* Encoding: UTF-8. /** ================================================================================================================. /** IMPORTANT NOTES. /** /** 1. ITEM NUMBERING /** In the process of bulilding a valid "baseline" of RTs, some "filler items" which had been initially selected were subsequently filtered out for validity reasons. /** Thus, RTs to items A8, A9, and A10 were excluded because they were almost perfectly correlated to RTs to item A7 (r>0.98); /** as reported by the survey organization (FORS), the most likely cause is a technical problem in the collection of RTs. /** Likewise, item A11 was excluded because the number of missing values (6.0%) slightly exceed the threshold of 5% finally retained for inclusion. /** Filtering out irrelevant items is done via the commands temporary / select if. Thus, only the relevant items are included in the analysis throughout! /** Considering that item numbering is involved in many portions of this (very long) syntax, I gave up renumbering the items to avoid errors in adapting the syntax. /** However, please note that ALL ITEMS ARE RENAMED AT THE END OF THE SYNTAX; thus, filler items A1-A14 are renumbered to a new range of A1-A10. /** Accordingly, the numbering of items in the final datafile (S2015_RT7.sav) is perfectly consistent with the numbering used throughout the paper. /** 2. SPSS WARNINGS. /** On several occasions, executing the syntax will return the following message: /** "Warning # 3211: On at least one case, the value of the weight variable was zero, negative, or missing (...)". /** This warning is NOT a cause for concern and can safely be ignored. /** As a matter of fact, the weighting variable (weight_d) is missing for the 3972 respondents who did not complete the survey. /** Only the 5337 respondents who did complete the survey were assigned a positive weight and are thus included in the analyses. /** 3. EMPIRICAL ANALYSES. /** Analyses presented in the paper and its accompaniying Appendices are performed by the syntax and will appear in the SPSS Output. /** Whenever a statistical analysis refers to a specific Table or Figure, it will be designated in the syntax by: @@@ OUTPUT @@@. /** Two exceptions are worth noting. First, purely descriptive analyses (e.g., Table 1, Figure 1, Figure A.3) are omitted. However, stastical evidence can be /** easily obtained via frequency tables in the final datafile. Second, some analyses summarized in the Appendix can be obtained from the author upon request /** (although some of them could easily be performed on the basis of available data). /** 4. DATA HANDLING. /** Make sure that ALL NECESSARY FILES (Selects postelectoral survey AND Response times file) ARE SAVED IN THE SAME DIRECTORY. /** Then SPECIFY THE CURRENT WORKING DIRECTORY by using the command "cd". The procedure is explained in the next lines of the syntax. /** ================================================================================================================. /** ================================================================================================================. /** SPECIFY THE CURRENT WORKING DIRECTORY. /** In the following command, please replace the "..." by the actual location of the directory where you saved the necessary datafiles. cd "..." /** Example: cd "C:\Users\Lionel Marquis\Documents\PROF\RESEARCH\SELECTS\Response times (Selects 2015)" /** ================================================================================================================. /** 1. MERGING FILES (RTs + Selects sample). get file="PES timing merged only RT.sav"/keep=all. rename variables (ID1FORS=userid). SORT CASES BY userid(A). MATCH FILES /FILE=* /FILE="726_Selects2015_PES_Data_v1.02.sav" /BY userid. EXECUTE. save outfile="S2015_RT1.sav"/keep=all. /** ========================================================================================================. get file="S2015_RT1.sav"/keep=all. WEIGHT BY weight_d. rename variables (IMODE=IMETHOD). var lab IMETHOD "Data collection method: CATI vs. CAWI". val lab IMETHOD 1"CAWI" 2"CATI". /** 2. COMPUTING RESPONSE TIMES.. /** Selection criterion : Less than 5 percent missing values (DK, NA, etc.) for each item. /** Categorization: Attitudinal questions (OA); behavioral/factual questions (OB); knowledge questions (OK). /** 2.1. SELECTING ITEMS. compute oa1.tim=f15420.tim. if missing(f15420) oa1.tim=99999. compute oa2.tim=f15430.tim. if missing(f15430) oa2.tim=99999. compute oa3.tim=f15440.tim. if missing(f15440) oa3.tim=99999. compute oa4.tim=f15470.tim. if missing(f15470) oa4.tim=99999. compute oa5.tim=f15480.tim. if missing(f15480) oa5.tim=99999. compute oa6.tim=f15490.tim. if missing(f15490) oa6.tim=99999. compute oa7.tim=f15760.tim. if missing(f15760) oa7.tim=99999. compute oa8.tim=f15761.tim. if missing(f15761) oa8.tim=99999. compute oa9.tim=f15762.tim. if missing(f15762) oa9.tim=99999. compute oa10.tim=f15763.tim. if missing(f15763) oa10.tim=99999. compute oa11.tim=f15800a.tim. count oa11.mis=f15801,f15802,f15803,f15804(missing). if oa11.mis gt 1 oa11.tim=99999. compute oa12.tim=f10100.tim. if missing(f10100) oa12.tim=99999. compute oa13.tim=f13700.tim. if missing(f13700) oa13.tim=99999. compute oa14.tim=f14600.tim. if missing(f14600) oa14.tim=99999. compute ob1.tim=f13400a.tim. count ob1.mis=f13401,f13402,f13403,f13404,f13405(missing). if ob1.mis gt 1 ob1.tim=99999. compute ob2.tim=f13300.tim. if missing(f13300) ob2.tim=99999. compute ob3.tim=f11100.tim. if missing(f11100) ob3.tim=99999. compute ob4.tim=f21310.tim. if missing(f21310) ob4.tim=99999. compute ob5.tim=f20600.tim. if missing(f20600) ob5.tim=99999. compute ob6.tim=f20601.tim. if missing(f20601) ob6.tim=99999. compute ob7.tim=f20602.tim. if missing(f20602) ob7.tim=99999. compute ob8.tim=99999. if IMETHOD=1 and not missing(f20300) ob8.tim=f20300Cawi.tim. if IMETHOD=2 and not missing(f20300) ob8.tim=f20300Cati.tim. compute ob9.tim=f20500.tim. if missing(f20500) ob9.tim=99999. compute ob10.tim=f20210.tim. if missing(f20210) ob10.tim=99999. compute ok1.tim=f15900.tim. compute ok2.tim=f16000.tim. compute ok3.tim=f16100.tim. compute ok4.tim=f16300.tim. missing values oa1.tim,oa2.tim,oa3.tim,oa4.tim,oa5.tim,oa6.tim,oa7.tim,oa8.tim,oa9.tim,oa10.tim,oa11.tim,oa12.tim,oa13.tim,oa14.tim, ob1.tim,ob2.tim,ob3.tim,ob4.tim,ob5.tim,ob6.tim,ob7.tim,ob8.tim,ob9.tim,ob10.tim (99999). /** 2.2. COMPUTING THE 2ND AND 98TH PERCENTILES. recode LingReg (1=1)(2=2)(3,4=3) into lregion. val lab lregion 1"German-speaking" 2"French-speaking" 3"Italian/Rumantsch-speaking". SORT CASES BY lregion IMETHOD. SPLIT FILE LAYERED BY lregion IMETHOD. FREQUENCIES VARIABLES=oa1.tim oa2.tim oa3.tim oa4.tim oa5.tim oa6.tim oa7.tim oa8.tim oa9.tim oa10.tim oa11.tim oa12.tim oa13.tim oa14.tim ob1.tim ob2.tim ob3.tim ob4.tim ob5.tim ob6.tim ob7.tim ob8.tim ob9.tim ob10.tim ok1.tim ok2.tim ok3.tim ok4.tim /FORMAT=NOTABLE /PERCENTILES=2.0 98.0 /ORDER=ANALYSIS. SPLIT FILE OFF. /** On the basis of the above descriptive analysis, I now assign empirical values to compute "trimmed" statistics (2nd and 98th percentiles). /** NB: For the Italian/Rumantsch region, CATI method, the 97th percentile was used to replace the 98th perc. (which does not exist for this distribution, cf. N=47). do if lregion=1 and IMETHOD=1. compute oa1.02pct=7. compute oa1.98pct=77. compute oa2.02pct=4. compute oa2.98pct=36. compute oa3.02pct=6. compute oa3.98pct=67. compute oa4.02pct=5. compute oa4.98pct=66.8395373999992. compute oa5.02pct=5. compute oa5.98pct=45.5429147999994. compute oa6.02pct=4. compute oa6.98pct=34. compute oa7.02pct=4. compute oa7.98pct=43. compute oa8.02pct=4. compute oa8.98pct=40. compute oa9.02pct=4. compute oa9.98pct=40. compute oa10.02pct=4. compute oa10.98pct=40. compute oa11.02pct=19. compute oa11.98pct=150. compute oa12.02pct=3. compute oa12.98pct=39. compute oa13.02pct=4. compute oa13.98pct=42. compute oa14.02pct=3. compute oa14.98pct=34. compute ob1.02pct=11. compute ob1.98pct=77. compute ob2.02pct=4. compute ob2.98pct=26. compute ob3.02pct=6. compute ob3.98pct=53. compute ob4.02pct=8. compute ob4.98pct=86.3759475999968. compute ob5.02pct=3. compute ob5.98pct=20. compute ob6.02pct=2. compute ob6.98pct=10. compute ob7.02pct=3. compute ob7.98pct=26. compute ob8.02pct=8. compute ob8.98pct=80. compute ob9.02pct=2. compute ob9.98pct=16. compute ob10.02pct=5. compute ob10.98pct=24. compute ok1.02pct=6. compute ok1.98pct=204.942098799998. compute ok2.02pct=3. compute ok2.98pct=107.140369399999. compute ok3.02pct=3. compute ok3.98pct=76. compute ok4.02pct=3.60109059999661. compute ok4.98pct=92. end if. do if lregion=1 and IMETHOD=2. compute oa1.02pct=16. compute oa1.98pct=92.0671371999997. compute oa2.02pct=10. compute oa2.98pct=71. compute oa3.02pct=13. compute oa3.98pct=98.1414248498818. compute oa4.02pct=13. compute oa4.98pct=66. compute oa5.02pct=12. compute oa5.98pct=88. compute oa6.02pct=9. compute oa6.98pct=65. compute oa7.02pct=1.75396771419598. compute oa7.98pct=50.9582694000005. compute oa8.02pct=1.54616909005667. compute oa8.98pct=51. compute oa9.02pct=1. compute oa9.98pct=50. compute oa10.02pct=1. compute oa10.98pct=50. compute