*******************************************************************
*Millennials participation in surveys (CRONOS)
*******************************************************************

*sort each data and save with a different name

use "G:\CRONOS\CRONOS_INTEGRATED 0 6 ESS8_e01.stata\CRONOS_ESS8_e01.dta", clear
sort idno
save "G:\CRONOS\integratedw06.dta", replace

use "G:\CRONOS\CRONOS_Paradata_e01.stata\CRONOS_Paradata_Basic_e01.dta", clear
sort idno
save "G:\CRONOS\paradata.dta", replace

*Merge 

use "G:\CRONOS\integratedw06.dta", clear
merge 1:1 idno using "G:\CRONOS\paradata.dta"
drop if _merge==2
*drop 2 cases that only have missings in the integrated file
save "G:\merged.dta", replace


*Create age cohorts

gen silent=.
replace silent=0 if 1943<=yrbrn & yrbrn<2003
replace silent=1 if yrbrn<1943
tab silent

gen boomers=.
replace boomers=0 if (1960<yrbrn & yrbrn<2003) | yrbrn<1943
replace boomers=1 if 1943<=yrbrn & yrbrn<=1960
tab boomers

gen genX=.
replace genX=0 if (1981<yrbrn & yrbrn<2003) | yrbrn<1961
replace genX=1 if 1961<=yrbrn & yrbrn<=1981
tab genX

gen millennials=.
replace millennials=0 if yrbrn<1982 & yrbrn>=1922
replace millennials=1 if 1982<=yrbrn & yrbrn<=2003
tab millennials

gen cohort="."
replace cohort="millennials" if millennials==1
replace cohort="genX" if genX==1
replace cohort="boomers" if boomers==1
replace cohort="silent" if silent==1
tab cohort

sort cntry
by cntry: tab cohort

*Drop 5 people without information on age 
drop if cohort=="."



***************************
*Participation in surveys
***************************

*we consider they were invited if the status is "invited", "inactive", "breakoff", "completed without break" "completed after break"
*we consider they were not invited if the status is "not invited" "email could not be delivered", "cannot be reached"
gen noinvited=0
replace noinvited=noinvited+1 if w0disp_status==12 | w0disp_status==13 | w0disp_status==22 | w0disp_status==31 | w0disp_status==32 
replace noinvited=noinvited+1 if w1disp_status==12 | w1disp_status==13 | w1disp_status==22 | w1disp_status==31 | w1disp_status==32  
replace noinvited=noinvited+1 if w2disp_status==12 | w2disp_status==13 | w2disp_status==22 | w2disp_status==31 | w2disp_status==32  
replace noinvited=noinvited+1 if w3disp_status==12 | w3disp_status==13 | w3disp_status==22 | w3disp_status==31 | w3disp_status==32  
replace noinvited=noinvited+1 if w4disp_status==12 | w4disp_status==13 | w4disp_status==22 | w4disp_status==31 | w4disp_status==32  
replace noinvited=noinvited+1 if w5disp_status==12 | w5disp_status==13 | w5disp_status==22 | w5disp_status==31 | w5disp_status==32  
replace noinvited=noinvited+1 if w6disp_status==12 | w6disp_status==13 | w6disp_status==22 | w6disp_status==31 | w6disp_status==32   
tab noinvited 

*we consider that they started if status is "breakoff" "completed without break" or "completed after break"
gen nostarted=0
replace nostarted=nostarted+1 if w0disp_status==22 | w0disp_status==31 | w0disp_status==32 
replace nostarted=nostarted+1 if w1disp_status==22 | w1disp_status==31 | w1disp_status==32
replace nostarted=nostarted+1 if w2disp_status==22 | w2disp_status==31 | w2disp_status==32
replace nostarted=nostarted+1 if w3disp_status==22 | w3disp_status==31 | w3disp_status==32
replace nostarted=nostarted+1 if w4disp_status==22 | w4disp_status==31 | w4disp_status==32
replace nostarted=nostarted+1 if w5disp_status==22 | w5disp_status==31 | w5disp_status==32
replace nostarted=nostarted+1 if w6disp_status==22 | w6disp_status==31 | w6disp_status==32
tab nostarted

