** user=**** ** password=**** ** package=spss ** project=lis set printback=listing . * GERMANY LIS 1989 : SPSS Example CAMSIS matching to DE 89 data. get file= ge89p . * Current occ unit (isco68 minors)=pocc. * (incomplete) Employment status info from ptypewk only. * Derive employment status variable. fre var= ptypewk. compute status=ptypewk. recode status (212 thru 233=1) (241 thru 243=2) (251 thru 253=3) (1,111,121,131,141,151,311,321,411,421,431,451, 511 thru 543 =4) (else=9) . *recode status (212 thru 233=1) (241 thru 243=2) (251 thru 253=3) (1,111,121,131,141,151,311,321,411,421,431,451, 511 thru 543 =4) (2,112,122,123,113,132,133,142,143,152, 312,313,322,323,412,413,422,423,432,433,452,453 =5) (else=9) . * the latter is an alternative handling of employees. add value labels status 1 "Self-Employed, no employees" 2 "Self-Employed, with employees" 3 "Family Assistant" 4 "Public service" 5 "Employee (contract)" 6 "Worker (non-contract)" 9 "Unknown". variable label status "Employment status". cro status by psex. * Key indexing variable is minbst. compute minbst=-999. if (pocc > 0 ) minbst=pocc*10 + status . temp. select if (pocc > 0). means tables=minbst by status . * Match with CAMSIS file. sort cases by minbst. sav out= cstemp . get file="u:/camsis/csdeis89.sav" . descriptives var=all. sort cases by minbst. match files table=* /in=csocc /file=cstemp /in=source /by=minbst. fre var=csocc source. cro csocc by source. select if (source=1). * (this removes those occbst values which are not * represented in the LIS sample). cro source by psex. * Adjust: Current values are within gender. * also can make a cross-gender CAMSIS variable : . compute cgentcs=-999. if (psex=1) cgentcs=htcs . if (psex=2) cgentcs=wtcs . compute cgenscs=-999. if (psex=1) cgenscs=hrcs . if (psex=2) cgenscs=wrcs . * Assess coverage of variables. descriptives var= minbst cgentcs cgenscs siops isei . temp. select if (pocc > 0). descriptives var= minbst cgentcs cgenscs siops isei . temp. select if (pocc > 0 & siops > 0). descriptives var= minbst cgentcs cgenscs siops isei . * note that the last summary shows us that there are quite a lot * of cases with occs but without camsis or isei, also some with * camsis scores but have missing siops by ISMF translation. * Quick assessment: compare values by a national specific * education recode. compute educ=peduc. recode educ (0,1,2=1) (3 thru 8=2) (9 thru 15=3) (else=-9) . add value labels educ 1 "None or low level" 2 "Intermediate" 3 "Higher and/or University". variable label educ "Germany education categories". fre var=educ . temp. select if (pocc > 0). descriptives var= educ cgentcs cgenscs siops isei . temp. select if (pocc > 0 & siops > 0 & psex=1). means tables= cgentcs cgenscs siops isei by educ /statistics=anova. temp. select if (pocc > 0 & siops > 0 & psex=2). means tables= cgentcs cgenscs siops isei by educ /statistics=anova. /* For interest here are the Eta-squared statistics with education from above: Men: CAMSIS title only : 0.35 CAMSIS title by status : 0.29 SIOPS : 0.36 ISEI: 0.38 Women: CAMSIS title only : 0.18 CAMSIS title by status : 0.17 SIOPS : 0.24 ISEI: 0.23 */