oa11.02pct=37.4976593999997. compute oa11.98pct=176. compute oa12.02pct=8. compute oa12.98pct=52.7306211999997. compute oa13.02pct=9. compute oa13.98pct=52. compute oa14.02pct=10. compute oa14.98pct=49.9550952000001. compute ob1.02pct=30.6212904000002. compute ob1.98pct=123. compute ob2.02pct=8. compute ob2.98pct=43. compute ob3.02pct=5. compute ob3.98pct=65. compute ob4.02pct=7. compute ob4.98pct=67.9296176. compute ob5.02pct=5. compute ob5.98pct=27. compute ob6.02pct=2. compute ob6.98pct=14. compute ob7.02pct=5. compute ob7.98pct=40.1759482000001. compute ob8.02pct=9. compute ob8.98pct=73. compute ob9.02pct=5. compute ob9.98pct=26. compute ob10.02pct=4. compute ob10.98pct=26. compute ok1.02pct=10. compute ok1.98pct=78. compute ok2.02pct=6. compute ok2.98pct=64.4283752000001. compute ok3.02pct=7. compute ok3.98pct=48.6628138. compute ok4.02pct=8. compute ok4.98pct=49. end if. do if lregion=2 and IMETHOD=1. compute oa1.02pct=7.03756499999983. compute oa1.98pct=79. compute oa2.02pct=5. compute oa2.98pct=56.4755746000002. compute oa3.02pct=5. compute oa3.98pct=72.3143276. compute oa4.02pct=5. compute oa4.98pct=61.663920274216. compute oa5.02pct=5. compute oa5.98pct=64. compute oa6.02pct=4. compute oa6.98pct=39. compute oa7.02pct=4. compute oa7.98pct=42. compute oa8.02pct=4. compute oa8.98pct=42. compute oa9.02pct=4. compute oa9.98pct=41.2413909999999. compute oa10.02pct=3. compute oa10.98pct=41. compute oa11.02pct=20. compute oa11.98pct=141.174831071326. compute oa12.02pct=3. compute oa12.98pct=42. compute oa13.02pct=4. compute oa13.98pct=46.2181067999998. compute oa14.02pct=3. compute oa14.98pct=40. compute ob1.02pct=10.4816237999996. compute ob1.98pct=85. compute ob2.02pct=5. compute ob2.98pct=36.7449185999999. compute ob3.02pct=7. compute ob3.98pct=47.8012300000001. compute ob4.02pct=9. compute ob4.98pct=91.7041411704963. compute ob5.02pct=3. compute ob5.98pct=22. compute ob6.02pct=2. compute ob6.98pct=12. compute ob7.02pct=3. compute ob7.98pct=33. compute ob8.02pct=8. compute ob8.98pct=79. compute ob9.02pct=2. compute ob9.98pct=19. compute ob10.02pct=5. compute ob10.98pct=25. compute ok1.02pct=3. compute ok1.98pct=281.391291999999. compute ok2.02pct=3. compute ok2.98pct=111.2154576. compute ok3.02pct=3. compute ok3.98pct=126.9581776. compute ok4.02pct=3. compute ok4.98pct=145.5046336. end if. do if lregion=2 and IMETHOD=2. compute oa1.02pct=16. compute oa1.98pct=81.6400795565149. compute oa2.02pct=10.262701. compute oa2.98pct=84.092638. compute oa3.02pct=11.1891482000001. compute oa3.98pct=68. compute oa4.02pct=14.2082335350458. compute oa4.98pct=64.804102. compute oa5.02pct=14. compute oa5.98pct=63.7574440000003. compute oa6.02pct=11.8918026. compute oa6.98pct=74.5620845499565. compute oa7.02pct=7. compute oa7.98pct=36.9682152. compute oa8.02pct=7. compute oa8.98pct=35.8833048. compute oa9.02pct=6. compute oa9.98pct=36.9682152. compute oa10.02pct=6. compute oa10.98pct=36.7871122020269. compute oa11.02pct=38.4245358000002. compute oa11.98pct=173.176044179256. compute oa12.02pct=8. compute oa12.98pct=47.4576704. compute oa13.02pct=10. compute oa13.98pct=46. compute oa14.02pct=8. compute oa14.98pct=44. compute ob1.02pct=31. compute ob1.98pct=130.725953197072. compute ob2.02pct=8. compute ob2.98pct=37.1839752. compute ob3.02pct=10. compute ob3.98pct=59.7579509965723. compute ob4.02pct=8.10236860000009. compute ob4.98pct=58.9397428. compute ob5.02pct=5. compute ob5.98pct=26. compute ob6.02pct=2. compute ob6.98pct=12.9958609785113. compute ob7.02pct=4.31359060000001. compute ob7.98pct=41. compute ob8.02pct=8.34354240000008. compute ob8.98pct=54.8479076. compute ob9.02pct=4. compute ob9.98pct=24.9825530024122. compute ob10.02pct=3. compute ob10.98pct=19.9719988000001. compute ok1.02pct=10.5893048000001. compute ok1.98pct=65.7848182472179. compute ok2.02pct=7. compute ok2.98pct=46.3289780373244. compute ok3.02pct=7.3434048. compute ok3.98pct=38. compute ok4.02pct=8. compute ok4.98pct=48. end if. do if lregion=3 and IMETHOD=1. compute oa1.02pct=7. compute oa1.98pct=116.512194049537. compute oa2.02pct=4. compute oa2.98pct=54.9483980653671. compute oa3.02pct=4.66640999999998. compute oa3.98pct=70.4445991499339. compute oa4.02pct=4. compute oa4.98pct=59. compute oa5.02pct=4.00021239999993. compute oa5.98pct=121.064864429137. compute oa6.02pct=4.28978039999998. compute oa6.98pct=56.2090961942449. compute oa7.02pct=4.18579799999995. compute oa7.98pct=44.268362. compute oa8.02pct=4. compute oa8.98pct=43.8084708. compute oa9.02pct=4. compute oa9.98pct=43.7377844. compute oa10.02pct=4. compute oa10.98pct=43.1808912. compute oa11.02pct=18.1136135999999. compute oa11.98pct=154.891197420489. compute oa12.02pct=3. compute oa12.98pct=30.775584. compute oa13.02pct=3.93932999999997. compute oa13.98pct=43.61333. compute oa14.02pct=3. compute oa14.98pct=37.2035995896232. compute ob1.02pct=9.07144439999991. compute ob1.98pct=80.0114978748348. compute ob2.02pct=4. compute ob2.98pct=53.893577605159. compute ob3.02pct=7. compute ob3.98pct=60.9217997948116. compute ob4.02pct=9. compute ob4.98pct=122.3075232. compute ob5.02pct=3. compute ob5.98pct=23.6674331916557. compute ob6.02pct=2. compute ob6.98pct=10.4608998974058. compute ob7.02pct=2.21895759999999. compute ob7.98pct=32.9308998974058. compute ob8.02pct=8.47550239999986. compute ob8.98pct=72.3072665982705. compute ob9.02pct=2. compute ob9.98pct=25.8435995896233. compute ob10.02pct=5. compute ob10.98pct=32.254499487029. compute ok1.02pct=5.79754799999998. compute ok1.98pct=308.225399384435. compute ok2.02pct=2. compute ok2.98pct=187.708998974058. compute ok3.02pct=3. compute ok3.98pct=165.367521618057. compute ok4.02pct=2. compute ok4.98pct=205.385197127363. end if. do if lregion=3 and IMETHOD=2. compute oa1.02pct=18.2947808881723. compute oa1.98pct=107.32574527334. compute oa2.02pct=10.637072. compute oa2.98pct=97.9759343397331. compute oa3.02pct=14.659042. compute oa3.98pct=93.3352484244467. compute oa4.02pct=13.7613272. compute oa4.98pct=108.45149054668. compute oa5.02pct=12. compute oa5.98pct=96.6871544775025. compute oa6.02pct=8.37520519999998. compute oa6.98pct=72.8152484244467. compute oa7.02pct=6.40249719999997. compute oa7.98pct=49.6297651326395. compute oa8.02pct=6.40249719999999. compute oa8.98pct=49.9038121061116. compute oa9.02pct=6. compute oa9.98pct=49.9038121061116. compute oa10.02pct=6. compute oa10.98pct=49.4219060530558. compute oa11.02pct=41.1673904440861. compute oa11.98pct=201.77149054668. compute oa12.02pct=5.6754172. compute oa12.98pct=85.5816711856954. compute oa13.02pct=8.40249719999998. compute oa13.98pct=58.2014634325072. compute oa14.02pct=7.36175119999995. compute oa14.98pct=58.3752484244466. compute ob1.02pct=24.5619759999999. compute ob1.98pct=214.243866334456. compute ob2.02pct=7.67541719999998. compute ob2.98pct=46.9038121061116. compute ob3.02pct=4.68087559999999. compute ob3.98pct=112.265718159167. compute ob4.02pct=7.37266799999996. compute ob4.98pct=66.7838121061117. compute ob5.02pct=3.413414. compute ob5.98pct=34.1757181591675. compute ob6.02pct=1.68633399999999. compute ob6.98pct=17.3919060530558. compute ob7.02pct=4.41341399999998. compute ob7.98pct=51.7433423713908. compute ob8.02pct=7.35083439999997. compute ob8.98pct=86.4609936977863. compute ob9.02pct=4. compute ob9.98pct=35.7504968488932. compute ob10.02pct=3.413414. compute ob10.98pct=28.3919060530558. compute ok1.02pct=10.413414. compute ok1.98pct=96.9595302652791. compute ok2.02pct=5.83826029605744. compute ok2.98pct=54.7838121061117. compute ok3.02pct=7.413414. compute ok3.98pct=56.1757181591675. compute ok4.02pct=9.41341399999997. compute ok4.98pct=58.3919060530558. end if. /** 2.3. COMPUTING THE TRIMMED VARIABLES. /** NB: The suffix "tmd" is used for "trimmed". compute oa1.tmd=oa1.tim. if oa1.tim lt oa1.02pct oa1.tmd=oa1.02pct. if oa1.tim gt oa1.98pct oa1.tmd=oa1.98pct. compute oa2.tmd=oa2.tim. if oa2.tim lt oa2.02pct oa2.tmd=oa2.02pct. if oa2.tim gt oa2.98pct oa2.tmd=oa2.98pct. compute oa3.tmd=oa3.tim. if oa3.tim lt oa3.02pct oa3.tmd=oa3.02pct. if oa3.tim gt oa3.98pct oa3.tmd=oa3.98pct. compute oa4.tmd=oa4.tim. if oa4.tim lt oa4.02pct oa4.tmd=oa4.02pct. if oa4.tim gt oa4.98pct oa4.tmd=oa4.98pct. compute oa5.tmd=oa5.tim. if oa5.tim lt oa5.02pct oa5.tmd=oa5.02pct. if oa5.tim gt oa5.98pct oa5.tmd=oa5.98pct. compute oa6.tmd=oa6.tim. if oa6.tim lt oa6.02pct oa6.tmd=oa6.02pct. if oa6.tim gt oa6.98pct oa6.tmd=oa6.98pct. compute oa7.tmd=oa7.tim. if oa7.tim lt oa7.02pct oa7.tmd=oa7.02pct. if oa7.tim gt oa7.98pct oa7.tmd=oa7.98pct. compute oa8.tmd=oa8.tim. if oa8.tim lt oa8.02pct oa8.tmd=oa8.02pct. if oa8.tim gt oa8.98pct oa8.tmd=oa8.98pct. compute oa9.tmd=oa9.tim. if oa9.tim lt oa9.02pct oa9.tmd=oa9.02pct. if oa9.tim gt oa9.98pct oa9.tmd=oa9.98pct. compute oa10.tmd=oa10.tim. if oa10.tim lt oa10.02pct oa10.tmd=oa10.02pct. if oa10.tim gt oa10.98pct oa10.tmd=oa10.98pct. compute oa11.tmd=oa11.tim. if oa11.tim lt oa11.02pct oa11.tmd=oa11.02pct. if oa11.tim gt oa11.98pct oa11.tmd=oa11.98pct. compute oa12.tmd=oa12.tim. if oa12.tim lt oa12.02pct oa12.tmd=oa12.02pct. if oa12.tim gt oa12.98pct oa12.tmd=oa12.98pct. compute oa13.tmd=oa13.tim. if oa13.tim lt oa13.02pct oa13.tmd=oa13.02pct. if oa13.tim