713a.ASC) and/or Siegel 1971 (reproduced as COR713e.ASC). c. If the Major Industry category corresponding to the split Occupation line is itself split (as, for example, Durable Goods Manufacturing Industries (107-259) in Occupation "441=Foremen n.e.c."), then the score for the Allocated Industry category (in this example, 267 for Manufacturing, Durable Goods) shall be the weighted average of the Occupation/Industry splits for that Major Industry Category. These scores shall be weighted by the 1950 Census for Duncan SEI, and by the 1960 Census for Siegel PRESTIGE. They are computed in COR713j.XLS. d. If the Major Industry category is split within an Occupation category (as above), AND it is the ONLY Major Industry category included in the Occupation split line (that is, there is no "All Other Industries" category within the Split Occupation line), then the score for the Allocated Industry category shall be the default score for the entire Occupation line, as designated in Duncan 1961c (reproduced as COR713d.ASC), and Siegel 1971 (reproduced as COR713e.ASC). D. FINAL LIST OF DUNCAN SEI/SIEGEL PRESTIGE SCORE CHANGES 1970-Basis Category Duncan SEI Siegel PRESTIGE ------------------- -------------------- -------------------- Occ. Ind. COW Old New Err. Old New Err. ---- ---- --- --- --- ---- --- --- ---- 092 714 656 B2 094 714 656 B2 131 720 840 B2 532 783 B2 156 530 620 B2 468 470 B2 180 518 519 B1 215 <=899 667 630 C3 423 406 C3 215 907 667 720 C3 423 433 C3 215 917 667 720 C3 423 433 C3 215 927 667 540 C3 423 416 C3 215 937 667 560 C3 423 384 C3 215 947 667 630 C3,C6d 423 416 C3,C6d 215 999** 667 630 C3 423 406 C3 222 <=899 673 660 C3 607 606 C3 222 907 673 840 C3 607 645 C3 222 917 673 840 C3 607 645 C3 222 927 673 660 C3 607 620 C3 222 937 673 540 C3 607 598 C3 222 947 673 660 C3,C6d 607 620 C3,C6d 222 999** 673 660 C3 607 606 C3 245 499 1-3 760 729 C6c 245 499 9* 760 729 C6c 245 499 4-6 490 431 C6c 245 699 1-3 590 560 C6b 245 699 9* 590 560 C6b 245 699 4-6 440 430 C6b 245 699 412 416 C6b 245 719 1-3 840 846 C6c 245 719 9* 840 846 C6c 245 719 4-6 760 791 C6c 245 719 508 637 C6c 245 767 1-3 530 624 C6c 245 767 9* 530 624 C6c 245 767 4-6 340 433 C6c 245 767 290 443 C6c 245 907-947 4-6 620 490 C4 245 907-947 508 503 C4 245 -9 -9 620 564 C2,C6c 245 <=058 -9 555 566 C2 245 067-078 -9 555 537 C2 245 107-399 -9 700 724 C2 245 407-429 -9 570 613 C2 245 447-479 -9 600 731 C2 245 499 -9 600 648 C2,C6b 245 507-599 -9 645 641 C2 245 607-608 -9 625 620 C2 245 609-627 -9 575 570 C2 245 628-638 -9 415 367 C2 245 639-647 -9 675 676 C2 245 648 -9 320 326 C2 245 649 -9 675 676 C2 245 657-658 -9 670 663 C2 245 667-668 -9 635 616 C2 245 669 -9 380 374 C2 245 677-698 -9 540 517 C2 245 699 -9 540 465 C2,C6c 245 717-718 -9 800 808 C2 245 719 -9 800 832 C2,C6c 245 727-748 -9 735 727 C2 245 749-757 -9 415 391 C2 245 758-759 -9 435 364 C2 245 767 -9 435 491 C2,C6c 245 769-799 -9 455 437 C2 245 807-899 -9 555 566 C2 245 907-947 -9 620 566 C2,C6c 245 999** -9 620 566 C2 260 423 424 B1 441 058 497 380 C6a 441 267 497 545 C6c 441 399 497 493 C6c 441 499 497 445 C6c 452 058 412 380 C6a 452 399 412 380 C6a 452 499 412 420 C6c 475 270 360 B2 408 339 B2 563 211 210 B1 640 <029 070 100 C5 264 263 C5 640 029 165 100 C6a 264 263 C5,C6a 640 058 165 100 C5,C6d 264 263 C5,C6d 640 067-947 070 100 C5 264 263 C5 640 999** 070 100 C5 264 263 C5 672 283 282 B1 690 058 190 200 C6a 284 323 C6a 690 267 190 184 C6c 284 290 C6b 690 399 190 163 C6c 284 275 C6b 690 499 190 181 C6c 284 351 C6c 690 699 200 170 C6a 690 719 190 200 C6a 284 323 C6a 690 799 190 110 C6a 284 259 C6a 690 899 190 200 C6a 284 323 C6a 690 947 190 170 C6a 284 334 C6a 692 058 193 200 C6a 293 323 C6a 692 267 193 184 C6c 293 290 C6b 692 399 193 163 C6c 293 275 C6b 692 499 193 181 C6c 293 351 C6c 692 699 193 170 C6a 692 719 193 200 C6a 293 323 C6a 692 799 193 110 C6a 293 259 C6a 692 899 193 200 C6a 293 323 C6a 692 947 193 170 C6a 293 334 C6a 694 058 192 200 C6a 291 323 C6a 694 267 192 184 C6c 291 290 C6b 694 399 192 163 C6c 291 275 C6b 694 499 192 181 C6c 291 351 C6c 694 699 192 170 C6a 694 719 192 200 C6a 291 323 C6a 694 799 192 110 C6a 291 259 C6a 694 899 192 200 C6a 291 323 C6a 694 947 192 170 C6a 291 334 C6a 695 058 192 200 C6a 291 323 C6a 695 267 192 184 C6c 291 290 C6b 695 399 192 163 C6c 291 275 C6b 695 499 192 181 C6c 291 351 C6c 695 699 192 170 C6a 695 719 192 200 C6a 291 323 C6a 695 799 192 110 C6a 291 259 C6a 695 899 192 200 C6a 291 323 C6a 695 947 192 170 C6a 291 334 C6a 764 086 076 B1 185 175 B1 780 <058 200 060 C1 780 058 082 060 C1,C6a 780 067-078 180 070 C1 780 107-108 070 030 C1 780 109 090 020 C1 780 118 090 050 C1 780 119 230 140 C1 780 127-128 100 050 C1 780 137 210 070 C1 780 138 150 050 C1 780 139 170 090 C1 780 147 120 040 C1 780 148-149 150 060 C1 780 157-158 160 070 C1 780 159 150 100 C1 780 167-168 150 100 C1 780 169 140 090 C1 780 177 220 100 C1 780 178 210 140 C1 780 179 220 100 C1 780 187 220 100 C1 780 188-189 321 170 C1 780 197-198 220 100 C1 780 199-209 260 140 C1 780 219 210 130 C1 780 227 340 150 C1 780 228 160 020 C1 780 229-238 230 080 C1 780 239 230 100 C1 780 247 230 100 C1 780 248 400 160 C1 780 249 280 110 C1 780 257 230 100 C1 780 258 150 100 C1 780 259 160 120 C1 780 267 082 072 C1,C6c 780 268 160 080 C1 780 269 220 130 C1 780 278 090 060 C1 780 279 140 060 C1 780 287 150 100 C1 780 288 120 100 C1 780 289 190 160 C1 780 297 110 050 C1 780 298 190 140 C1 780 299 020 000 C1 780 307 210 040 C1 780 308 080 090 C1 780 309 140 140 C1 780 317 020 010 C1 780 318 100 060 C1 780 319 220 110 C1 780 327 170 060 C1 780 328 190 060 C1 780 329 190 080 C1 780 337 170 100 C1 780 338-339 310 230 C1 780 347-348 230 080 C1 780 349 090 040 C1 780 357 260 220 C1 780 358 230 080 C1 780 359 150 080 C1 780 367-369 230 080 C1 780 377 560 260 C1 780 378 140 030 C1 780 379 220 120 C1 780 387 220 120 C1 780 388 100 020 C1 780 389 090 100 C1 780 397 140 120 C1 780 398 160 080 C1 780 399 082 091 C1,C6c 780 407 150 030 C1 780 408-429 230 090 C1 780 447-479 210 060 C1 780 499 082 050 C1,C6c 780 507-698 170 120 C1 780 699 082 120 C1,C6a 780 707-718 200 060 C1 780 719 082 060 C1,C6a 780 727-759 190 090 C1 780 767 082 090 