*we consider they breakoff if the status is "breakoff"

gen nobreakoff=0
replace nobreakoff=nobreakoff+1 if w0disp_status==22
replace nobreakoff=nobreakoff+1 if w1disp_status==22
replace nobreakoff=nobreakoff+1 if w2disp_status==22
replace nobreakoff=nobreakoff+1 if w3disp_status==22
replace nobreakoff=nobreakoff+1 if w4disp_status==22
replace nobreakoff=nobreakoff+1 if w5disp_status==22
replace nobreakoff=nobreakoff+1 if w6disp_status==22
tab nobreakoff


******Participation rate

*number of surveys started (waves 0 to 6) divided by the number of surveys to which he/she was invited (same waves)
*we report the average of these individual participation rates, for each cohort and country

gen PR=100*nostarted/noinvited
sum PR

by cntry: sum PR
by cntry: sum PR if silent==1
by cntry: sum PR if boomers==1
by cntry: sum PR if genX==1
by cntry: sum PR if millennials==1

*tests significance differences between millennials and other cohorts 

*EE
prtesti 260 .8160 310 .8790 
prtesti 260 .8160 198 .9106
prtesti 260 .8160 38 .9831

*GB
prtesti 213 .6100 357 .7181
prtesti 213 .6100 290 .8250
prtesti 213 .6100 61 .8366

*SI
prtesti 223 .7604 287 .8625
prtesti 223 .7604 165 .8920
prtesti 223 .7604 30 .8786


******Breakoff rate

*number of times he/she abandoned a survey divided by the number of times he/she started a survey (waves 0 to 6). 
*we report the average of these individual break-off rates, for each cohort and country

gen BR=400*nobreakoff/nostarted
sum BR

by cntry: sum BR
by cntry: sum BR if silent==1
by cntry: sum BR if boomers==1
by cntry: sum BR if genX==1
by cntry: sum BR if millennials==1

*EE
prtesti 260 .1284 310 .0653 
prtesti 260 .1284 198 .0497
prtesti 260 .1284 38 .0000

*GB
prtesti 213 .1854 357 .1051
prtesti 213 .1854 290 .0430
prtesti 213 .1854 61 .0901

*SI
prtesti 223 .0711 287 .0418
prtesti 223 .0711 165 .0302
prtesti 223 .0711 30 .0824



******Rate of surveys answered with smartphones

*number of surveys he/she participated using a smartphone (device type used for most items) divided by the number of times he/she started a survey (waves 0 to 6)
*we report the average of these individual rates of surveys completed through smartphones for each cohort and country

gen nosurveyssmartph=0
replace nosurveyssmartph=nosurveyssmartph+1 if w2main_device==2
replace nosurveyssmartph=nosurveyssmartph+1 if w3main_device==2
replace nosurveyssmartph=nosurveyssmartph+1 if w4main_device==2
replace nosurveyssmartph=nosurveyssmartph+1 if w5main_device==2
replace nosurveyssmartph=nosurveyssmartph+1 if w6main_device==2
tab nosurveyssmartph

gen SR=100*nosurveyssmartph/nostarted
sum SR

by cntry: sum SR
by cntry: sum SR if silent==1
by cntry: sum SR if boomers==1
by cntry: sum SR if genX==1
by cntry: sum SR if millennials==1

*EE
prtesti 260 .2328 310 .1296 
prtesti 260 .2328 198 .0961
prtesti 260 .2328 38 .2055

*GB
prtesti 213 .4471 357 .3709
prtesti 213 .4471 290 .1724
prtesti 213 .4471 61 .1633