gt oa13.98pct oa13.tmd=oa13.98pct. compute oa14.tmd=oa14.tim. if oa14.tim lt oa14.02pct oa14.tmd=oa14.02pct. if oa14.tim gt oa14.98pct oa14.tmd=oa14.98pct. compute ob1.tmd=ob1.tim. if ob1.tim lt ob1.02pct ob1.tmd=ob1.02pct. if ob1.tim gt ob1.98pct ob1.tmd=ob1.98pct. compute ob2.tmd=ob2.tim. if ob2.tim lt ob2.02pct ob2.tmd=ob2.02pct. if ob2.tim gt ob2.98pct ob2.tmd=ob2.98pct. compute ob3.tmd=ob3.tim. if ob3.tim lt ob3.02pct ob3.tmd=ob3.02pct. if ob3.tim gt ob3.98pct ob3.tmd=ob3.98pct. compute ob4.tmd=ob4.tim. if ob4.tim lt ob4.02pct ob4.tmd=ob4.02pct. if ob4.tim gt ob4.98pct ob4.tmd=ob4.98pct. compute ob5.tmd=ob5.tim. if ob5.tim lt ob5.02pct ob5.tmd=ob5.02pct. if ob5.tim gt ob5.98pct ob5.tmd=ob5.98pct. compute ob6.tmd=ob6.tim. if ob6.tim lt ob6.02pct ob6.tmd=ob6.02pct. if ob6.tim gt ob6.98pct ob6.tmd=ob6.98pct. compute ob7.tmd=ob7.tim. if ob7.tim lt ob7.02pct ob7.tmd=ob7.02pct. if ob7.tim gt ob7.98pct ob7.tmd=ob7.98pct. compute ob8.tmd=ob8.tim. if ob8.tim lt ob8.02pct ob8.tmd=ob8.02pct. if ob8.tim gt ob8.98pct ob8.tmd=ob8.98pct. compute ob9.tmd=ob9.tim. if ob9.tim lt ob9.02pct ob9.tmd=ob9.02pct. if ob9.tim gt ob9.98pct ob9.tmd=ob9.98pct. compute ob10.tmd=ob10.tim. if ob10.tim lt ob10.02pct ob10.tmd=ob10.02pct. if ob10.tim gt ob10.98pct ob10.tmd=ob10.98pct. compute ok1.tmd=ok1.tim. if ok1.tim lt ok1.02pct ok1.tmd=ok1.02pct. if ok1.tim gt ok1.98pct ok1.tmd=ok1.98pct. compute ok2.tmd=ok2.tim. if ok2.tim lt ok2.02pct ok2.tmd=ok2.02pct. if ok2.tim gt ok2.98pct ok2.tmd=ok2.98pct. compute ok3.tmd=ok3.tim. if ok3.tim lt ok3.02pct ok3.tmd=ok3.02pct. if ok3.tim gt ok3.98pct ok3.tmd=ok3.98pct. compute ok4.tmd=ok4.tim. if ok4.tim lt ok4.02pct ok4.tmd=ok4.02pct. if ok4.tim gt ok4.98pct ok4.tmd=ok4.98pct. /** 2.4. COMPUTING THE TRIMMED SUMMARY STATISTICS (MEAN, MEDIAN, SD). AGGREGATE /OUTFILE=* MODE=ADDVARIABLES /BREAK=lregion IMETHOD /oa1.timM=MEAN(oa1.tmd) /oa1.timMD=MEDIAN(oa1.tmd) /oa1.timSD=SD(oa1.tmd) /oa2.timM=MEAN(oa2.tmd) /oa2.timMD=MEDIAN(oa2.tmd) /oa2.timSD=SD(oa2.tmd) /oa3.timM=MEAN(oa3.tmd) /oa3.timMD=MEDIAN(oa3.tmd) /oa3.timSD=SD(oa3.tmd) /oa4.timM=MEAN(oa4.tmd) /oa4.timMD=MEDIAN(oa4.tmd) /oa4.timSD=SD(oa4.tmd) /oa5.timM=MEAN(oa5.tmd) /oa5.timMD=MEDIAN(oa5.tmd) /oa5.timSD=SD(oa5.tmd) /oa6.timM=MEAN(oa6.tmd) /oa6.timMD=MEDIAN(oa6.tmd) /oa6.timSD=SD(oa6.tmd) /oa7.timM=MEAN(oa7.tmd) /oa7.timMD=MEDIAN(oa7.tmd) /oa7.timSD=SD(oa7.tmd) /oa8.timM=MEAN(oa8.tmd) /oa8.timMD=MEDIAN(oa8.tmd) /oa8.timSD=SD(oa8.tmd) /oa9.timM=MEAN(oa9.tmd) /oa9.timMD=MEDIAN(oa9.tmd) /oa9.timSD=SD(oa9.tmd) /oa10.timM=MEAN(oa10.tmd) /oa10.timMD=MEDIAN(oa10.tmd) /oa10.timSD=SD(oa10.tmd) /oa11.timM=MEAN(oa11.tmd) /oa11.timMD=MEDIAN(oa11.tmd) /oa11.timSD=SD(oa11.tmd) /oa12.timM=MEAN(oa12.tmd) /oa12.timMD=MEDIAN(oa12.tmd) /oa12.timSD=SD(oa12.tmd) /oa13.timM=MEAN(oa13.tmd) /oa13.timMD=MEDIAN(oa13.tmd) /oa13.timSD=SD(oa13.tmd) /oa14.timM=MEAN(oa14.tmd) /oa14.timMD=MEDIAN(oa14.tmd) /oa14.timSD=SD(oa14.tmd) /ob1.timM=MEAN(ob1.tmd) /ob1.timMD=MEDIAN(ob1.tmd) /ob1.timSD=SD(ob1.tmd) /ob2.timM=MEAN(ob2.tmd) /ob2.timMD=MEDIAN(ob2.tmd) /ob2.timSD=SD(ob2.tmd) /ob3.timM=MEAN(ob3.tmd) /ob3.timMD=MEDIAN(ob3.tmd) /ob3.timSD=SD(ob3.tmd) /ob4.timM=MEAN(ob4.tmd) /ob4.timMD=MEDIAN(ob4.tmd) /ob4.timSD=SD(ob4.tmd) /ob5.timM=MEAN(ob5.tmd) /ob5.timMD=MEDIAN(ob5.tmd) /ob5.timSD=SD(ob5.tmd) /ob6.timM=MEAN(ob6.tmd) /ob6.timMD=MEDIAN(ob6.tmd) /ob6.timSD=SD(ob6.tmd) /ob7.timM=MEAN(ob7.tmd) /ob7.timMD=MEDIAN(ob7.tmd) /ob7.timSD=SD(ob7.tmd) /ob8.timM=MEAN(ob8.tmd) /ob8.timMD=MEDIAN(ob8.tmd) /ob8.timSD=SD(ob8.tmd) /ob9.timM=MEAN(ob9.tmd) /ob9.timMD=MEDIAN(ob9.tmd) /ob9.timSD=SD(ob9.tmd) /ob10.timM=MEAN(ob10.tmd) /ob10.timMD=MEDIAN(ob10.tmd) /ob10.timSD=SD(ob10.tmd) /ok1.timM=MEAN(ok1.tmd) /ok1.timMD=MEDIAN(ok1.tmd) /ok1.timSD=SD(ok1.tmd) /ok2.timM=MEAN(ok2.tmd) /ok2.timMD=MEDIAN(ok2.tmd) /ok2.timSD=SD(ok2.tmd) /ok3.timM=MEAN(ok3.tmd) /ok3.timMD=MEDIAN(ok3.tmd) /ok3.timSD=SD(ok3.tmd) /ok4.timM=MEAN(ok4.tmd) /ok4.timMD=MEDIAN(ok4.tmd) /ok4.timSD=SD(ok4.tmd). /** 2.5. CORRECTING FOR EXTREMELY LONG RESPONSE TIMES (timR). compute oa1.timR=oa1.tim. compute oa2.timR=oa2.tim. compute oa3.timR=oa3.tim. compute oa4.timR=oa4.tim. compute oa5.timR=oa5.tim. compute oa6.timR=oa6.tim. compute oa7.timR=oa7.tim. compute oa8.timR=oa8.tim. compute oa9.timR=oa9.tim. compute oa10.timR=oa10.tim. compute oa11.timR=oa11.tim. compute oa12.timR=oa12.tim. compute oa13.timR=oa13.tim. compute oa14.timR=oa14.tim. compute ob1.timR=ob1.tim. compute ob2.timR=ob2.tim. compute ob3.timR=ob3.tim. compute ob4.timR=ob4.tim. compute ob5.timR=ob5.tim. compute ob6.timR=ob6.tim. compute ob7.timR=ob7.tim. compute ob8.timR=ob8.tim. compute ob9.timR=ob9.tim. compute ob10.timR=ob10.tim. compute ok1.timR=ok1.tim. compute ok2.timR=ok2.tim. compute ok3.timR=ok3.tim. compute ok4.timR=ok4.tim. if oa1.tim gt (oa1.timM+3*(oa1.timSD)) oa1.timR=oa1.timMD. if oa2.tim gt (oa2.timM+3*(oa2.timSD)) oa2.timR=oa2.timMD. if oa3.tim gt (oa3.timM+3*(oa3.timSD)) oa3.timR=oa3.timMD. if oa4.tim gt (oa4.timM+3*(oa4.timSD)) oa4.timR=oa4.timMD. if oa5.tim gt (oa5.timM+3*(oa5.timSD)) oa5.timR=oa5.timMD. if oa6.tim gt (oa6.timM+3*(oa6.timSD)) oa6.timR=oa6.timMD. if oa7.tim gt (oa7.timM+3*(oa7.timSD)) oa7.timR=oa7.timMD. if oa8.tim gt (oa8.timM+3*(oa8.timSD)) oa8.timR=oa8.timMD. if oa9.tim gt (oa9.timM+3*(oa9.timSD)) oa9.timR=oa9.timMD. if oa10.tim gt (oa10.timM+3*(oa10.timSD)) oa10.timR=oa10.timMD. if oa11.tim gt (oa11.timM+3*(oa11.timSD)) oa11.timR=oa11.timMD. if oa12.tim gt (oa12.timM+3*(oa12.timSD)) oa12.timR=oa12.timMD. if oa13.tim gt (oa13.timM+3*(oa13.timSD)) oa13.timR=oa13.timMD. if oa14.tim gt (oa14.timM+3*(oa14.timSD)) oa14.timR=oa14.timMD. if ob1.tim gt (ob1.timM+3*(ob1.timSD)) ob1.timR=ob1.timMD. if ob2.tim gt (ob2.timM+3*(ob2.timSD)) ob2.timR=ob2.timMD. if ob3.tim gt (ob3.timM+3*(ob3.timSD)) ob3.timR=ob3.timMD. if ob4.tim gt (ob4.timM+3*(ob4.timSD)) ob4.timR=ob4.timMD. if ob5.tim gt (ob5.timM+3*(ob5.timSD)) ob5.timR=ob5.timMD. if ob6.tim gt (ob6.timM+3*(ob6.timSD)) ob6.timR=ob6.timMD. if ob7.tim gt (ob7.timM+3*(ob7.timSD)) ob7.timR=ob7.timMD. if ob8.tim gt (ob8.timM+3*(ob8.timSD)) ob8.timR=ob8.timMD. if ob9.tim gt (ob9.timM+3*(ob9.timSD)) ob9.timR=ob9.timMD. if ob10.tim gt (ob10.timM+3*(ob10.timSD)) ob10.timR=ob10.timMD. if ok1.tim gt (ok1.timM+3*(ok1.timSD)) ok1.timR=ok1.timMD. if ok2.tim gt (ok2.timM+3*(ok2.timSD)) ok2.timR=ok2.timMD. if ok3.tim gt (ok3.timM+3*(ok3.timSD)) ok3.timR=ok3.timMD. if ok4.tim gt (ok4.timM+3*(ok4.timSD)) ok4.timR=ok4.timMD. /** 2.6. CREATING NEW RTS (fR) INCLUDING DK/NA AS THE MEDIAN. count OMISS=oa1.timR to oa7.timR,oa12.timR to oa14.timR,ob1.timR to ok4.timR (missing). do if OMISS < 14. /** NB: only about 0.2% of cases do not satisfy this condition. compute oa1.fR=oa1.timR. recode f15420 (1=1)(else=copy) into f15420NMV. if any(f15420NMV,8,9) oa1.fR=oa1.timMD. compute oa2.fR=oa2.timR. recode f15430 (1=1)(else=copy) into f15430NMV. if any(f15430NMV,8,9) oa2.fR=oa2.timMD. compute oa3.fR=oa3.timR. recode f15440 (1=1)(else=copy) into f15440NMV. if any(f15440NMV,8,9) oa3.fR=oa3.timMD. compute oa4.fR=oa4.timR. recode f15470 (1=1)(else=copy) into f15470NMV. if any(f15470NMV,8,9) oa4.fR=oa4.timMD. compute oa5.fR=oa5.timR. recode f15480 (1=1)(else=copy) into f15480NMV. if any(f15480NMV,8,9) oa5.fR=oa5.timMD. compute oa6.fR=oa6.timR. recode f15490 (1=1)(else=copy) into f15490NMV. if any(f15490NMV,8,9) oa6.fR=oa6.timMD. compute oa7.fR=oa7.timR. recode f15760 (1=1)(else=copy) into f15760NMV. if any(f15760NMV,8,9) oa7.fR=oa7.timMD. compute oa8.fR=oa8.timR. recode f15761 (1=1)(else=copy) into f15761NMV. if any(f15761NMV,8,9) oa8.fR=oa8.timMD. compute oa9.fR=oa9.timR. recode f15762 (1=1)(else=copy) into f15762NMV. if any(f15762NMV,8,9) oa9.fR=oa9.timMD. compute oa10.fR=oa10.timR. recode f15763 (1=1)(else=copy) into f15763NMV. if any(f15763NMV,8,9) oa10.fR=oa10.timMD. compute oa11.fR=oa11.timR. count oa11fRmiss=f15801,f15802,f15803,f15804 (8,9). count oa11fRsysmis=f15801,f15802,f15803,f15804 (sysmis). if (oa11fRsysmis NE 4) AND (oa11fRmiss gt 1) oa11.fR=oa11.timMD. compute oa12.fR=oa12.timR. recode f10100 (1=1)(else=copy) into f10100NMV. if any(f10100NMV,8,9) oa12.fR=oa12.timMD. compute oa13.fR=oa13.timR. recode f13700 (1=1)(else=copy) into f13700NMV. if any(f13700NMV,8,9) oa13.fR=oa13.timMD. compute oa14.fR=oa14.timR. recode f14600 (1=1)(else=copy) into f14600NMV. if any(f14600NMV,8,9) oa14.fR=oa14.timMD. compute ob1.fR=ob1.timR. count ob1fRmiss=f13401,f13402,f13403,f13404,f13405 (8,9). if ob1fRmiss gt 1 ob1.fR=ob1.timMD. compute ob2.fR=ob2.timR. recode f13300 (1=1)(else=copy) into f13300NMV. if any(f13300NMV,8,9) ob2.fR=ob2.timMD. compute ob3.fR=ob3.timR. recode f11100 (1=1)(else=copy) into f11100NMV. if any(f11100NMV,8,9) ob3.fR=ob3.timMD. compute ob4.fR=ob4.timR. recode f21310 (1=1)(else=copy) into f21310NMV. if any(f21310NMV,98,99) ob4.fR=ob4.timMD. compute ob5.fR=ob5.timR. recode f20600 (1=1)(else=copy) into f20600NMV. if any(f20600NMV,8,9) ob5.fR=ob5.timMD. compute ob6.fR=ob6.timR. recode f20601 (1=1)(else=copy) into f20601NMV. if any(f20601NMV,8,9) ob6.fR=ob6.timMD. compute ob7.fR=ob7.timR. recode f20602 (1=1)(else=copy) into f20602NMV. if any(f20602NMV,8,9) ob7.fR=ob7.timMD. compute ob8.fR=ob8.timR. recode