C1,C6a 780 769-798 110 050 C1 780 799 082 050 C1,C6a 780 807-897 200 060 C1 780 899 082 060 C1,C6a 780 907-937 170 070 C1 780 947 082 070 C1,C6a 780 999** 200 060 C1 785 <058 200 060 C1 785 058 083 060 C1,C6a 785 067-078 180 070 C1 785 107-108 070 030 C1 785 109 090 020 C1 785 118 090 050 C1 785 119 230 140 C1 785 127-128 100 050 C1 785 137 210 070 C1 785 138 150 050 C1 785 139 170 090 C1 785 147 120 040 C1 785 148-149 150 060 C1 785 157-158 160 070 C1 785 159 150 100 C1 785 167-168 150 100 C1 785 169 140 090 C1 785 177 220 100 C1 785 178 210 140 C1 785 179 220 100 C1 785 187 220 100 C1 785 188-189 321 170 C1 785 197-198 220 100 C1 785 199-209 260 140 C1 785 219 210 130 C1 785 227 340 150 C1 785 228 160 020 C1 785 229-238 230 080 C1 785 239 230 100 C1 785 247 230 100 C1 785 248 400 160 C1 785 249 280 110 C1 785 257 230 100 C1 785 258 150 100 C1 785 259 160 120 C1 785 267 083 072 C1,C6c 785 268 160 080 C1 785 269 220 130 C1 785 278 090 060 C1 785 279 140 060 C1 785 287 150 100 C1 785 288 120 100 C1 785 289 190 160 C1 785 297 110 050 C1 785 298 190 140 C1 785 299 020 000 C1 785 307 210 040 C1 785 308 080 090 C1 785 309 140 140 C1 785 317 020 010 C1 785 318 100 060 C1 785 319 220 110 C1 785 327 170 060 C1 785 328 190 060 C1 785 329 190 080 C1 785 337 170 100 C1 785 338-339 310 230 C1 785 347-348 230 080 C1 785 349 090 040 C1 785 357 260 220 C1 785 358 230 080 C1 785 359 150 080 C1 785 367-369 230 080 C1 785 377 560 260 C1 785 378 140 030 C1 785 379 220 120 C1 785 387 220 120 C1 785 388 100 020 C1 785 389 090 100 C1 785 397 140 120 C1 785 398 160 080 C1 785 399 083 091 C1,C6c 785 407 150 030 C1 785 408-429 230 090 C1 785 447-479 210 060 C1 785 499 083 050 C1,C6c 785 507-698 170 120 C1 785 699 083 120 C1,C6a 785 707-718 200 060 C1 785 719 083 060 C1,C6a 785 727-759 190 090 C1 785 767 083 090 C1,C6a 785 769-798 110 050 C1 785 799 083 050 C1,C6a 785 807-897 200 060 C1 785 899 083 060 C1,C6a 785 907-937 170 070 C1 785 947 083 070 C1,C6a 785 999** 200 060 C1 964 -9 380 405 C2 964 9* 340 390 C2 * Indicates individual missing data for COW; i.e., cow^=-9 and (cow<1 or cow>6). ** Indicates individual missing data for IND; i.e., ind^=-9 and (ind<1 or ind>947). E. WHAT DIFFERENCE DOES IT MAKE? Following are the old and new Means and Standard Deviations for some selected, important variables using the old and new Duncan SEI and Siegel PRESTIGE scores: Duncan SEI SiegelPREST -------------- -------------- Variable Wave Old New Old New ------------- ---- ---- ---- ---- ---- OCHH57U 1 346.4 344.9 398.4 398.4 (Head's Occ,R75) (231.3) (232.3) (118.9) (118.9) OCMO57 1 328.3 327.6 338.8 338.8 (Mom's Occ,R75) (198.2) (198.4) (124.2) (124.2) OCSIB 1 456.9 456.3 431.9 431.9 (Sib's Occ,R75) (235.9) (236.9) (136.8) (137.0) OCZU 1 556.2 556.3 515.2 515.3 (R's HS Asp,R75) (216.9) (217.3) (124.8) (124.9) OCX1U 1 431.7 431.2 402.2 402.2 (R's 1st Job,R75) (230.5) (231.3) (142.7) (142.8) OCX70 1 486.8 486.5 446.2 446.2 (R's 1970 Job,R75) (232.7) (233.2) (136.9) (137.2) OCR74 1 479.0 478.9 437.4 437.6 (R's 1974 Job,R75) (232.3) (232.8) (137.8) (138.2) OCXCURU 1 476.3 476.1 435.4 435.5 (R's c/l Job,R75) (230.5) (231.1) (137.8) (138.1) OCR10 1 521.7 522.1 479.2 479.5 (R's 1975 Asp,R75) (213.0) (213.9) (125.1) (125.6) SPOCASP1 1 510.8 508.6 461.2 461.3 (R's HS Asp,R57) (216.5) (215.7) (132.2) (132.4) XOCHH57 1 328.8 327.3 391.8 391.8 (Head's Occ,S77) (224.9) (225.5) (117.9) (117.9) XOCZ 1 540.3 539.9 515.2 515.5 (S's HS Asp,S77) (227.1) (228.9) (127.7) (127.9) XOCX1 1 407.2 405.8 390.3 390.3 (S's 1st Job,S77) (234.6) (236.2) (144.7) (144.7) XOCX70 1 447.4 446.6 423.2 423.0 (S's 1970 Job,S77) (232.7) (233.4) (137.8) (137.9) XOCR76 1 469.3 469.0 432.9 432.9 (S's 1976 Job,S77) (233.3) (234.6) (137.9) (138.1) XOCXCUR 1 466.6 466.2 429.3 429.4 (S's c/l Job,S77) (231.0) (232.5) (137.8) (138.0) XOCR10 1 516.4 516.8 474.9 475.1 (S's 1977 Asp,S77) (219.1) (220.3) (124.0) (124.3) RFU06J1C 2 479.3 479.3 436.6 436.7 (R's 1975 Job,R92) (227.9) (228.5) (137.5) (137.9) RFU34JCE 2 498.6 498.6 445.7 445.7 (R's c/l Job,R92) (227.9) (228.4) (136.6) (136.8) RKU39SSF 2 464.5 463.9 432.8 432.8 (S's 1992 Job,R92) (236.2) (237.3) (139.7) (139.9) RFU03JAF 2 529.3 525.1 474.1 474.2 (R's 1975 Asp,R92) (228.4) (227.8) (133.8) (134.2) SFU06REC 3 438.8 438.2 414.2 414.4 (S's 1st Job,S94) (240.2) (241.3) (146.5) (146.7) SFU29REC 3 472.7 472.9 436.2 436.4 (S's 1977 Job,S94) (236.7) (237.7) (138.0) (138.4) SFU53RED 3 488.0 488.0 441.1 441.3 (S's c/l Job,S94) (233.4) (234.2) (139.1) (139.3) Duncan SEI -------------- Head's Occ (R75), OCHH57U by: Variable Wave Old New Diff ----------------------------- ------------- ---- ----- ----- ----- R's 1st Job OCX1U 1,1 .2652 .2646 -.0006 R's c/l Job(R75) OCXCURU 1,1 .2705 .2698 -.0007 R's 1975 Job(R92) RFU06J1C 1,2 .2530 .2525 -.0005 R's c/l Job(R92) RFU34JCE 1,2 .2372 .2366 -.0006 Sib's 1975 Job(R75), OCSIB by: ------------------------------ R's 1975 Job (R75) OCXCURU 1,1 .2280 .2275 -.0005 S's 1977 Job (S77) XOCXCUR 1,1 .7929 .7912 -.0017 R's 1975 Job(R92) RFU06J1C 1,2 .2192 .2194 +.0002 R's c/l Job(R92) RFU34JCE 1,2 .2148 .2152 +.0004 S's 1992 Job(R92) RKU39SSF 1,2 .6612 .6625 +.0013 S's 1977 Job(S94) SFU29REC 1,3 .7747 .7737 -.0010 S's c/l Job(S94) SFU53RED 1,3 .6367 .6378 +.0011 R's First Job (R75), OCX1U by: ------------------------------ R's c/l Job (R75) OCXCURU 1,1 .5466 .5465 -.0001 S's First Job (S77) XOCX1 1,1 .2816 .2825 +.0009 R's 1975 Job (R92) RFU06J1C 1,2 .5362 .5358 -.0004 R's c/l Job (R92) RFU34JCE 1,2 .4919 .4913 -.0006 S's First Job (S94) SFU06REC 1,3 .2250 .2253 +.0003 R's C/L Job (R75), OCXCURU by: ------------------------------ S's 1977 Job (S77) XOCXCUR 1,1 .2253 .2244 -.0009 R's 1975 Job (R92) RFU06J1C 1,2 .8084 .8080 -.0004 R's c/l Job (R92) RFU34JCE 1,2 .6327 .6335 +.0008 S's 1977 Job (S94) SFU29REC 1,3 .2199 .2198 -.0001 Head's Job (S77), XOCHH57 by: ----------------------------- S's First Job (S77) XOCX1 1,1 .2696 .2681 -.0015 S's c/l Job (S77) XOCXCUR 1,1 .2266 .2274 +.0008 S's First Job (S94) SFU06REC 1,3 .2631 .2636 +.0005 S's 1977 Job (S94) SFU29REC 1,3 .2127 .2137 +.0010 S's c/l Job (S94) SFU53RED 1,3 .2138 .2130 -.0008 S's First Job (S77), XOCX1 