f20300 (1=1)(else=copy) into f20300NMV. if any(f20300NMV,99) ob8.fR=ob8.timMD. compute ob9.fR=ob9.timR. recode f20500 (1=1)(else=copy) into f20500NMV. if any(f20500NMV,99) ob9.fR=ob9.timMD. compute ob10.fR=ob10.timR. recode f20210 (1=1)(else=copy) into f20210NMV. if any(f20210NMV,9998,9999) ob10.fR=ob10.timMD. end if. compute ok1.fR=ok1.timR. compute ok2.fR=ok2.timR. compute ok3.fR=ok3.timR. compute ok4.fR=ok4.timR. /** 3. KNOWLEDGE SCORES recode f15900r f16000r f16100r f16300r (8,9=0)(else=copy) into know1,know2,know3,know4. value labels know1,know2,know3,know4 0"wrong" 1"correct". compute knowtot=sum(know1 to know4). /** 4. PREDICTORS. /** AGE AND GENDER. compute age=2015-birthyear_sample. compute age_sq=age*age. compute age_cb=age*age*age. recode age (18 thru 24=1)(25 thru 34=2)(35 thru 44=3)(45 thru 54=4)(55 thru 64=5)(65 thru 74=6)(75 thru highest=7) into age_group. val lab age_group 1"18-24" 2"25-34" 3"35-44" 4"45-54" 5"55-64" 6"65-74" 7"75+". recode age_group (1,2,3=1)(4,5=2)(6,7=3) into age_group3. val lab age_group3 1"18-44" 2"45-64" 3"65+". recode age (18 thru 35=1)(36 thru 50=2)(51 thru 65=3)(66 thru highest=4) into age_group4. val lab age_group4 1"18-35" 2"36-50" 3"51-65" 4"66+". recode gender (1=0)(2=1) into female. compute female_age=female*age. /** CAWI AND INTERNET. recode IMETHOD (1=1)(2=0) into CAWI. val lab CAWI 0"CATI" 1"CAWI". recode f20602 (1=2)(2=1)(3,4,5=0) into internet. val lab internet 0"occasional/no use" 1"regular use" 2"daily use". recode f20602 (1=3)(2=2)(3=1)(4,5=0) into internet4. val lab internet4 0"no access/never" 1"once/wk or less" 2"more than once/wk" 3"everyday". /** EDUCATION. recode educ_rr (sysmis=4)(else=copy) into education. val lab education 1"compulsory" 2"basic vocational" 3"vocational" 4"diploma school" 5"high school" 6"higher vocational" 7"vocational university" 8"university". recode education (1,2=1)(3=2)(4=3)(5=4)(6=5)(7=6)(8=7) into education7. val lab education7 1"compulsory/basic vocational" 2"vocational" 3"diploma school" 4"high school" 5"higher vocational" 6"vocational university" 7"university". recode education7 (1,2=1)(3,4,5=2)(6,7=3) into education3. val lab education3 1"low" 2"middle" 3"high". /** POLITICAL INTEREST. recode f10100 (1=3)(2=2)(3=1)(4=0) into polint. val lab polint 0"not interested at all" 1"rather not interested" 2"rather interested" 3"very interested". /** INTENSITY OF POLITICAL PREFERENCES. recode f15200 (0,10=5)(1,9=4)(2,8=3)(3,7=2)(4,6=1)(else=0) into LRint. /** TIME ELAPSED BETWEEN ELECTION AND INTERVIEW. compute datelection=date.dmy(18,10,2015). formats datelection(date11). compute dayselec=datediff(intdate,datelection,'days'). /** PROFESSIONAL ROLE. recode situar (1=1)(else=0) into PRexecutive. recode situar (2=1)(else=0) into PRsupervisory. recode situar (3=1)(else=0) into PRoperative. recode situar (4=1)(else=0) into PRselfemployed. recode f21400 (3,5,6=1)(else=0) into PRinactive. compute PRmissing=1. if sum(PRexecutive to PRinactive) > 0 PRmissing=0. /** NB: Categories 3, 5, and 6 for f21400 (i.e., in training, pensioner for age or health reasons) are missing for situar; /** in contrast, categories 4 and 7 for f21400 (i.e., homemaker and unemployed) do have a valid code in the situar variable and are coded accordingly. /** SWISS VS. FOREIGN ORIGINS. recode f20210 (8100=0)(8200 thru 9997=1) into foreign. val lab foreign 0"Swiss at birth" 1"foreigner at birth". /** RELIGION. recode f20760 (1=1)(2 thru 6=0) into relig_prot. recode f20760 (2=1)(1,3,4,5,6=0) into relig_cath. recode f20760 (3,4,5=1)(1,2,6=0) into relig_oth. recode f20760 (6=1)(1 thru 5=0) into relig_none. save outfile="S2015_RT2.sav"/keep=all. /** ========================================================================================================. get file="S2015_RT2.sav"/keep=all. /** CREATING NEW VARIABLES AT THE ITEM LEVEL (cf. Yan & Tourangeau 2008). compute oa1.MV= 4.8. compute oa2.MV= 4.8. compute oa3.MV= 2.7. compute oa4.MV= 3.9. compute oa5.MV= 3.9. compute oa6.MV= 3.7. compute oa7.MV= 3.6. compute oa8.MV= 3.9. compute oa9.MV= 3.4. compute oa10.MV= 3.1. compute oa11.MV= 5.7. compute oa12.MV= 0.6. compute oa13.MV= 2. compute oa14.MV= 1.7. compute ob1.MV= 4.2. compute ob2.MV= 0.6. compute ob3.MV= 0.8. compute ob4.MV= 1.1. compute ob5.MV= 0.6. compute ob6.MV= 0.3. compute ob7.MV= 0.3. compute ob8.MV= 0.5. compute ob9.MV= 0.3. compute ob10.MV= 0.3. compute ok1.MV= 8. compute ok2.MV= 8.5. compute ok3.MV= 11.2. compute ok4.MV= 17.1. compute oa1.POS= 76. compute oa2.POS= 77. compute oa3.POS= 78. compute oa4.POS= 79. compute oa5.POS= 80. compute oa6.POS= 81. compute oa7.POS= 82. compute oa8.POS= 83. compute oa9.POS= 84. compute oa10.POS= 85. compute oa11.POS= 86. compute oa12.POS= 4. compute oa13.POS= 53. compute oa14.POS= 61. compute ob1.POS= 52. compute ob2.POS= 46. compute ob3.POS= 10. compute ob4.POS= 112. compute ob5.POS= 98. compute ob6.POS= 99. compute ob7.POS= 100. compute ob8.POS= 101. compute ob9.POS= 102. compute ob10.POS= 105. compute ok1.POS= 94. compute ok2.POS= 95. compute ok3.POS= 96. compute ok4.POS= 97. compute oa1.SCA= 1. compute oa2.SCA= 1. compute oa3.SCA= 1. compute oa4.SCA= 1. compute oa5.SCA= 1. compute oa6.SCA= 1. compute oa7.SCA= 0. compute oa8.SCA= 0. compute oa9.SCA= 0. compute oa10.SCA= 0. compute oa11.SCA= 1. compute oa12.SCA= 1. compute oa13.SCA= 1. compute oa14.SCA= 1. compute ob1.SCA= 1. compute ob2.SCA= 0. compute ob3.SCA= 0. compute ob4.SCA= 1. compute ob5.SCA= 0. compute ob6.SCA= 0. compute ob7.SCA= 1. compute ob8.SCA= 0. compute ob9.SCA= 0. compute ob10.SCA= 0. compute ok1.SCA= 0. compute ok2.SCA= 0. compute ok3.SCA= 0. compute ok4.SCA= 0. do if IMETHOD=1. compute oa1.WRD= 31. compute oa2.WRD= 17. compute oa3.WRD= 24. compute oa4.WRD= 33. compute oa5.WRD= 29. compute oa6.WRD= 17. compute oa7.WRD= 36. compute oa8.WRD= 3. compute oa9.WRD= 4. compute oa10.WRD= 4. compute oa11.WRD= 96. compute oa12.WRD= 12. compute oa13.WRD= 16. compute oa14.WRD= 14. compute ob1.WRD= 15. compute ob2.WRD= 32. compute ob3.WRD= 31. compute ob4.WRD= 23. compute ob5.WRD= 22. compute ob6.WRD= 7. compute ob7.WRD= 18. compute ob8.WRD= 12. compute ob9.WRD= 9. compute ob10.WRD= 7. compute ok1.WRD= 27. compute ok2.WRD= 11. compute ok3.WRD= 14. compute ok4.WRD= 16. end if. do if IMETHOD=2. compute oa1.WRD= 44. compute oa2.WRD= 33. compute oa3.WRD= 35. compute oa4.WRD= 43. compute oa5.WRD= 35. compute oa6.WRD= 25. compute oa7.WRD= 36. compute oa8.WRD= 3. compute oa9.WRD= 4. compute oa10.WRD= 4. compute oa11.WRD= 96. compute oa12.WRD= 12. compute oa13.WRD= 25. compute oa14.WRD= 25. compute ob1.WRD= 43. compute ob2.WRD= 27. compute ob3.WRD= 31. compute ob4.WRD= 23. compute ob5.WRD= 22. compute ob6.WRD= 7. compute ob7.WRD= 18. compute ob8.WRD= 12. compute ob9.WRD= 9. compute ob10.WRD= 7. compute ok1.WRD= 18. compute ok2.WRD= 13. compute ok3.WRD= 16. compute ok4.WRD= 17. end if. compute oa1.CTG= 5. compute oa2.CTG= 5. compute oa3.CTG= 5. compute oa4.CTG= 5. compute oa5.CTG= 5. compute oa6.CTG= 5. compute oa7.CTG= 2. compute oa8.CTG= 2. compute oa9.CTG= 2. compute oa10.CTG= 2. compute oa11.CTG= 4. compute oa12.CTG= 4. compute oa13.CTG= 4. compute oa14.CTG= 5. compute ob1.CTG= 20. compute ob2.CTG= 2. compute ob3.CTG= 4. compute ob4.CTG= 13. compute ob5.CTG= 2. compute ob6.CTG= 2. compute ob7.CTG= 4. compute ob8.CTG= 15. compute ob9.CTG= 10. compute ob10.CTG= 8. compute ok1.CTG= 5. compute ok2.CTG= 5. compute ok3.CTG= 5. compute ok4.CTG= 5. compute oa1.TYP= 0. compute oa2.TYP= 0. compute oa3.TYP= 0. compute oa4.TYP= 0. compute oa5.TYP= 0. compute oa6.TYP= 0. compute oa7.TYP= 0. compute oa8.TYP= 0. compute oa9.TYP= 0. compute oa10.TYP= 0. compute oa11.TYP= 0. compute oa12.TYP= 0. compute oa13.TYP= 0. compute oa14.TYP= 0. compute ob1.TYP= 1. compute ob2.TYP= 1. compute ob3.TYP= 1. compute ob4.TYP= 1. compute ob5.TYP= 1. compute ob6.TYP= 1. compute ob7.TYP= 1. compute ob8.TYP= 1. compute ob9.TYP= 1. compute ob10.TYP= 1. compute ok1.TYP= 2. compute ok2.TYP= 2. compute ok3.TYP= 2. compute ok4.TYP= 2. compute oa1.KNO= 99. compute oa2.KNO= 99. compute oa3.KNO= 99. compute oa4.KNO= 99. compute oa5.KNO= 99. compute oa6.KNO= 99. compute oa7.KNO= 99. compute oa8.KNO= 99. compute oa9.KNO= 99. compute oa10.KNO= 99. compute oa11.KNO= 99. compute oa12.KNO= 99. compute oa13.KNO= 99. compute oa14.KNO= 99. compute ob1.KNO= 99. compute ob2.KNO= 99. compute ob3.KNO= 99. compute ob4.KNO= 99. compute ob5.KNO= 99. compute ob6.KNO= 99. compute ob7.KNO= 99. compute ob8.KNO= 99. compute ob9.KNO= 99. compute ob10.KNO= 99. compute ok1.KNO= know1. compute ok2.KNO= know2. compute ok3.KNO= know3. compute ok4.KNO= know4. missing values oa1.KNO to ok4.KNO (99). compute oa1.KNOW=99. compute oa2.KNOW= 99. compute oa3.KNOW= 99. compute oa4.KNOW= 99. compute oa5.KNOW= 99. compute oa6.KNOW= 99. compute oa7.KNOW= 99. compute oa8.KNOW= 99. compute oa9.KNOW= 99. compute oa10.KNOW= 99. compute oa11.KNOW= 99. compute oa12.KNOW= 99. compute oa13.KNOW= 99. compute oa14.KNOW= 99. compute ob1.KNOW= 99. compute ob2.KNOW= 99. compute ob3.KNOW= 99. compute ob4.KNOW= 99. compute ob5.KNOW= 99. compute ob6.KNOW= 99. compute ob7.KNOW= 99. compute ob8.KNOW= 99. compute ob9.KNOW= 99. compute ob10.KNOW= 99. compute ok1.KNOW=f15900. compute ok2.KNOW=f16000. compute ok3.KNOW=f16100. compute ok4.KNOW=f16300. missing values oa1.KNOW to ok4.KNOW (8,9,99). varstocases /make RTR from oa1.fR oa2.fR oa3.fR oa4.fR oa5.fR oa6.fR oa7.fR oa8.fR oa9.fR oa10.fR oa11.fR oa12.fR oa13.fR oa14.fR ob1.fR ob2.fR ob3.fR ob4.fR ob5.fR ob6.fR ob7.fR ob8.fR ob9.fR ob10.fR ok1.fR ok2.fR ok3.fR ok4.fR /make ITEM.MV from oa1.MV to ok4.MV /make ITEM.POS from oa1.POS to ok4.POS /make ITEM.SCA from oa1.SCA to ok4.SCA /make ITEM.WRD from oa1.WRD to ok4.WRD /make ITEM.CTG from oa1.CTG to ok4.CTG /make ITEM.TYP from oa1.TYP to ok4.TYP /make ITEM.KNO from oa1.KNO to ok4.KNO /make ITEM.KNOW from oa1.KNOW to ok4.KNOW /index trial. save outfile="S2015_RT3.sav"/COMPRESSED/ keep=userid oa1.timR oa2.timR oa3.timR oa4.timR oa5.timR oa6.timR oa7.timR oa8.timR oa9.timR oa10.timR oa11.timR oa12.timR oa13.timR oa14.timR ob1.timR ob2.timR ob3.timR ob4.timR ob5.timR ob6.timR ob7.timR ob8.timR ob9.timR ob10.timR ok1.timR ok2.timR ok3.timR ok4.timR trial RTR ITEM.MV ITEM.POS ITEM.SCA ITEM.WRD ITEM.CTG ITEM.TYP ITEM.KNO ITEM.KNOW CAWI internet education lregion age age_sq female polint LRint dayselec foreign PRexecutive PRsupervisory PRoperative PRselfemployed PRinactive PRmissing relig_prot relig_cath relig_oth relig_none. /** ========================================================================================================. get file="S2015_RT3.sav"/keep=all. /** Exclude items oa8, oa9, oa10 (perfectly correlated to oa7) and oa11 (missing values = 6.0%, i.e. > 5%); this is done via the commands temporary / select if. /** Take away weighting of cases. WEIGHT OFF. /** ==========================================. /** ESTIMATION OF CROSS-CLASSIFIED MODELS OF RTS. /** ==========================================. /** NB: Because the SPSS MIXED procedure takes a very long time to converge (i.e., hours or days, depending on one's system), we stopped the estimation /** after 4 iterations. However, we estimated the same models using the (much quicker) Stata MIXED procedure and got exactly the same results. As a matter of fact, /** residuals from the SPSS MIXED procedure (with incomplete convergence) were identical to residuals from the full Stata MIXED procedure (r>0.99999). /** The results of the Stata analysis can be obtained upon request from the authors. /** MODEL 1-0: Null model for RT. /** @@@ TABLE A.2 -- Model 0 (empty model, not shown) for Model 1 @@@. /** NB: The empty model is necessarty to compute the "Percent variance explained" in Table A.2. temporary. select if not any(trial,8,9,10,11). MIXED RTR /CRITERIA=CIN(95) MXITER(4) MXSTEP(10) SCORING(1) SINGULAR(0.000000000001) HCONVERGE(0, ABSOLUTE) LCONVERGE(0, ABSOLUTE) PCONVERGE(0.000001, ABSOLUTE) /METHOD=REML /PRINT=G R SOLUTION TESTCOV CORB COVB CPS /RANDOM=INTERCEPT | SUBJECT(trial) COVTYPE(ID) /RANDOM=INTERCEPT | SUBJECT(userid) COVTYPE(ID) /SAVE=PRED RESID. /** MODEL 1-1: Full cross-classified model for both CATI & CATI (with interactions) for RTR. /** @@@ TABLE A.2 -- Model 1 @@@. temporary. select if not any(trial,8,9,10,11). MIXED RTR BY CAWI lregion foreign ITEM.SCA ITEM.TYP female WITH internet education age polint LRint dayselec PRexecutive PRsupervisory PRoperative PRselfemployed PRinactive PRmissing ITEM.MV ITEM.POS ITEM.WRD ITEM.CTG /CRITERIA=CIN(95) MXITER(4) MXSTEP(10) SCORING(1) SINGULAR(0.000000000001) HCONVERGE(0, ABSOLUTE) LCONVERGE(0, ABSOLUTE) PCONVERGE(0.000001, ABSOLUTE) /FIXED=CAWI education internet lregion female age polint LRint dayselec foreign PRexecutive PRsupervisory PRoperative PRselfemployed PRinactive PRmissing ITEM.MV ITEM.POS ITEM.SCA ITEM.WRD ITEM.CTG ITEM.TYP CAWI*internet ITEM.TYP*internet ITEM.TYP*CAWI | SSTYPE(3) /METHOD=REML /PRINT=G R SOLUTION TESTCOV CORB COVB CPS /RANDOM=INTERCEPT | SUBJECT(trial) COVTYPE(ID) /RANDOM=INTERCEPT | SUBJECT(userid) COVTYPE(ID) /SAVE=PRED RESID. /** MODEL 1-1B: Full cross-classified model for both CATI & CATI (with interactions), INCLUDING AGE SQUARED, for RTR. /** @@@ TABLE A.2 -- Model not shown; tested to respond to one reviewer's suggestion -- see Appendix footnote #6 @@@. temporary. select if not any(trial,8,9,10,11). MIXED RTR BY CAWI lregion foreign ITEM.SCA ITEM.TYP female WITH internet education age age_sq polint LRint dayselec PRexecutive PRsupervisory PRoperative PRselfemployed PRinactive PRmissing ITEM.MV ITEM.POS ITEM.WRD ITEM.CTG /CRITERIA=CIN(95) MXITER(4) MXSTEP(10) SCORING(1) SINGULAR(0.000000000001) HCONVERGE(0, ABSOLUTE) LCONVERGE(0, ABSOLUTE) PCONVERGE(0.000001, ABSOLUTE) /FIXED=CAWI education internet lregion female age age_sq polint LRint dayselec foreign PRexecutive PRsupervisory PRoperative PRselfemployed PRinactive PRmissing ITEM.MV ITEM.POS ITEM.SCA ITEM.WRD ITEM.CTG ITEM.TYP CAWI*internet ITEM.TYP*internet ITEM.TYP*CAWI | SSTYPE(3) /METHOD=REML /PRINT=G R SOLUTION TESTCOV CORB COVB CPS /RANDOM=INTERCEPT | SUBJECT(trial) COVTYPE(ID) /RANDOM=INTERCEPT | SUBJECT(userid) COVTYPE(ID) /SAVE=PRED RESID. /** MODEL 2-0: Null model for RTR among CATI respondents. /** CAWI and its interactions with other variables are removed and the model is tested separately for CAWI and CATI respondents. /** @@@ TABLE A.2 -- Model 0 (empty model, not shown) for Model 2 @@@. /** NB: The empty model is necessarty to compute the "Percent variance explained" in Table A.2. temporary. select if not any(trial,8,9,10,11) AND CAWI=0. MIXED RTR /CRITERIA=CIN(95) MXITER(4) MXSTEP(10) SCORING(1) SINGULAR(0.000000000001) HCONVERGE(0, ABSOLUTE) LCONVERGE(0, ABSOLUTE) PCONVERGE(0.000001, ABSOLUTE) /METHOD=REML /PRINT=G R SOLUTION TESTCOV CORB COVB CPS /RANDOM=INTERCEPT | SUBJECT(trial) COVTYPE(ID) /RANDOM=INTERCEPT | SUBJECT(userid) COVTYPE(ID) /SAVE=PRED RESID. /** MODEL 2-1: Full cross-classified model for RTR among CATI respondents. /** CAWI and its interactions with other variables are removed and the model is tested separately for CAWI and CATI respondents. /** @@@ TABLE A.2 -- Model 2 @@@. temporary. select if not any(trial,8,9,10,11) AND CAWI=0. MIXED RTR BY lregion foreign ITEM.SCA ITEM.TYP female WITH internet education age polint LRint dayselec PRexecutive PRsupervisory PRoperative PRselfemployed PRinactive PRmissing ITEM.MV ITEM.POS ITEM.WRD ITEM.CTG /CRITERIA=CIN(95) MXITER(4) MXSTEP(10) SCORING(1) SINGULAR(0.000000000001) HCONVERGE(0, ABSOLUTE) LCONVERGE(0, ABSOLUTE) PCONVERGE(0.000001, ABSOLUTE) /FIXED=education internet lregion female age polint LRint dayselec foreign PRexecutive PRsupervisory PRoperative PRselfemployed PRinactive PRmissing ITEM.MV ITEM.POS ITEM.SCA ITEM.WRD ITEM.CTG ITEM.TYP ITEM.TYP*internet | SSTYPE(3) /METHOD=REML /PRINT=G R SOLUTION TESTCOV CORB COVB CPS /RANDOM=INTERCEPT | SUBJECT(trial) COVTYPE(ID) /RANDOM=INTERCEPT | SUBJECT(userid) COVTYPE(ID) /SAVE=PRED RESID. /** MODEL 3-0: Null model for RTR among CAWI respondents. /** CAWI and its interactions with other variables are removed and the model is tested separately for CAWI and CATI respondents. /** @@@ TABLE A.2 -- Model 0 (empty model, not shown) for Model 3 @@@. /** NB: The empty model is necessarty to compute the "Percent variance explained" in Table A.2. temporary. select if not any(trial,8,9,10,11) AND CAWI=1. MIXED RTR /CRITERIA=CIN(95) MXITER(4) MXSTEP(10) SCORING(1) SINGULAR(0.000000000001) HCONVERGE(0, ABSOLUTE) LCONVERGE(0, ABSOLUTE) PCONVERGE(0.000001, ABSOLUTE) /METHOD=REML /PRINT=G R SOLUTION TESTCOV CORB COVB CPS /RANDOM=INTERCEPT | SUBJECT(trial) COVTYPE(ID) /RANDOM=INTERCEPT | SUBJECT(userid) COVTYPE(ID) /SAVE=PRED RESID. /** MODEL 3-1: Full cross-classified model for RTR among CAWI respondents. /** CAWI and its interactions with other variables are removed and the model is tested separately for CAWI and CATI respondents. /** @@@ TABLE A.2 -- Model 3 @@@. temporary. select if not any(trial,8,9,10,11) AND CAWI=1. MIXED RTR BY lregion foreign ITEM.SCA ITEM.TYP female WITH internet education age polint LRint dayselec PRexecutive PRsupervisory PRoperative PRselfemployed PRinactive PRmissing ITEM.MV ITEM.POS ITEM.WRD ITEM.CTG /CRITERIA=CIN(95) MXITER(4) MXSTEP(10) SCORING(1) SINGULAR(0.000000000001) HCONVERGE(0, ABSOLUTE) LCONVERGE(0, ABSOLUTE) PCONVERGE(0.000001, ABSOLUTE) /FIXED=education internet lregion female age polint LRint dayselec foreign PRexecutive PRsupervisory PRoperative PRselfemployed PRinactive PRmissing ITEM.MV ITEM.POS ITEM.SCA ITEM.WRD ITEM.CTG ITEM.TYP ITEM.TYP*internet | SSTYPE(3) /METHOD=REML /PRINT=G R SOLUTION TESTCOV CORB COVB CPS /RANDOM=INTERCEPT | SUBJECT(trial) COVTYPE(ID) /RANDOM=INTERCEPT | SUBJECT(userid) COVTYPE(ID) /SAVE=PRED RESID. /** Renaming predicted values and residuals. rename variables (PRED_1=PRED_10_RTR)(RESID_1=RESID_10_RTR) (PRED_2=PRED_11_RTR)(RESID_2=RESID_11_RTR) (PRED_3=PRED_11B_RTR)(RESID_3=RESID_11B_RTR) (PRED_4=PRED_20_RTR)(RESID_4=RESID_20_RTR) (PRED_5=PRED_21_RTR)(RESID_5=RESID_21_RTR) (PRED_6=PRED_30_RTR)(RESID_6=RESID_30_RTR) (PRED_7=PRED_31_RTR)(RESID_7=RESID_31_RTR). if CAWI=0 PRED_FIN_RTR=PRED_21_RTR. if CAWI=0 RESID_FIN_RTR=RESID_21_RTR. if CAWI=1 PRED_FIN_RTR=PRED_31_RTR. if CAWI=1 RESID_FIN_RTR=RESID_31_RTR. WEIGHT by weight_d. /** Recoding of residuals into discrete