by: ------------------------------ S's c/l Job (S77) XOCXCUR 1,1 .6187 .6169 -.0018 S's First Job (S94) SFU06REC 1,3 .9352 .9343 -.0009 S's 1977 Job (S94) SFU29REC 1,3 .6296 .6270 -.0026 S's c/l Job (S94) SFU53RED 1,3 .5103 .5092 -.0011 S's C/L Job (S77), XOCXCUR by: ------------------------------ S's 1977 Job (S94) SFU29REC 1,3 .9976 .9976 - 0 - S's c/l Job (S94) SFU53RED 1,3 .6346 .6354 +.0008 R's 1975 Job (R92), RFU06J1C by: -------------------------------- R's c/l Job (R92) RFU34JCE 2,2 .6999 .6998 -.0001 S's 1977 Job (S94) SFU29REC 2,3 .2116 .2113 -.0003 R's C/L Job (R92), RFU34JCE by: ------------------------------- S's 1992 Job (R92) RKU39SSF 2,2 .2041 .2037 -.0004 S's c/l Job (S94) SFU53RED 2,3 .1948 .1957 -.0003 S's First Job (S94), SFU06REC by: --------------------------------- S's 1977 Job (S94) SFU29REC 3,3 .6503 .6486 -.0017 S's c/l Job (S94) SFU53RED 3,3 .5807 .5796 -.0011 S's 1977 Job (S94), SFU29REC by: -------------------------------- S's c/l Job (S94) SFU53RED 3,3 .6986 .6990 +.0004 R's 1957 Asps (R57), SPOCASP1 by: --------------------------------- R's 1957 Asps (R75) OCZU 1,1 .6516 .6550 +.0034 R's First Job (R75) OCX1U 1,1 .4645 .4714 +.0069 R's c/l Job (R75) OCXCURU 1,1 .4456 .4495 +.0039 R's 1975 Job (R92) RFU06J1C 1,2 .4432 .4467 +.0035 R's c/l Job (R92) RFU34JCE 1,2 .3904 .3941 +.0037 R's 1957 Asps (R75), OCZU by: ----------------------------- R's 1975 Asps (R75) OCR10 1,1 .4569 .4569 - 0 - S's 1957 Asps (S77) XOCZ 1,1 .1807 .1818 +.0011 R's First Job (R75) OCX1U 1,1 .4957 .4948 -.0009 R's c/l Job (R75) OCXCURU 1,1 .4546 .4529 -.0017 R's 1975 Job (R92) RFU06J1C 1,2 .4448 .4438 -.0010 R's c/l Job (R92) RFU34JCE 1,2 .4003 .3982 -.0021 R's 1975 Asps (R75), OCR10 by: ------------------------------ R's 1975 Asps (R92) RFU03JAF 1,2 .6483 .6521 +.0038 S's 1975 Asps (S77) XOCR10 1,1 .1601 .1588 -.0013 R's c/l Job (R75) OCXCURU 1,1 .6987 .6992 +.0005 R's 1975 Job (R92) RFU06J1C 1,2 .6055 .6050 -.0005 R's c/l Job (R92) RFU34JCE 1,2 .5297 .5286 -.0011 R's 1975 Asps (R92), RFU03JAF by: --------------------------------- R's c/l Job (R75) OCXCURU 1,2 .7291 .7290 -.0001 R's 1975 Job (R92) RFU06J1C 2,2 .8559 .8540 -.0019 R's c/l Job (R92) RFU34JCE 2,2 .7509 .7488 -.0021 S's 1957 Asps (S77), XOCZ by: ----------------------------- S's 1975 Asps (S77) XOCR10 1,1 .4131 .4148 +.0027 S's First Job (S77) XOCX1 1,1 .4532 .4505 -.0027 S's c/l Job (S77) XOCXCUR 1,1 .3999 .4007 +.0008 S's First Job (S94) SFU06REC 1,3 .4537 .4542 +.0005 S's 1977 Job (S94) SFU29REC 1,3 .4426 .4436 +.0010 S's c/l Job (S94) SFU53RED 1,3 .3426 .3466 +.0040 S's 1977 Asps (S77), XOCR10 by: ------------------------------- S's c/l Job (S77) XOCXCUR 1,1 .6658 .6618 -.0040 S's 1977 Job (S94) SFU29REC 1,3 .6787 .6752 -.0035 S's c/l Job (S94) SFU53RED 1,3 .5230 .5216 -.0014 In general, the means for Duncan SEI tend to decrease slightly and the standard deviations increase slightly, EXCEPT FOR ASPIRATIONS VARIABLES. Aspirations variables have different patterns. The means and standard deviations for Siegel Prestige tend to remain steady or both slightly increase, with very few exceptions. The correlations were done only for Duncan's SEI (since Prestige is not much used in most models of status attainment.) In general, the differences between the old and new measures are slight, EXCEPT FOR ASPIRATIONS. Some general patterns include: - The correlation between Head's Occ and R's Occ DECREASES SLIGHTLY - The corr. between R's First Job and any later jobs DECREASES SLIGHTLY - The corr. between S's First Job and any later jobs DECREASES SLIGHTLY - Aspirations show by far the greatest changes... - The corr. between original 1957 aspirations and any of R's later jobs INCREASE MODERATELY - The corr. between retrospective 1957 aspirations and any of R's later jobs DECREASE SLIGHTLY to MODERATELY - The corr. between retrospective high school aspirations (siblings) and any of S's later occupations INCREASES MODERATELY - The corr. between sibling's mid-life aspiration and later jobs tends to DECREASE MODERATELY *************************************************************************** *** I V . M S E I 2 / T S E I 2 *** *************************************************************************** The MSEI2 and TSEI2 scores are found in the Appendix to Stevens and Featherman's 1981 article, "A Revised Socioeconomic Indix of Occupational Status." These scores DO NOT CONTAIN SPLITS--so they are very easy to check! The scores were typed into COR512, before being edited into code used to assign these scores to the Waves 2 and 3 data. MSEI2 and TSEI2 are NOT assigned to the Wave 1 data, excepting the 1970-basis aspirations data recoded in 1994 per COR617. Three errors were found in the MSEI2 and TSEI2 scores: (1)Assignment of MSEI2 and TSEI2 scores to 1970 Occupation '580=Former Members of the Armed Services'; (2)Error in TSEI2 score for 1970 Occupation '572=Not Specified Apprentices'; and (3)want to use weighted averages for Allocated Occupation categories--not the MSEI2/TSEI2 computed from the 1970 Census data. (1) ASSIGNMENT OF MSEI2 AND TSEI2 SCORES TO 1970 OCCUPATION '580=FORMER MEMBERS OF THE ARMED SERVICES' Although Duncan originally assigned an SEI score to "Members of the Armed Forces" (1960 Occupation 555), the WLS has never used this score, as outlined in MEMO122.ASC. (Siegel never assigned a PRESTIGE score for this category). Stevens and Featherman *did* assign MSEI2 and TSEI2 scores to this Occupation category, and these scores (MSEI2=2821; TSEI2=3247) were used in COR512, and copied to COR521C. Because we are not assigning SEI and PRESTIGE scores to this category (per MEMO122), we should not be assigning MSEI2 and TSEI2 scores, either. (2) ERROR IN TSEI2 SCORE FOR 1970 OCCUPATION '572=NOT SPECIFIED APPRENTICES' This is a case of an unclear footnote. In Stevens and Featherman's list of TSEI2 scores in their appendix, they do not give scores for 1970 Occupations '572=Not Specified Apprentices' and '575=Craftsmen and Kindred Workers, n.e.c.' Rather, they provide a footnote which reads (S&F, p.30): "Source tapes provided by the U.S. Census did not include data for these two