intervals. RANK VARIABLES = RESID_21_RTR by trial /NTILES (100) INTO RESID_2_100Tiles. RANK VARIABLES = RESID_21_RTR by trial /NTILES (50) INTO RESID_2_50Tiles. RANK VARIABLES = RESID_21_RTR by trial /NTILES (20) INTO RESID_2_20Tiles. RANK VARIABLES = RESID_31_RTR by trial /NTILES (100) INTO RESID_3_100Tiles. RANK VARIABLES = RESID_31_RTR by trial /NTILES (50) INTO RESID_3_50Tiles. RANK VARIABLES = RESID_31_RTR by trial /NTILES (20) INTO RESID_3_20Tiles. RANK VARIABLES = RESID_FIN_RTR by trial /NTILES (100) INTO RESID_F_100Tiles. RANK VARIABLES = RESID_FIN_RTR by trial /NTILES (50) INTO RESID_F_50Tiles. RANK VARIABLES = RESID_FIN_RTR by trial /NTILES (20) INTO RESID_F_20Tiles. /** Creating separate Ntiles for CATI and CAWI respondents in a single variable. compute trialC=(CAWI*100)+trial. RANK VARIABLES = RESID_FIN_RTR by trialC /NTILES (20) INTO RESID_FS_20Tiles. save outfile="S2015_RT4.sav"/keep=all. /** ============================================================================================. get file="S2015_RT4.sav"/keep=all. SORT CASES BY userid(A). CASESTOVARS /ID=userid /INDEX=trial /FIXED=CAWI internet education lregion age age_sq foreign polint weight_d female LRint dayselec PRexecutive PRsupervisory PRoperative PRselfemployed PRinactive PRmissing relig_prot relig_cath relig_oth relig_none oa1.timR oa2.timR oa3.timR oa4.timR oa5.timR oa6.timR oa7.timR oa8.timR oa9.timR oa10.timR oa11.timR oa12.timR oa13.timR oa14.timR ob1.timR ob2.timR ob3.timR ob4.timR ob5.timR ob6.timR ob7.timR ob8.timR ob9.timR ob10.timR ok1.timR ok2.timR ok3.timR ok4.timR. save outfile="S2015_RT5.sav"/keep=all. /** ============================================================================================. get file="S2015_RT5.sav"/keep=all. MATCH FILES /FILE=* /FILE="726_Selects2015_PES_Data_v1.02.sav" /RENAME (weight_d = d0) /BY userid /DROP= d0. WEIGHT BY weight_d. save outfile="S2015_RT6.sav"/keep=all. /** ============================================================================================. get file="S2015_RT6.sav"/keep=all. /** COMPUTING 10-TILES RESIDUALS DISTRIBUTIONS FOR KNOWLEDGE ITEMS. recode RESID_FS_20Tiles.25 (1,2=1)(3,4=2)(5,6=3)(7,8=4)(9,10=5)(11,12=6)(13,14=7)(15,16=8)(17,18=9)(19,20=10) into RESID_FS_10Tiles.KNOW1. recode RESID_FS_20Tiles.26 (1,2=1)(3,4=2)(5,6=3)(7,8=4)(9,10=5)(11,12=6)(13,14=7)(15,16=8)(17,18=9)(19,20=10) into RESID_FS_10Tiles.KNOW2. recode RESID_FS_20Tiles.27 (1,2=1)(3,4=2)(5,6=3)(7,8=4)(9,10=5)(11,12=6)(13,14=7)(15,16=8)(17,18=9)(19,20=10) into RESID_FS_10Tiles.KNOW3. recode RESID_FS_20Tiles.28 (1,2=1)(3,4=2)(5,6=3)(7,8=4)(9,10=5)(11,12=6)(13,14=7)(15,16=8)(17,18=9)(19,20=10) into RESID_FS_10Tiles.KNOW4. rename variables (RESID_FS_20Tiles.25=RESID_FS_20Tiles.KNOW1) (RESID_FS_20Tiles.26=RESID_FS_20Tiles.KNOW2) (RESID_FS_20Tiles.27=RESID_FS_20Tiles.KNOW3) (RESID_FS_20Tiles.28=RESID_FS_20Tiles.KNOW4). /** ============================================. /** RELATING RT RESIDUALS TO KNOWLEDGE "HIT RATES". /** ============================================. recode f15900r f16000r f16100r f16300r (8,9=0)(else=copy) into know1,know2,know3,know4. value labels know1,know2,know3,know4 0"wrong" 1"correct". /** @@@ FIGURE 3 @@@. /** CATI SAMPLE (10 intervals). temporary. select if CAWI=0. MEANS TABLES=know1 BY RESID_FS_10Tiles.KNOW1 /CELLS=MEAN COUNT STDDEV. temporary. select if CAWI=0. MEANS TABLES=know2 BY RESID_FS_10Tiles.KNOW2 /CELLS=MEAN COUNT STDDEV. temporary. select if CAWI=0. MEANS TABLES=know3 BY RESID_FS_10Tiles.KNOW3 /CELLS=MEAN COUNT STDDEV. temporary. select if CAWI=0. MEANS TABLES=know4 BY RESID_FS_10Tiles.KNOW4 /CELLS=MEAN COUNT STDDEV. /** CAWI SAMPLE (20 intervals). temporary. select if CAWI=1. MEANS TABLES=know1 BY RESID_FS_20Tiles.KNOW1 /CELLS=MEAN COUNT STDDEV. temporary. select if CAWI=1. MEANS TABLES=know2 BY RESID_FS_20Tiles.KNOW2 /CELLS=MEAN COUNT STDDEV. temporary. select if CAWI=1. MEANS TABLES=know3 BY RESID_FS_20Tiles.KNOW3 /CELLS=MEAN COUNT STDDEV. temporary. select if CAWI=1. MEANS TABLES=know4 BY RESID_FS_20Tiles.KNOW4 /CELLS=MEAN COUNT STDDEV. /** SOCIODEMOGRAPHIC PREDICTORS FOR VALIDITY ANALYSIS. /** LINGUISTIC REGION (dummies). recode lregion (1=1)(2,3=0) into region_alem. recode lregion (2=1)(1,3=0) into region_rom. recode lregion (3=1)(1,2=0) into region_ital. /** DEGREE OF URBANIZATION (Eurostat). recode degurba (1=1)(2,3=0) into urba_high. recode degurba (2=1)(1,3=0) into urba_mid. recode degurba (3=1)(1,2=0) into urba_low. /** INCOME GROUPS.. recode f28910 (1,2,3,4,98,99,sysmis=1)(else=0) into income_lowdk. recode f28910 (5,6,7,8,9=1)(else=0) into income_mid. recode f28910 (10,11,12,13,14,15=1)(else=0) into income_high. /** OVERALL EXPOSURE TO INFORMATION. /** NB: Measure based on 3 sub-scales: (1) political discussions; (2) # of information sources; (3) attention to news. /** 1. Attention to news. compute newsattention=mean(f13401,f13402,f13403,f13404,f13405). /** 2. Number of information sources. count infosource=f13661,f13662,f13665,f13666,f13669 (1). /** 3. Political discussions. compute poldisc=99. if f13300=2 poldisc=0. if f13305=3 poldisc=1. if f13305=2 poldisc=2. if f13305=1 poldisc=3. missing values poldisc(99). value labels poldisc 0"never" 1"rarely" 2"sometimes" 3"often". /** Overall exposure to information. recode newsattention (1 thru 1.5=0)(1.51 thru 2=1)(2.01 thru 2.5=2)(2.51 thru 4=3) into info_dim1. recode infosource (0=0)(1=1)(2=2)(3 thru 5=3) into info_dim2. compute info_dim3=poldisc. compute info_exp=mean.2(info_dim1,info_dim2,info_dim3). /** ============================================. /** ANALYSIS OF CONVERGENT AND PREDICTIVE VALIDITY. /** ============================================. /** CONVERGENT VALIDITY: COMPUTING A "CHEATING" VARIABLE. compute cheat_1A=0. compute cheat_1B=0. compute cheat_1C=0. compute cheat_2A=0. compute cheat_2B=0. compute cheat_2C=0. if (RESID_FS_20Tiles.KNOW2 gt 16) OR (RESID_FS_20Tiles.KNOW3 gt 18) cheat_1A=1. if (RESID_FS_20Tiles.KNOW2 gt 16 AND CAWI=1) OR (RESID_FS_20Tiles.KNOW3 gt 18) cheat_1B=1. if (RESID_FS_20Tiles.KNOW2 gt 16 AND CAWI=1) OR (RESID_FS_20Tiles.KNOW3 gt 18 AND CAWI=1) cheat_1C=1. if (RESID_FS_20Tiles.KNOW2 gt 16) AND (RESID_FS_20Tiles.KNOW3 gt 18) cheat_2A=1. if (RESID_FS_20Tiles.KNOW2 gt 16 AND CAWI=1) AND (RESID_FS_20Tiles.KNOW3 gt 18) cheat_2B=1. if (RESID_FS_20Tiles.KNOW2 gt 16 AND CAWI=1) AND (RESID_FS_20Tiles.KNOW3 gt 18 AND CAWI=1) cheat_2C=1. /** @@@ TABLE 2 @@@. /** NB: To enhance readability of the age effect, the coefficient (and S.E.) was multiplied by 10 in Table 2, thus expressing the effect of age in decades. compute cheat=cheat_1A. compute knowscale_r0=mean.4(know1,know2,know3,know4). compute cheat_polint=cheat*polint. compute cheat_education=cheat*education. compute cheat_income_mid=cheat*income_mid. compute cheat_income_high=cheat*income_high. compute cheat_rom=cheat*region_rom. compute cheat_ital=cheat*region_ital. compute cheat_female=cheat*female. compute cheat_age=cheat*age. compute cheat_urba_mid=cheat*urba_mid. compute cheat_urba_high=cheat*urba_high. REGRESSION /MISSING LISTWISE /STATISTICS COEFF OUTS R ANOVA CHANGE SELECTION /CRITERIA=PIN(.001) POUT(.01) /NOORIGIN /DEPENDENT knowscale_r0 /METHOD=ENTER age female education income_mid income_high region_rom region_ital urba_high urba_mid polint cheat cheat_polint cheat_education cheat_income_mid cheat_income_high cheat_rom cheat_ital cheat_female cheat_age cheat_urba_mid cheat_urba_high. /** @@@ TABLE A.4 @@@. /** Choosing which of the six cheating measures shall be used for tests. /** One example: compute cheat=cheat_2A. compute cheat_polint=cheat*polint. compute cheat_education=cheat*education. compute cheat_income_mid=cheat*income_mid. compute cheat_income_high=cheat*income_high. compute cheat_rom=cheat*region_rom. compute cheat_ital=cheat*region_ital. compute cheat_female=cheat*female. compute cheat_age=cheat*age. compute cheat_urba_mid=cheat*urba_mid. compute cheat_urba_high=cheat*urba_high. REGRESSION /MISSING LISTWISE /STATISTICS COEFF OUTS R ANOVA CHANGE SELECTION /CRITERIA=PIN(.001) POUT(.01) /NOORIGIN /DEPENDENT knowscale_r0 /METHOD=ENTER age female education income_mid income_high region_rom region_ital urba_high urba_mid polint cheat cheat_polint cheat_education cheat_income_mid cheat_income_high cheat_rom cheat_ital cheat_female cheat_age cheat_urba_mid cheat_urba_high. /** PREDICTIVE VALIDITY: COMPUTING KNOWLEDGE SCALES. /** A stands for the most inclusive treatment; C stands for the least inclusive treatment. compute know2_r1a=know2. compute know3_r1a=know3. if RESID_FS_20Tiles.KNOW2 gt 16 AND know2=1 know2_r1a=0.5. if RESID_FS_20Tiles.KNOW3 gt 18 AND know3=1 know3_r1a=0.5. compute know2_r1b=know2. compute know3_r1b=know3. if RESID_FS_20Tiles.KNOW2 gt 16 AND CAWI=1 AND know2=1 know2_r1b=0.5. if RESID_FS_20Tiles.KNOW3 gt 18 AND know3=1 know3_r1b=0.5. compute know2_r1c=know2. compute know3_r1c=know3. if RESID_FS_20Tiles.KNOW2 gt 16 AND CAWI=1 AND know2=1 know2_r1c=0.5. if RESID_FS_20Tiles.KNOW3 gt 18 AND CAWI=1 AND know3=1 know3_r1c=0.5. compute know2_r2a=know2. compute know3_r2a=know3. if RESID_FS_20Tiles.KNOW2 gt 16 AND know2=1 know2_r2a=0. if RESID_FS_20Tiles.KNOW3 gt 18 AND know3=1 know3_r2a=0. compute know2_r2b=know2. compute know3_r2b=know3. if RESID_FS_20Tiles.KNOW2 gt 16 AND CAWI=1 AND know2=1 know2_r2b=0. if RESID_FS_20Tiles.KNOW3 gt 18 AND know3=1 know3_r2b=0. compute know2_r2c=know2. compute know3_r2c=know3. if RESID_FS_20Tiles.KNOW2 gt 16 AND CAWI=1 AND know2=1 know2_r2c=0. if RESID_FS_20Tiles.KNOW3 gt 18 AND CAWI=1 AND know3=1 know3_r2c=0. compute know1_r3a=((21-RESID_FS_20Tiles.KNOW1)/20)*know1. if missing(RESID_FS_20Tiles.KNOW1) know1_r3a=know1. compute know2_r3a=((21-RESID_FS_20Tiles.KNOW2)/20)*know2. if missing(RESID_FS_20Tiles.KNOW2) know2_r3a=know2. compute know3_r3a=((21-RESID_FS_20Tiles.KNOW3)/20)*know3. if missing(RESID_FS_20Tiles.KNOW3) know3_r3a=know3. compute know4_r3a=((21-RESID_FS_20Tiles.KNOW4)/20)*know4. if missing(RESID_FS_20Tiles.KNOW4) know4_r3a=know4. compute know1_r3b=((21-RESID_FS_20Tiles.KNOW1)/20)*know1. if missing(RESID_FS_20Tiles.KNOW1) OR CAWI=0 know1_r3b=know1. compute know2_r3b=((21-RESID_FS_20Tiles.KNOW2)/20)*know2. if missing(RESID_FS_20Tiles.KNOW2) OR CAWI=0 know2_r3b=know2. compute know3_r3b=((21-RESID_FS_20Tiles.KNOW3)/20)*know3. if missing(RESID_FS_20Tiles.KNOW3) OR CAWI=0 know3_r3b=know3. compute know4_r3b=((21-RESID_FS_20Tiles.KNOW4)/20)*know4. if missing(RESID_FS_20Tiles.KNOW4) OR CAWI=0 know4_r3b=know4. compute know1_r4a=(ln(21-RESID_FS_20Tiles.KNOW1)/ln(20))*know1. if missing(RESID_FS_20Tiles.KNOW1) know1_r4a=know1. compute know2_r4a=(ln(21-RESID_FS_20Tiles.KNOW2)/ln(20))*know2. if missing(RESID_FS_20Tiles.KNOW2) know2_r4a=know2. compute know3_r4a=(ln(21-RESID_FS_20Tiles.KNOW3)/ln(20))*know3. if missing(RESID_FS_20Tiles.KNOW3) know3_r4a=know3. compute know4_r4a=(ln(21-RESID_FS_20Tiles.KNOW4)/ln(20))*know4. if missing(RESID_FS_20Tiles.KNOW4) know4_r4a=know4. compute know1_r4b=(ln(21-RESID_FS_20Tiles.KNOW1)/ln(20))*know1. if missing(RESID_FS_20Tiles.KNOW1) OR CAWI=0 know1_r4b=know1. compute know2_r4b=(ln(21-RESID_FS_20Tiles.KNOW2)/ln(20))*know2. if missing(RESID_FS_20Tiles.KNOW2) OR CAWI=0 know2_r4b=know2. compute know3_r4b=(ln(21-RESID_FS_20Tiles.KNOW3)/ln(20))*know3. if missing(RESID_FS_20Tiles.KNOW3) OR CAWI=0 know3_r4b=know3. compute know4_r4b=(ln(21-RESID_FS_20Tiles.KNOW4)/ln(20))*know4. if missing(RESID_FS_20Tiles.KNOW4) OR CAWI=0 know4_r4b=know4. /** Knowledge scales #5 and #6 are built to test the alternative hypothesis that RTs in themselves, and not cheating, allow to improve the validity of knowledge scales. /** Here I reverse the logic and build scales which penalize slow respondents on K1 and K4, NOT on K2 and K3. compute know1_r5=know1. if RESID_FS_20Tiles.KNOW1 gt 18 AND know1=1 know1_r5=0.5. compute know2_r5=know2. compute know3_r5=know3. compute know4_r5=know4. if RESID_FS_20Tiles.KNOW4 gt 18 AND know4=1 know4_r5=0.5. compute know1_r6=know1. if RESID_FS_20Tiles.KNOW1 gt 18 AND know1=1 know1_r6=0. compute know2_r6=know2. compute know3_r6=know3. compute know4_r6=know4. if RESID_FS_20Tiles.KNOW4 gt 18 AND know4=1 know4_r6=0. /** COMPUTING KNOWLEDGE SCALES. compute knowscale_r0=mean.4(know1,know2,know3,know4). compute knowscale_r1a=mean.4(know1,know4,know2_r1a,know3_r1a). compute knowscale_r1b=mean.4(know1,know4,know2_r1b,know3_r1b). compute knowscale_r1c=mean.4(know1,know4,know2_r1c,know3_r1c). compute knowscale_r2a=mean.4(know1,know4,know2_r2a,know3_r2a). compute knowscale_r2b=mean.4(know1,know4,know2_r2b,know3_r2b). compute knowscale_r2c=mean.4(know1,know4,know2_r2c,know3_r2c). compute knowscale_r3a=mean.4(know1_r3a,know2_r3a,know3_r3a,know4_r3a). compute knowscale_r3b=mean.4(know1_r3b,know2_r3b,know3_r3b,know4_r3b). compute knowscale_r3c=mean.4(know1,know2_r3b,know3_r3b,know4). compute knowscale_r4a=mean.4(know1_r4a,know2_r4a,know3_r4a,know4_r4a). compute knowscale_r4b=mean.4(know1_r4b,know2_r4b,know3_r4b,know4_r4b). compute knowscale_r4c=mean.4(know1,know2_r4b,know3_r4b,know4). compute knowscale_r5=mean.4(know1_r5,know2_r5,know3_r5,know4_r5). compute knowscale_r6=mean.4(know1_r6,know2_r6,know3_r6,know4_r6). /** @@@ TABLE 3, LEFT PANEL @@@. /** F12500 (direct democratic participation, 0-10) as dependent. /** Model 1. REGRESSION /MISSING LISTWISE /STATISTICS COEFF OUTS R ANOVA CHANGE SELECTION /CRITERIA=PIN(.001) POUT(.01) /NOORIGIN /DEPENDENT f12500 /METHOD=ENTER age female education income_mid income_high region_rom region_ital urba_mid urba_high polint info_exp knowscale_r0. /** Model 2. REGRESSION /MISSING LISTWISE /STATISTICS COEFF OUTS R ANOVA CHANGE SELECTION /CRITERIA=PIN(.001) POUT(.01) /NOORIGIN /DEPENDENT f12500 /METHOD=ENTER age female education income_mid income_high region_rom region_ital urba_mid urba_high polint info_exp knowscale_r1a. /** Model 3. REGRESSION /MISSING LISTWISE /STATISTICS COEFF OUTS R ANOVA CHANGE SELECTION /CRITERIA=PIN(.001) POUT(.01) /NOORIGIN /DEPENDENT f12500 /METHOD=ENTER age female education income_mid income_high region_rom region_ital urba_mid urba_high polint info_exp knowscale_r2a. /** @@@ TABLE 3, RIGHT PANEL @@@. /** F11100R (electoral participation, 0-1) as dependent. /** NB: I use multinomial regression (instead of logistic regression) to get BIC values. /** Model 4. NOMREG f11100r (BASE=FIRST ORDER=ASCENDING) WITH age female education income_mid income_high region_rom region_ital urba_mid urba_high polint info_exp knowscale_r0 /CRITERIA CIN(95) DELTA(0) MXITER(100) MXSTEP(5) CHKSEP(20) LCONVERGE(0) PCONVERGE(0.000001) SINGULAR(0.00000001) /MODEL /STEPWISE=PIN(.05) POUT(0.1) MINEFFECT(0) RULE(SINGLE) ENTRYMETHOD(LR) REMOVALMETHOD(LR) /INTERCEPT=INCLUDE /PRINT=PARAMETER SUMMARY LRT CPS STEP MFI IC. /** Model 5. NOMREG f11100r (BASE=FIRST ORDER=ASCENDING) WITH age female education income_mid income_high region_rom region_ital urba_mid urba_high polint info_exp knowscale_r1a /CRITERIA CIN(95) DELTA(0) MXITER(100) MXSTEP(5) CHKSEP(20) LCONVERGE(0) PCONVERGE(0.000001) SINGULAR(0.00000001) /MODEL /STEPWISE=PIN(.05) POUT(0.1) MINEFFECT(0) RULE(SINGLE) ENTRYMETHOD(LR) REMOVALMETHOD(LR) /INTERCEPT=INCLUDE /PRINT=PARAMETER SUMMARY LRT CPS STEP MFI IC. /** Model 6. NOMREG f11100r (BASE=FIRST ORDER=ASCENDING) WITH age female education income_mid income_high region_rom region_ital urba_mid urba_high polint info_exp knowscale_r2a /CRITERIA CIN(95) DELTA(0) MXITER(100) MXSTEP(5) CHKSEP(20) LCONVERGE(0) PCONVERGE(0.000001) SINGULAR(0.00000001) /MODEL /STEPWISE=PIN(.05) POUT(0.1) MINEFFECT(0) RULE(SINGLE) ENTRYMETHOD(LR) REMOVALMETHOD(LR) /INTERCEPT=INCLUDE /PRINT=PARAMETER SUMMARY LRT CPS STEP MFI IC. /** @@@ TABLES A.6 AND A.7 @@@. /** Choosing which knowledge scale shall be used for tests. /** One example: compute knowscale_test=knowscale_r3a. REGRESSION /MISSING LISTWISE /STATISTICS COEFF OUTS R ANOVA CHANGE SELECTION /CRITERIA=PIN(.001) POUT(.01) /NOORIGIN /DEPENDENT f12500 /METHOD=ENTER age female education income_mid income_high region_rom region_ital urba_mid urba_high polint info_exp knowscale_test. NOMREG f11100r (BASE=FIRST ORDER=ASCENDING) WITH age female education income_mid income_high region_rom region_ital urba_mid urba_high polint info_exp knowscale_test /CRITERIA CIN(95) DELTA(0) MXITER(100) MXSTEP(5) CHKSEP(20) LCONVERGE(0) PCONVERGE(0.000001) SINGULAR(0.00000001) /MODEL /STEPWISE=PIN(.05) POUT(0.1) MINEFFECT(0) RULE(SINGLE) ENTRYMETHOD(LR) REMOVALMETHOD(LR) /INTERCEPT=INCLUDE /PRINT=PARAMETER SUMMARY LRT CPS STEP MFI IC. /** ===================. /** ADDITIONAL ANALYSES. /** ===================. /** First-hand analysis of cheating behavior. /** @@@ FIGURE A.1 @@@. RANK VARIABLES = ok1.timR by CAWI /NTILES (10) INTO OK1_10Tiles. RANK VARIABLES = ok2.timR by CAWI /NTILES (10) INTO OK2_10Tiles. RANK VARIABLES = ok3.timR by CAWI /NTILES (10) INTO OK3_10Tiles. RANK VARIABLES = ok4.timR by CAWI /NTILES (10) INTO OK4_10Tiles. MEANS TABLES=know1 BY OK1_10Tiles /CELLS=MEAN COUNT STDDEV. MEANS TABLES=know2 BY OK2_10Tiles /CELLS=MEAN COUNT STDDEV. MEANS TABLES=know3 BY OK3_10Tiles /CELLS=MEAN COUNT STDDEV. MEANS TABLES=know4 BY OK4_10Tiles /CELLS=MEAN COUNT STDDEV. /** Estimating reclassification of individuals by the residuals analysis (in terms of percentile positions). /** @@@ FOOTNOTE #18 @@@. RANK VARIABLES = RTR.25 /NTILES (100) INTO RTR.25_100Tiles. RANK VARIABLES = RTR.26 /NTILES (100) INTO RTR.26_100Tiles. RANK VARIABLES = RTR.27 /NTILES (100) INTO RTR.27_100Tiles. RANK VARIABLES = RTR.28 /NTILES (100) INTO RTR.28_100Tiles. compute DIFF.25_tiles=abs(RTR.25_100Tiles-RESID_F_100Tiles.25). compute DIFF.26_tiles=abs(RTR.26_100Tiles-RESID_F_100Tiles.26). compute DIFF.27_tiles=abs(RTR.27_100Tiles-RESID_F_100Tiles.27). compute DIFF.28_tiles=abs(RTR.28_100Tiles-RESID_F_100Tiles.28). compute DIFF.25A_tiles=(RTR.25_100Tiles-RESID_F_100Tiles.25). compute DIFF.26A_tiles=(RTR.26_100Tiles-RESID_F_100Tiles.26). compute DIFF.27A_tiles=(RTR.27_100Tiles-RESID_F_100Tiles.27). compute DIFF.28A_tiles=(RTR.28_100Tiles-RESID_F_100Tiles.28). FREQUENCIES VARIABLES=DIFF.25_tiles DIFF.26_tiles DIFF.27_tiles DIFF.28_tiles /FORMAT=NOTABLE /STATISTICS=MEAN /ORDER=ANALYSIS. /** RENAMING RTs. rename variables (oa1.timR=A1)(oa2.timR=A2)(oa3.timR=A3)(oa4.timR=A4)(oa5.timR=A5)(oa6.timR=A6)(oa7.timR=A7)(oa12.timR=A8)(oa13.timR=A9)(oa14.timR=A10) (ob1.timR=B1)(ob2.timR=B2)(ob3.timR=B3)(ob4.timR=B4)(ob5.timR=B5)(ob6.timR=B6)(ob7.timR=B7)(ob8.timR=B8)(ob9.timR=B9)(ob10.timR=B10) (ok1.timR=K1)(ok2.timR=K2)(ok3.timR=K3)(ok4.timR=K4). rename