titles. Consequently, updated scores could not be computed. See Hauser and Featherman (1977, p. 326) for approximations: 572:39.00 (MSEI2); 575:25.70 (MSEI2)/25.30 (TSEI2)." Looking in H&F 1977 (and also FSD as well as my own Excel spreadsheet with the FSD scores, COR713h.XLS) I can verify the following scores: 1970 Male Total Occup. SEI SEI ------ ----- ----- 572 39.0 <> 575 25.7 25.3 Recall that in H&F, as well as FSD, the TOTAL-BASED scores are only reported if they differ from the MALE-BASED scores. The S&F footnote failed to mention that the TSEI2 score, therefore, should be the same as the MSEI2 score of 39.0 for 1970 Occupation 572 (the scores for 575 are correct). When the S&F list was typed in COR512, the TSEI2 score '9999' was entered for Occ=572. When COR512 was turned into computer code used to map TSEI2 and MSEI2 scores for the Waves 2 and 3 data, this '9999' score was somehow turned into '9900'. COR713.SAS will list the TSEI2 score for 1970 Occupation 572 as 3900, not 9900. (Surprisingly, this does affect at least one case!) (3) WANT TO USE WEIGHTED AVERAGES FOR ALLOCATED OCCUPATION CATEGORIES When Stevens and Featherman created the MSEI2/TSEI2 scores, they just used the Occupational Education and Income scores for the 12 Allocated Occupation categories as they appeared in the Nam-Powers data. Because the WLS uses these categories to correspond to an occupation in the Major group--rather than a random assignment for missing data--we should not use these scores (which are based on the population of people with missing occupation data), and should instead use a weighted average of MSEI2/TSEI2 for the entire Major Occupation category corresponding to the Allocated code. MSEI2 should be weighted using only men in the Nam-Powers tabulations; TSEI2 should be weighted using all workers in the ECLF. COR713j.XLS contains the calculations of the weighted MSEI2/TSEI2 scores. A. FINAL LIST OF MSEI2/TSEI2 SCORE CHANGES 1970-Basis Category MSEI2 TSEI2 ------------------- -------------------- -------------------- Occ. Old New Err. Old New Err. ---- ---- ---- ---- ---- ---- ---- 196 5671 6853 (3) 5675 6665 (3) 246 4107 5310 (3) 4007 5064 (3) 296 3329 4435 (3) 2867 3721 (3) 396 2965 3275 (3) 2719 2913 (3) 572 9900 3900 (2) 580 2821 9999 (1) 3247 9999 (1) 586 2249 2615 (3) 2388 2637 (3) 696 1801 1849 (3) 1797 1839 (3) 726 1794 2020 (3) 1989 2161 (3) 796 1637 1625 (3) 1906 1902 (3) 806 2160 2263 (3) 2352 2425 (3) 846 1441 1429 (3) 1741 1733 (3) 976 1951 2093 (3) 1964 2102 (3) 986 1335 1398 (3) 1381 1553 (3) B. WHAT DIFFERENCE DOES IT MAKE? In general VERY FEW WLS VARIABLES had any changes at all in their MSEI2/TSEI2 scores. The only ones that changed were those with many Allocated Codes. The following list are the only three variables with changes in their mean MSEI2 and/or TSEI2 scores: MSEI2 TSEI2 ---------------- ---------------- Variable Wave Old New Old New ------------- ---- ---- ---- ---- ---- SPOCASP1 1 4322.2 4321.4 (1957 aspirations-R57) (2046.6) (2045.6) OCNXYR1 1 2941.2 2941.3 2576.1 2563.8 (1957 Expectation-R57) (1153.4) (1153.4) (866.1) (801.3) RJU24SK 2 3857.3 3856.4 (Asps for Sel.Kid-R92) (1979.1) (1977.7) *************************************************************************** *** V . O C C U P A T I O N A L E D U C A T I O N / *** *** O C C U P A T I O N A L I N C O M E S C O R E S *** *************************************************************************** Finally, I checked the scores we are using for 1970 Occupational Education and Occupational Income percentages. The list of these scores provided by Warren (COR638) originally was only reported to two digits, and we wanted to use 3-digit scores in order to be able to reproduce the *2* digit started logit transformations of these percentages used in several published articles (e.g., Hauser and Warren 1997; Warren, Sheridan and Hauser 1998; Hauser, Sheridan and Warren 1999). When I re-ran the SPSS code provided in COR638 to produce Occupational Education and Occupational Income scores rounded to three rather than two digits (COR684), I did not have access to the ORIGINAL documentation provided by Nam and Powers (at first--I received it later.) One will see upon reading the list of 871 Occupation lines in the Nam-Powers data that many Occupation lines in the Crafts and Operatives Major categories are out of order. Because I did not have this list, nor were there Occupation/ Industry codes attached to the 871 individual lines of output, all I could do was match the TOTAL N lines to the TOTAL N lines in Warren's original spreadsheet (COR684)!! I did some checking to make sure I lined up the Occupations correctly at the time, but because we wanted this check of the 1970-basis scores to be the FINAL, DEFINITIVE WORD, I did some re-checking just to be certain. First, I re-wrote the SPSS code into SAS code, and only included the TOTAL scores, rather than splitting by race, sex, and hours/weeks worked. I also eliminated any other definitions of Occupational Education and Income that were coded in the original document. The SAS programs I used, and the much smaller ASCII data files containing only the TOTAL N's and percentages can be found in COR713k.TAR. Details regarding the averaged lines (for Allocated Occupation and Industry categories) can be found in COR713j.XLS. (Recall that original Nam-Powers data files are included in COR638.) Second, I checked the scores used for 1970 Allocation codes. Recall that these are 12 Occupations codes that the Census uses to ALLOCATE an Occupation code for those with nonresponse on this item. These 12 codes *ARE* included in the Nam-Powers data. The problem with using those scores as-is for WLS purposes is that the Occupational Education and Income scores are actually computed using the cases where the Census allocated an Occupation based on nonresponse. The WLS uses these Allocation codes to indicate an observation that cannot be coded into any more detail than a Major Occupation category (occurs for Aspirations). So, the scores used for the Allocation categories should be weighted averages of the ENTIRE Major Occupation category, rather than the few individuals with nonresponse. These averages are computed in COR713j.XLS. Third, the Occupation lines that are split by Industry and/or Class-of-Worker in the Nam-Powers data should have default scores for variables that do not have an Industry and/or Class-of-Worker defined. These should be weighted averages of the entire category, much like the Allocated codes above should be weighted averages of the scores for the entire Major Occupation category. These avarages are computed in COR713j.XLS. Fourth, I checked that the Industry/Class-of-Worker Split line definitions used in the WLS code (COR684) matched those indicated in the Nam-Powers documentation. Ten main types of errors were found. (1) The first type were minor rounding errors. These were caused because COR684 outputs the scores to 4 digits, and then rounded to 3. This can cause the 3rd digit to be higher than it should be (for example, an Occupational Education score such as .79745 would have been output as .7975, then rounded to .798, when in fact it should have been rounded to .797.) (2) The second type of error occurred for "default" Occupation lines (either Split lines or Allocation codes, where a weighted average of Occupational Education/Income scores was computed.) In three places, an Occupation line was inadvertently left out of the demoninator. For Occupation 296, the N for Occupation 280 was included in the calculation of both Occupational Education and Occupational Income for '296=Sales Workers-- Allocated'. Because 280 is itself a "default" line that is a weighted average of the scores for 281-285, it should not have been included. Occupation '983=Laundresses, Private Household' is split by whether the Laundress lives in or out of the household. The very small number of Laundresses who 'live in' was inadvertently left out of the calculation of Occupational Education and Occupational Income; however, this ommission only affected the Occupational Education score. (3) The third type of error occurred because Class-of-Worker categories were switched or mislabelled. (This occurred for Occupational Income scores only.) For 1970 Occupation '245=Managers and Administrators n.e.c.', Industry Splits 727-767 and 669-698 had their Salaried scores in the Self- Employed slot, and vice versa. (4) The fourth type of error occurred because the original code in COR638 did not included the Class-of-Worker designation of '6' (unpaid family worker, not further specified), and those people with COW=6 were treated as having individual missing data, rather than a "self-employed" designation as they should have. This only affects those in occupations split by class-of-worker --specifically occupations 245, 142, 143, 144, and 964. (5) The fifth type of changes to be made are for missing data cases (individual-type missing data, as defined in the Duncan SEI/Siegel PRESTIGE section III above, error type C2.) When an individual observation was missing an Industry and/or Class-of-Worker code for a Split occupation line, then COR684 assigned the default score for the ENTIRE Occupation line, rather than the "Other Industries" line. Follow the rules set out in Section III (C-6) above. For the cases where Class-of-Worker is missing, (Occs. 245, 142, 143, 144, 964 are affected), assume the default is Salaried Worker for 245 (as per Duncan's instructions, see Section IIIC.) For the others, just use the weighted average of the entire category for missing Class-of-Worker for any reason. (6) The sixth change made to the Occupational Education and Income scores is to include the splitting of Occupations 215 and 222 (as we did for the Ducan SEI and Siegel PRESTIGE scores in Section III, part C3). (7) For Occupations 692 and 695, the Industry category 389 (Leather Footwear) had no observations. I computed a weighted average of the entire "Leather Products" Industry categories (Ind=388-397), and used this for 692 and 694. These calculations are found in COR713j.XLS. (8) For some reason a "Don't Know" response for Detailed Occupation was coded to a '9999' missing value for Occupational Education and Income scores, rather than '9997' as every other mapping of 1970-basis Occupation did. This should change. (9) Finally, as for the Duncan SEI/Siegel PRESTIGE scores, we should account for the Allocated Industry categories in the 1970 Census Detailed Industry classification system. Where the entire Major Industry category is used as a Split line, that score will be used. Where the Major Industry category is itself split within a Split Occupation, the weighted average of the Industry splits will be used. These are computed in COR713j.XLS. (10) The Census defines SELF-EMPLOYED, BUSINESS INCORPORATED (Class-of-Worker = 3) as a SALARIED worker--NOT a self- employed worker as we define it in the WLS. The original code for assigning Occupational Education and Income scores set COW=3 to Self-Employed rather than Salaried. Change so that COW>=4 and COW<=6 are the self-employed codes. (These changes are NOT listed below!!!!) The following changes have been made to COR684, and will be included in COR713.SAS: A. FINAL LIST OF OCCUPATIONAL EDUCATION AND OCCUPATIONAL INCOME SCORE CHANGES 1970-Basis Category Occupational Education Occupational Income ------------------- ---------------------- --------------------- Occ. Ind. COW Old New Err. Old New Err. ---- ---- --- --- --- ---- --- --- ---- 022 798 797 (1) 042 482 481 (1) 053 791 790 (1) 061 573 572 (1) 081 150 149 (1) 102 973 972 (1) 114 994 993 (1) 141 673 672 (1) 142 6 947 914 (4) 165 080 (4) 143 1,3-6 026 025 (1) 143 6 730 644 (4) 085 025 (4) 144 -9 969 968 (1) 144 9* 969 968 (5) 144 6 969 952 (4) 290 176 (4) 163 769 768 (1) 165 636 635 (1) 180 220 219 (1) 183 604 603 (1) 193 283 282 (1) 205 410 409 (1) 215 907 364 383 (6) 461 552 (6) 215 917 364 383 (6) 461 552 (6) 215 927 364 340 (6) 461 310 (6) 215 937 364 315 (6) 461 304 (6) 222 907 530 595 (6) 563 709 (6) 222 917 530 595 (6) 563 709 (6) 222 927 530 553 (6) 563 509 (6) 222 937 530 415 (6) 563 368 (6) 230 281 280 (1) 231 372 371 (1) 245 499 1-2 627 424 (9) 564 641 (9) 245 499 3-6 367 233 (9) 422 472 (9) 245 499 6 586 233 (9,4) 542 