variables (RTR.1=A1R)(RTR.2=A2R)(RTR.3=A3R)(RTR.4=A4R)(RTR.5=A5R)(RTR.6=A6R)(RTR.7=A7R)(RTR.12=A8R)(RTR.13=A9R)(RTR.14=A10R) (RTR.15=B1R)(RTR.16=B2R)(RTR.17=B3R)(RTR.18=B4R)(RTR.19=B5R)(RTR.20=B6R)(RTR.21=B7R)(RTR.22=B8R)(RTR.23=B9R)(RTR.24=B10R) (RTR.25=K1R)(RTR.26=K2R)(RTR.27=K3R)(RTR.28=K4R). /** PRINCIPAL COMPONENT ANALYSIS OF RECODED ITEMS. /** @@@ FIGURE A.2 @@@. FACTOR /VARIABLES A1 A2 A3 A4 A5 A6 A7 A8 A9 A10 B1 B2 B3 B4 B5 B6 B7 B8 B9 B10 K1 K2 K3 K4 /MISSING LISTWISE /ANALYSIS A1 A2 A3 A4 A5 A6 A7 A8 A9 A10 B1 B2 B3 B4 B5 B6 B7 B8 B9 B10 K1 K2 K3 K4 /PRINT INITIAL KMO EXTRACTION ROTATION /PLOT ROTATION /CRITERIA MINEIGEN(1) ITERATE(100) /EXTRACTION PC /CRITERIA ITERATE(100) DELTA(0) /ROTATION OBLIMIN /METHOD=CORRELATION. temporary. select if CAWI=0. FACTOR /VARIABLES A1 A2 A3 A4 A5 A6 A7 A8 A9 A10 B1 B2 B3 B4 B5 B6 B7 B8 B9 B10 K1 K2 K3 K4 /MISSING LISTWISE /ANALYSIS A1 A2 A3 A4 A5 A6 A7 A8 A9 A10 B1 B2 B3 B4 B5 B6 B7 B8 B9 B10 K1 K2 K3 K4 /PRINT INITIAL KMO EXTRACTION ROTATION /PLOT ROTATION /CRITERIA MINEIGEN(1) ITERATE(100) /EXTRACTION PC /CRITERIA ITERATE(100) DELTA(0) /ROTATION OBLIMIN /METHOD=CORRELATION. temporary. select if CAWI=1. FACTOR /VARIABLES A1 A2 A3 A4 A5 A6 A7 A8 A9 A10 B1 B2 B3 B4 B5 B6 B7 B8 B9 B10 K1 K2 K3 K4 /MISSING LISTWISE /ANALYSIS A1 A2 A3 A4 A5 A6 A7 A8 A9 A10 B1 B2 B3 B4 B5 B6 B7 B8 B9 B10 K1 K2 K3 K4 /PRINT INITIAL KMO EXTRACTION ROTATION /PLOT ROTATION /CRITERIA MINEIGEN(1) ITERATE(100) /EXTRACTION PC /CRITERIA ITERATE(100) DELTA(0) /ROTATION OBLIMIN /METHOD=CORRELATION. /** Documenting the variables in the final dataset (see below). variable label A1 "RT for Item A1 (trimmed)". variable label A2 "RT for Item A2 (trimmed)". variable label A3 "RT for Item A3 (trimmed)". variable label A4 "RT for Item A4 (trimmed)". variable label A5 "RT for Item A5 (trimmed)". variable label A6 "RT for Item A6 (trimmed)". variable label A7 "RT for Item A7 (trimmed)". variable label A8 "RT for Item A8 (trimmed)". variable label A9 "RT for Item A9 (trimmed)". variable label A10 "RT for Item A10 (trimmed)". variable label B1 "RT for Item B1 (trimmed)". variable label B2 "RT for Item B2 (trimmed)". variable label B3 "RT for Item B3 (trimmed)". variable label B4 "RT for Item B4 (trimmed)". variable label B5 "RT for Item B5 (trimmed)". variable label B6 "RT for Item B6 (trimmed)". variable label B7 "RT for Item B7 (trimmed)". variable label B8 "RT for Item B8 (trimmed)". variable label B9 "RT for Item B9 (trimmed)". variable label B10 "RT for Item B10 (trimmed)". variable label K1 "RT for Item K1 (trimmed)". variable label K2 "RT for Item K2 (trimmed)". variable label K3 "RT for Item K3 (trimmed)". variable label K4 "RT for Item K4 (trimmed)". variable label A1R "RT for Item A1 (residual from CCM)". variable label A2R "RT for Item A2 (residual from CCM)". variable label A3R "RT for Item A3 (residual from CCM)". variable label A4R "RT for Item A4 (residual from CCM)". variable label A5R "RT for Item A5 (residual from CCM)". variable label A6R "RT for Item A6 (residual from CCM)". variable label A7R "RT for Item A7 (residual from CCM)". variable label A8R "RT for Item A8 (residual from CCM)". variable label A9R "RT for Item A9 (residual from CCM)". variable label A10R "RT for Item A10 (residual from CCM)". variable label B1R "RT for Item B1 (residual from CCM)". variable label B2R "RT for Item B2 (residual from CCM)". variable label B3R "RT for Item B3 (residual from CCM)". variable label B4R "RT for Item B4 (residual from CCM)". variable label B5R "RT for Item B5 (residual from CCM)". variable label B6R "RT for Item B6 (residual from CCM)". variable label B7R "RT for Item B7 (residual from CCM)". variable label B8R "RT for Item B8 (residual from CCM)". variable label B9R "RT for Item B9 (residual from CCM)". variable label B10R "RT for Item B10 (residual from CCM)". variable label K1R "RT for Item K1 (residual from CCM)". variable label K2R "RT for Item K2 (residual from CCM)". variable label K3R "RT for Item K3 (residual from CCM)". variable label K4R "RT for Item K4 (residual from CCM)". variable label RESID_FS_20Tiles.KNOW1 "RT for Item K1 (20 intervals)". variable label RESID_FS_20Tiles.KNOW2 "RT for Item K2 (20 intervals)". variable label RESID_FS_20Tiles.KNOW3 "RT for Item K3 (20 intervals)". variable label RESID_FS_20Tiles.KNOW4 "RT for Item K4 (20 intervals)". variable label RESID_FS_10Tiles.KNOW1 "RT for Item K1 (10 intervals)". variable label RESID_FS_10Tiles.KNOW2 "RT for Item K2 (10 intervals)". variable label RESID_FS_10Tiles.KNOW3 "RT for Item K3 (10 intervals)". variable label RESID_FS_10Tiles.KNOW4 "RT for Item K4 (10 intervals)". variable label cheat_1A "Cheater group, definition 1A; see Table A.3". variable label cheat_1B "Cheater group, definition 1B; see Table A.3". variable label cheat_1C "Cheater group, definition 1C; see Table A.3". variable label cheat_2A "Cheater group, definition 2A; see Table A.3". variable label cheat_2B "Cheater group, definition 2B; see Table A.3". variable label cheat_2C "Cheater group, definition 2C; see Table A.3". variable label know1 "Knowledge Q1: Number of parties in the Federal Council". variable label know2 "Knowledge Q2: Name of the Swiss president in 2015". variable label know3 "Knowledge Q3: Required number of signatures for a federal initiative". variable label know4 "Knowledge Q4: Party with most seats in the National Council". variable label knowscale_r0 "Knowledge scale (0-1): Original scale". variable label knowscale_r1a "Knowledge scale (0-1): Revised scale 1A; see Appendix A.5". variable label knowscale_r1b "Knowledge scale (0-1): Revised scale 1B; see Appendix A.5". variable label knowscale_r1c "Knowledge scale (0-1): Revised scale 1C; see Appendix A.5". variable label knowscale_r2a "Knowledge scale (0-1): Revised scale 2A; see Appendix A.5". variable label knowscale_r2b "Knowledge scale (0-1): Revised scale 2B; see Appendix A.5". variable label knowscale_r2c "Knowledge scale (0-1): Revised scale 2C; see Appendix A.5". variable label knowscale_r3a "Knowledge scale (0-1): Revised scale 3A; see Appendix A.5". variable label knowscale_r3b "Knowledge scale (0-1): Revised scale 3B; see Appendix A.5". variable label knowscale_r3c "Knowledge scale (0-1): Revised scale 3C; see Appendix A.5". variable label knowscale_r4a "Knowledge scale (0-1): Revised scale 4A; see Appendix A.5". variable label knowscale_r4b "Knowledge scale (0-1): Revised scale 4B; see Appendix A.5". variable label knowscale_r4c "Knowledge scale (0-1): Revised scale 4C; see Appendix A.5". variable label knowscale_r5 "Knowledge scale (0-1): Revised scale 5; see Appendix A.5". variable label knowscale_r6 "Knowledge scale (0-1): Revised scale 6; see Appendix A.5". variable label CAWI "Interviewing mode (CATI vs. CAWI)". variable label internet "Frequency of internet use (0-2)". variable label education "Highest educational level attained (1-8)". variable label lregion "Linguistic region (3 main areas)". variable label age "Respondent's age". variable label female "Gender: Female=1 (vs. male=0) dummy". variable label polint "Political interest (0-3)". variable label LRint "Intensity of left-right self-placement (0-5)". variable label foreign "Foreigner at birth, dummy (0-1)". variable label PRexecutive "Professional role: employee with executive function, dummy (0-1)". variable label PRsupervisory "Professional role: employee with supervisory/training function, dummy (0-1)". variable label PRoperative "Professional role: employee with operative function, dummy (0-1)". variable label PRselfemployed "Professional role: self-employed, dummy (0-1)". variable label PRinactive "Professional role: inactive (in training, retired), dummy (0-1)". variable label PRmissing "Professional role: other or missing, dummy (0-1)". variable label relig_prot "Religion: protestant, dummy (0-1)". variable label relig_cath "Religion: catholic, dummy (0-1)". variable label relig_oth "Religion: other, dummy (0-1)". variable label relig_none "Religion: none, dummy (0-1)". /** NB: The final content of the datafile is to be determined by each user. Here we propose to keep a subset of relevant variables. /** NB: In case you want to keep all variables, specify "keep=all". save outfile="S2015_RT7.sav" /keep=userid weight_d A1 A2 A3 A4 A5 A6 A7 A8 A9 A10 B1 B2 B3 B4 B5 B6 B7 B8 B9 B10 K1 K2 K3 K4 A1R A2R A3R A4R A5R A6R A7R A8R A9R A10R B1R B2R B3R B4R B5R B6R B7R B8R B9R B10R K1R K2R K3R K4R RESID_FS_20Tiles.KNOW1 RESID_FS_20Tiles.KNOW2 RESID_FS_20Tiles.KNOW3 RESID_FS_20Tiles.KNOW4 RESID_FS_10Tiles.KNOW1 RESID_FS_10Tiles.KNOW2 RESID_FS_10Tiles.KNOW3 RESID_FS_10Tiles.KNOW4 cheat_1A cheat_1B cheat_1C cheat_2A cheat_2B cheat_2C know1 know2 know3 know4 knowscale_r0 knowscale_r1a to knowscale_r6 CAWI internet education lregion age female polint LRint foreign PRexecutive PRsupervisory PRoperative PRselfemployed PRinactive PRmissing relig_prot relig_cath relig_oth relig_none. /** ==================================================================================.