472 (9,4) 245 499 -9 586 407 (9,5) 542 626 (9,5) 245 499 9* 586 424 (9,5) 542 641 (9,5) 245 639-647 -9 513 512 (1) 245 649 -9 513 512 (1) 245 667-668 3-5 453 452 (1) 245 669-698 1-2 329 422 (3) 245 669-698 3-5 422 329 (3) 245 699 1-2 377 315 (9) 329 411 (9) 245 699 3-5 297 218 (9) 422 337 (9) 245 699 6 342 218 (9,4) 381 337 (9,4) 245 699 -9 342 277 (9,5) 381 382 (9,5) 245 699 9* 342 315 (9,5) 381 411 (9,5) 245 727-767 1-2 458 611 (3) 245 727-767 3-5 611 458 (3) 245 -9 6 399 251 (4) 553 405 (4) 245 <017 6 399 367 (4) 553 422 (4) 245 017-058 6 586 367 (4) 542 422 (4) 245 067-078 6 273 202 (4) 629 499 (4) 245 107-267 6 542 267 (4) 772 482 (4) 245 268-399 6 532 381 (4) 713 504 (4) 245 407-429 6 369 232 (4) 596 460 (4) 245 447-479 6 463 236 (4) 671 579 (4) 245 507-599 6 428 312 (4) 633 522 (4) 245 607-608 6 353 308 (4) 474 468 (4) 245 609-627 6 376 290 (4) 365 325 (4) 245 628-638 6 184 137 (4) 319 245 (4) 245 639-647 6 298 288 (4) 512 506 (4) 245 648 6 134 128 (4) 280 320 (4) 245 649 6 298 288 (4) 512 506 (4) 245 657-658 6 373 331 (4) 408 424 (4) 245 667-668 6 364 317 (4) 524 452 (4) 245 669-698 6 342 297 (4) 381 329 (4) 245 707-719 6 614 603 (4) 660 668 (4) 245 727-767 6 467 318 (4) 566 458 (4) 245 769-799 6 259 219 (4) 245 807-947 6 586 367 (4) 542 422 (4) 245 >947 6 399 367 (4) 553 422 (4) 245 -9 9* 399 451 (5) 553 605 (5) 245 <017 9* 399 627 (5) 553 564 (5) 245 017-058 9* 586 627 (5) 542 564 (5) 245 067-078 9* 273 315 (5) 629 705 (5) 245 107-267 9* 542 570 (5) 772 802 (5) 245 268-399 9* 532 548 (5) 713 735 (5) 245 407-429 9* 369 390 (5) 596 617 (5) 245 447-479 9* 463 468 (5) 671 673 (5) 245 499 9* 407 424 (5) 626 641 (5) 245 507-599 9* 428 457 (5) 633 660 (5) 245 607-608 9* 353 379 (5) 474 478 (5) 245 609-627 9* 376 397 (5) 365 374 (5) 245 628-638 9* 184 222 (5) 319 381 (5) 245 639-647 9* 298 300 (5) 512 514 (5) 245 648 9* 134 144 (5) 280 219 (5) 245 649 9* 298 300 (5) 512 514 (5) 245 657-658 9* 373 391 (5) 408 401 (5) 245 667-668 9* 364 397 (5) 524 573 (5) 245 669-698 9* 342 377 (5) 381 422 (5) 245 699 9* 277 315 (5) 382 411 (5) 245 707-719 9* 614 615 (5) 660 659 (5) 245 727-767 9* 467 528 (5) 566 611 (5) 245 769-799 9* 259 290 (5) 245 807-947 9* 586 627 (5) 542 564 (5) 245 >947 9* 399 627 (5) 553 564 (5) 245 999** 3-6 251 367 (5) 405 422 (5) 245 999** 1-2 451 627 (5) 605 564 (5) 245 999** 9* 399 627 (5) 553 542 (5) 280 275 274 (1) 282 405 404 (1) 283 628-638 111 110 (1) 283 699 225 186 (9) 115 071 (9) 283 999** 186 225 (5) 071 115 (5) 284 <639 248 247 (1) 284 648 248 247 (1) 284 >649 248 247 (1) 284 699 271 263 (9) 248 303 (9) 284 999** 263 271 (5) 303 248 (5) 296 296 313 (2) 296 242 (2) 314 197 196 (1) 334 098 097 (1) 361 157 156 (1) 383 180 179 (1) 394 407-499 283 282 (1) 394 999** 294 287 (5) 120 111 (5) 395 999** 231 219 (5) 058 060 (5) 402 112 111 (1) 433 426 425 (1) 441 267 161 198 (9) 407 573 (9) 441 399 214 173 (9) 524 405 (9) 441 499 208 182 (9) 406 546 (9) 441 999** 184 208 (5) 483 406 (5) 452 067-078 277 276 (1) 452 999** 199 219 (5) 271 291 (5) 486 165 164 (1) 491 999** 144 135 (5) 152 107 (5) 492 999** 142 140 (5) 225 193 (5) 495 999** 137 148 (5) 258 247 (5) 503 039 038 (1) 542 045 044 (1) 550 083 082 (1) 601 086 085 (1) 612 047 046 (1) 089 088 (1) 613 017 016 (1) 620 052 051 (1) 640 <048 117 116 (1) 640 >049 117 116 (1) 640 999** 073 083 (5) 135 117 (5) 642 181 180 (1) 661 178 177 (1) 666 060 059 (1) 671 034 033 (1) 690 118 044 043 (1) 690 188-189 146 145 (1) 690 267 075 052 (9) 109 103 (9) 690 399 054 059 (9) 074 129 (9) 690 499 131 105 (9) 057 215 (9) 690 999** 058 131 (5) 113 057 (5) 692 158 064 063 (1) 692 267 000 053 (9) 106 105 (9) 692 319 017 016 (1) 692 389 163 015 (7) 065 024 (7) 692 399 071 060 (9) 085 107 (9) 692 499 163 055 (9) 065 144 (9) 692 999** 058 163 (5) 104 065 (5) 694 067-078 060 059 (1) 694 267 067 059 (9) 040 103 (9) 694 347 169 168 (1) 694 399 029 058 (9) 061 075 (9) 694 499 283 124 (9) 043 134 (9) 694 607-699 071 070 (1) 694 -9 076 075 (1) 694 999** 076 283 (5) 089 043 (5) 695 147 154 153 (1) 695 178 087 086 (1) 695 219 174 173 (1) 695 258 144 143 (1) 695 267 034 058 (9) 128 124 (9) 695 269 175 174 (1) 695 278 039 038 (1) 695 299 022 021 (1) 695 389 115 042 (7) 083 041 (7) 695 399 011 064 (9) 060 094 (9) 695 499 115 093 (9) 083 186 (9) 695 907-947 106 105 (1) 695 999** 064 115 (5) 112 083 (5) 715 417 358 357 (1) 715 999** 228 151 (5) 752 070 069 (1) 770 083 082 (1) 780 127 014 013 (1) 780 147 107 106 (1) 780 228 054 053 (1) 780 239-257 147 146 (1) 780 267 000 046 (9) 101 099 (9) 780 347-369 033 032 (1) 780 399 000 042 (9) 000 063 (9) 780 499 000 037 (9) 000 057 (9) 780 799 027 033 (9) 000 006 (9) 780 999** 051 077 (5) 075 036 (5) 785 119 079 078 (1) 785 137 079 078 (1) 785 138 079 078 (1) 785 177-198 040 039 (1) 785 259 051 050 (1) 785 267 000 059 (9) 022 038 (9) 785 399 000 059 (9) 015 043 (9) 785 499 000 064 (9) 000 058 (9) 785 799 027 094 (9) 038 022 (9) 785 347-369 046 045 (1) 785 777-798 113 112 (1) 785 999** 070 127 (5) 036 024 (5) 910 103 102 (1) 934 031 030 (1) 941 033 032 (1) 963 247 246 (1) 964 6 280 257 (4) 374 211 (4) 983 046 045 (2) 997 9999 9997 (8) 9999 9997 (8) * Indicates individual missing data for COW; i.e., cow^=-9 and (cow<1 or cow>6). ** Indicates individual missing data for IND; i.e., ind^=-9 and (ind<1 or ind>947). B. WHAT DIFFERENCE DOES IT MAKE? Following are the old and new Means and Standard Deviations for some selected, important variables in the Waves 2 and 3 data using the old and new Occupational Education and Occupational Income scores: Occ. Education Occ. Income -------------- -------------- Variable Wave Old New Old New ------------- ---- ---- ---- ---- ---- RFU06J1C 2 363.3 363.3 262.7 262.2 (R's 1975 Job,R92) (287.0) (287.1) (233.6) (233.6) RFU34JCE 2 380.3 380.3 293.1 292.5 (R's c/l Job,R92) (281.2) (281.3) (239.5) (239.4) RKU39SSF 2 352.5 352.5 269.0 268.0 (S's 1992 Job,R92) (285.8) (285.9) (228.4) (227.9) RFU03JAF 2 427.8 428.2 313.0 312.4 (R's 1975 Asp,R92) (299.3) (299.5) (243.3) (243.4) SFU06REC 3 337.2 337.1 200.2 200.2 (S's 1st Job,S94) (306.2) (306.2) (214.4) (214.5) SFU29REC 3 359.0 359.1 280.1 280.3 (S's 1977 Job,S94) (294.9) (295.0) (234.7) (235.4) SFU53RED 3 375.8 375.7 280.2 279.6 (S's c/l Job,S94) (287.3) (287.4) (237.3) (237.3) Occ. Educ. Occ.Inc. ------------------ ----------------- Variable Wave Old New Diff Old New Diff -------- ---- ---- ---- ---- --- --- ---- Head's Occ (R75), OCHH57U by: ----------------------------- R's 1st Job OCX1U 1,1 .2905 .2902 -.0003 .1746 .1733 -.0013 R's c/l Job(R75) OCXCURU 1,1 .2569 .2567 -.0002 .1487 .1494 +.0007 R's 1975 Job(R92) RFU06J1C 1,2 .2532 .2534 +.0002 .1486 .1500 +.0014 R's c/l Job(R92) RFU34JCE 1,2 .2299 .2300 +.0001 .1292 .1294 +.0002 Sib's 1975 Job(R75), OCSIB by: ------------------------------ R's 1975 Job (R75) OCXCURU 1,1 .2691 .2688 -.0003 .1296 .1277 -.0019 S's 1977 Job (S77) XOCXCUR 1,1 .8545 .8547 +.0002 .7166 .7217 +.0051 R's 1975 Job(R92) RFU06J1C 1,2 .2539 .2540 +.0001 .1273 .1275 +.0002 R's c/l Job(R92) RFU34JCE 1,2 .2324 .2329 +.0005 .1352 .1347 -.0005 S's 1992 Job(R92) RKU39SSF 1,2 .7358 .7363 +.0005 .5719 .5743 +.0024 S's 1977 Job(S94) SFU29REC 1,3 .8289 .8288 -.0001 .6843 .6857 +.0014 S's c/l Job(S94) SFU53RED 1,3 .7155 .7158 +.0003 .5512 .5524 +.0012 R's First Job (R75), OCX1U by: ------------------------------ R's c/l Job (R75) OCXCURU 1,1 .7245 .7253 +.0008 .5890 .5903 +.0013 S's First Job (S77) XOCX1 1,1 .3131 .3134 +.0003 .2675 .2673 -.0002 R's 1975 Job (R92) RFU06J1C 1,2 .7091 .7097 +.0006 .5752 .5771 +.0019 R's c/l Job (R92) RFU34JCE 1,2 .6435 .6440 +.0005 .4956 .4979 +.0023 S's First Job (S94) SFU06REC 1,3 .2701 .2699 -.0002 .1507 .1511 +.0004 R's C/L Job (R75), OCXCURU by: ------------------------------ S's 1977 Job (S77) XOCXCUR 1,1 .2467 .2470 +.0003 .2492 .2484 -.0008 R's 1975 Job (R92) RFU06J1C 1,2 .8623 .8623 - 0 - .7752 .7755 +.0003 R's c/l Job (R92) RFU34JCE 1,2 .7203 .7209 +.0006 .6098 .6102 +.0004 S's 1977 Job (S94) SFU29REC 1,3 .2553 .2558 +.0005 .1455 .1436 -.0019 Head's Job (S77), XOCHH57 by: ----------------------------- S's First Job (S77) XOCX1 1,1 .2788 .2785 -.0003 .1539 .1504 -.0025 S's c/l Job (S77) XOCXCUR 1,1 .2369 .2359 -.0010 .1197 .1214 +.0017 S's First Job (S94) SFU06REC 1,3 .2653 .2643 -.0010 .1391 .1357 -.0034 S's 1977 Job (S94) SFU29REC 1,3 .2134 .2130 -.0004 .1003 .1016 +.0013 S's c/l Job (S94) SFU53RED 1,3 .2075 .2055 -.0020 .0985 .0983 -.0002 S's First Job (S77), XOCX1 by: ------------------------------ S's c/l Job (S77) XOCXCUR 1,1 .7653 .7654 +.0001 .6154 .6173 +.0019 S's First Job (S94) SFU06REC 1,3 .9439 .9436 -.0003 .8935 .8941 +.0006 S's 1977 Job (S94) SFU29REC 1,3 .7663 .7665 +.0002 .6224 .6259 +.0035 S's c/l Job (S94) SFU53RED 1,3 .6426 .6418 -.0008 .5179 .5243 +.0064 S's C/L Job (S77), XOCXCUR by: ------------------------------ S's 1977 Job (S94) SFU29REC 1,3 .9991 .9991 - 0 - .9971 .9973 +.0002 S's c/l Job (S94) SFU53RED 1,3 .7390 .7386 -.0004 .6227 .6235 +.0008 R's 1975 Job (R92), RFU06J1C by: -------------------------------- R's c/l Job (R92) RFU34JCE 2,2 .7666 .7666 - 0 - .6845 .6856 +.0011 S's 1977 Job (S94) SFU29REC 2,3 .2451 .2451 - 0 - .1305 .1277 -.0028 R's C/L Job (R92), RFU34JCE by: ------------------------------- S's 1992 Job (R92) RKU39SSF 2,2 .2204 .2208 +.0004 .1390 .1397 +.0007 S's c/l Job (S94) SFU53RED 2,3 .2154 .2155 +.0001 .1086 .1077 -.0009 S's First Job (S94), SFU06REC by: --------------------------------- S's 1977 Job (S94) SFU29REC 3,3 .7587 .7586 -.0001 .6081 .6095 +.0014 S's c/l Job (S94) SFU53RED 3,3 .7138 .7137 -.0001 .5748 .5767 +.0019 S's 1977 Job (S94), SFU29REC by: -------------------------------- S's c/l Job (S94) SFU53RED 3,3 .7609 .7608 -.0001 .6511 .6521 +.0010 R's 1957 Asps (R57), SPOCASP1 by: --------------------------------- R's 1957 Asps (R75) OCZU 1,1 .6472 .6477 +.0005 .7548 .7547 -.0001 R's First Job (R75) OCX1U 1,1 .5273 .5273 - 0 - .5602 .5604 +.0002 R's c/l Job (R75) OCXCURU 1,1 .4975 .4979 +.0004 .5381 .5386 +.0005 R's 1975 Job (R92) RFU06J1C 1,2 .4990 .4993 +.0003 .5406 .5410 +.0004 R's c/l Job (R92) RFU34JCE 1,2 .4631 .4629 -.0002 .4614 .4618 +.0004 R's 1957 Asps (R75), OCZU by: ----------------------------- R's 1975 Asps (R75) OCR10 1,1 .4915 .4933 +.0018 .5335 .5342 +.0007 S's 1957 Asps (S77) XOCZ 1,1 .2344 .2359 +.0015 .2713 .2712 -.0001 R's First Job (R75) OCX1U 1,1 .5106 .5104 -.0002 .5833 .5832 -.0001 R's c/l Job (R75) OCXCURU 1,1 .4832 .4832 - 0 - .5704 .5701 -.0003 R's 1975 Job (R92) RFU06J1C 1,2 .4779 .4784 +.0005 .5689 .5690 +.0001 R's c/l Job (R92) RFU34JCE 1,2 .4446 .4449 +.0003 .4879 .4881 +.0002 R's 1975 Asps (R75), OCR10 by: ------------------------------ R's 1975 Asps (R92) RFU03JAF 1,2 .7085 .7094 +.0009 .6491 .6498 +.0007 S's 1975 Asps (S77) XOCR10 1,1 .1967 .1980 +.0013 .1805 .1804 -.0001 R's c/l Job (R75) OCXCURU 1,1 .7090 .7100 +.0010 .7388 .7399 +.0011 R's 1975 Job (R92) RFU06J1C 1,2 .6505 .6515 +.0010 .6386 .6390 +.0004 R's c/l Job (R92) RFU34JCE 1,2 .5981 .5984 +.0003 .5488 .5488 - 0 - R's 1975 Asps (R92), RFU03JAF by: --------------------------------- R's c/l Job (R75) OCXCURU 1,2 .7842 .7842 - 0 - .6934 .6945 +.0011 R's 1975 Job (R92) RFU06J1C 2,2 .8680 .8679 -.0001 .8626 .8633 +.0007 R's c/l Job (R92) RFU34JCE 2,2 .7956 .7957 +.0001 .7339 .7360 +.0021 S's 1957 Asps (S77), XOCZ by: ----------------------------- S's 1975 Asps (S77) XOCR10 1,1 .4395 .4379 -.0016 .4836 .4842 +.0006 S's First Job (S77) XOCX1 1,1 .3984 .3994 +.0010 .5062 .5033 -.0029 S's c/l Job (S77) XOCXCUR 1,1 .3973 .3972 -.0001 .4948 .4946 -.0002 S's First Job (S94) SFU06REC 1,3 .4053 .4061 +.0008 .4912 .4900 -.0012 S's 1977 Job (S94) SFU29REC 1,3 .4194 .4195 +.0001 .5146 .5163 +.0017 S's c/l Job (S94) SFU53RED 1,3 .3794 .3796 +.0002 .4463 .4521 +.0058 S's 1977 Asps (S77), XOCR10 by: ------------------------------- S's c/l Job (S77) XOCXCUR 1,1 .7057 .7052 -.0005 .6910 .6920 +.0010 S's 1977 Job (S94) SFU29REC 1,3 .7216 .7212 -.0004 .7004 .7018 +.0014 S's c/l Job (S94) SFU53RED 1,3 .6046 .6026 -.0020 .5557 .5570 +.0013 The means and standard deviations for both Occupational Education scores and Occupational Income scores change very little. The Occupational Income means change a bit more than the Occ. Educ. scores, but with no real trend. The correlations of both Occ. Ed. and Occ. Inc. scores also show very little change with the corrections. In general: - Corrs. of R's First Job with all subsequent jobs INCREASES SLIGHTLY (Occ. Ed. only) - Corrs. of R's 1975 Job with all subsequent jobs INCREASES SLIGHTLY (Occ. Ed. only) - Corrs. of Head's occ (sibling 1977 report) with all subsequent Sib jobs DECREASES SLIGHTLY (Occ. Ed. only) - R's retrospective report of 1957 aspirations shows SLIGHTLY INCREASED correlation with other aspiration reports (Occ. Ed. only) - R's midlife aspirations have SLIGHTLY INCREASED correlations with subsequent jobs (Occ. Ed. only) - S's retrospective report of 1957 aspirations shows SLIGHTLY INCREASED correlations with subsequent jobs (Occ. Ed. only) - Occ. Inc. correlations have larger changes than Occ. Ed. correls. - Occ. Inc. correlations corresponding to SIBLING variables tend to INCREASE MODERATELY--especially those correlated with 1994 Sib's current/last job - Occ. Inc. correlations for Sibling's Head by Sibling's First Job (1994 report) DECREASE MODERATELY *************************************************************************** *** V I . R E F E R E N C E S *** *************************************************************************** Blau, Peter M. and Otis Dudley Duncan. 1967. _The American Occupational Structure_. New York: The Free Press. Duncan, Otis Dudley. 1961a. "A Socioeconomic Index for All Occupations." Pp. 109-161 and 263-295 in Reiss, Albert J. (ed.) _Occupations and Social Status_. New York: The Free Press. Duncan, Otis Dudley. 1961b (March 5). "Scores on Socioeconomic Index for Extended Major Occupation Groups, 1950." Duncan, Otis Dudley. 1961c (October). "Socioeconomic Status Scores for Detailed Occupations." Siegel, Paul Mathew. 1971. "Prestige in the American Occupational Structure." Dissertation. University of Chicago. Featherman, David L., F. Lancaster Jones, and Robert M. Hauser. 1975. "Assumptions of Social Mobility Research in the U.S.: The Case of Occupational Status." _Social Science Research_. 4:329-360. (Also, IRP Reprint #183). Featherman, David L., Michael Sobel and David Dickens. 1975. "A Manual for Coding Occupation and Industries into Detailed 1970 categories and a Listing of 1970-Basis Duncan Socioeconomic and NORC Prestige Scores." CDE Working Paper #75-1. Center for Demography and Ecology. University of Wisconsin-Madison. U.S. Bureau of the Census. 1972. _1970 Occupation and Industry Classification Systems in Terms of Their 1960 Occupation and Industry Elements._ By John A. Priebe, Joan Heinkel, and Stanley Greene. Technical Paper No. 26. Washington, D.C.: U.S. Government Printing Office. Sheehy, L., D. Netkin, and J. Grant. "Social Factors in Aspirations and Achievements: Occupation-Industry Coding Handbook, 1974-1975." Hauser, Robert M. and David L. Featherman. 1977. _The Process of Stratification: Trends and Analyses_. New York: Academic Press. Stevens, Gillian and David L. Featherman. 1981. "A Revised Socioeconomic Index of Occupational Status." _Social Science Research_. 10:364-395. *************************************************************************** *** V I I . A T T A C H E D D O C U M E N T S *** *************************************************************************** COR713.ASC This document--details the status scales checked and the errors found in the 1970-basis mappings. COR713.SAS SAS code (module) that contains the final versions of DUNCAN SEI, SIEGEL PRESTIGE, MSEI2, TSEI2, OCCUPATIONAL EDUCATION SCORES, and OCCUPATIONAL INCOME SCORES used in the WLS data. Sets missing data appropriately for each wave. Sets scores appropriately depending on type of missing data for Split Occupation lines. Also maps MAJOR OCCUPATION and MAJOR INDUSTRY categories (but see MEMO133 for explanation of these.) To obtain the 1970-basis mappings from a SAS MACRO call, see COR713l.SAS and COR713m.SAS. COR713a.ASC Duncan's original 1950-basis SEI scores. From Duncan 1961a (Reiss), Appendix B-1. COR713b.ASC Duncan's original 1950-basis SEI scores for 1950 Major Occupation Categories. From Duncan 1961a (Reiss), Table VII-4. COR713c.ASC Duncan's 1950-basis SEI scores for 1950 Major Occupation Categories, some split by Industry and/or Class-of-Worker as in Blau and Duncan 1967. From Duncan 1961b (Mimeo). COR713d.ASC Duncan's 1950-basis SEI scores for 1960 Detailed Occupation Categories. Industry/Class-of-Worker Splits are included. From Duncan 1961c (Mimeo). COR713e.ASC Siegel's 1960-basis PRESTIGE scores for 1960 Detailed Occupation Categories. From Siegel 1971, Table 5. COR713f.ASC Footnote from Featherman, Jones and Hauser (1975), which lists the SEI and PRESTIGE scores for Major Occupation Categories (but rounds PRESTIGE to two digits!) COR713g.ASC Appendix B from Featherman, Sobel and Dickens (1975) which maps 1960-basis SEI and PRESTIGE scores (COR713d and COR713e above) to 1970 Detailed Occupation Categories. COR713h.XLS An EXCEL worksheet containing the data in Tech. Paper 26, Table 1. Instructions for using this spreadsheet are included in this COR (COR713.ASC). COR713i.XLS Mapping 1960 Detailed Industry Categories to the 1970 Detailed Industry Classification System - for use with COR713h.XLS. COR713j.XLS Calculations of SEI/MSEI/TSEI/Occ.Ed./Occ.Inc. score to use when a weighted average must be computed. For example, when class- of-worker is not known, or when an allocated occupation category is used to represent an entire major occupation category (rather than allocation for missing data, as in the Census usage.) COR713k.TAR SAS programs and ASCII data files for checking TOTAL-based Occupational Education and Occupational Income scores. To be used with the original Nam-Powers data (educ.dat and income.dat) found in COR638.ZIP. COR713l.SAS Code to call COR713m.SAS as a MACRO. Edit the template in COR713l.SAS to call COR713m.SAS as a SAS MACRO. COR713m.SAS COR713.SAS in a SAS MACRO format. This MACRO code should not be changed. Use the template in COR713l.SAS to get the scores from COR713m.SAS. MEMO133.ASC Describes each mapping used in the WLS for 1970-basis Occupation variables. Lists all applicable CORS and MEMOs related to these mappings. Lists all known errors in the data, how introduced, how and when fixed.