*SI
prtesti 223 .3750 287 .1912
prtesti 223 .3750 165 .1765
prtesti 223 .3750 30 .3740


*****************************
*Evaluation of surveys
*****************************

******Difficult surveys

replace w3q75=. if w3q75==-99 | w3q75==-33
replace w6q77=. if w6q77==-99 | w6q77==-33 | w6q77==-88

*create a variable counting how many times they said 3, 4 or 5
gen nodifficult=0
replace nodifficult=nodifficult+1 if w0q44==3 | w0q44==4 | w0q44==5
replace nodifficult=nodifficult+1 if w1q93==3 | w1q93==4 | w1q93==5
replace nodifficult=nodifficult+1 if w2q94==3 | w2q94==4 | w2q94==5
replace nodifficult=nodifficult+1 if w3q75==3 | w3q75==4 | w3q75==5
replace nodifficult=nodifficult+1 if w4q60==3 | w4q60==4 | w4q60==5
replace nodifficult=nodifficult+1 if w6q77==3 | w6q77==4 | w6q77==5 
tab nodifficult

gen noansweredDQ=0
replace noansweredDQ=noansweredDQ+1 if w0q44!=. 
replace noansweredDQ=noansweredDQ+1 if w1q93!=.
replace noansweredDQ=noansweredDQ+1 if w2q94!=.
replace noansweredDQ=noansweredDQ+1 if w3q75!=.
replace noansweredDQ=noansweredDQ+1 if w4q60!=.
replace noansweredDQ=noansweredDQ+1 if w6q77!=.
tab noansweredDQ

gen DR=100*nodifficult/(noansweredDQ)

by cntry: sum DR
by cntry: sum DR if silent==1
by cntry: sum DR if boomers==1
by cntry: sum DR if genX==1
by cntry: sum DR if millennials==1

*EE
prtesti 243 .15 194 .166 
prtesti 243 .15 301 .122
prtesti 243 .15 38 .251

*GB
prtesti 189 .147 285 .138
prtesti 189 .147 333 .115
prtesti 189 .147 58 .207

*SI
prtesti 205 .210 161 .209
prtesti 205 .210 279 .178
prtesti 205 .210 28 .374


******Liked/enjoyed surveys

replace w3q77=. if w3q77==-99 | w3q77==-77 | w3q77==-33
replace w6q80=. if w6q80==-99 | w6q80==-77 | w6q80==-88 | w6q80==-33

*create a variable counting how many times they said 3, 4 or 5
gen noliked=0
replace noliked=noliked+1 if w0q46==3 | w0q46==4 | w0q46==5
replace noliked=noliked+1 if w1q95==3 | w1q95==4 | w1q95==5
replace noliked=noliked+1 if w2q96==3 | w2q96==4 | w2q96==5
replace noliked=noliked+1 if w3q77==3 | w3q77==4 | w3q77==5
replace noliked=noliked+1 if w4q62==3 | w4q62==4 | w4q62==5
replace noliked=noliked+1 if w6q80==3 | w6q80==4 | w6q80==5 
tab noliked

gen noansweredLQ=0
replace noansweredLQ=noansweredLQ+1 if w0q46!=. 
replace noansweredLQ=noansweredLQ+1 if w1q95!=.
replace noansweredLQ=noansweredLQ+1 if w2q96!=.
replace noansweredLQ=noansweredLQ+1 if w3q77!=.
replace noansweredLQ=noansweredLQ+1 if w4q62!=.
replace noansweredLQ=noansweredLQ+1 if w6q80!=.
tab noansweredLQ

gen LR=100*noliked/(noansweredLQ)
tab LR

by cntry: sum LR
by cntry: sum LR if silent==1
by cntry: sum LR if boomers==1
by cntry: sum LR if genX==1
by cntry: sum LR if millennials==1


*EE
prtesti 243 .714 194 .682 
prtesti 243 .714 301 .667
prtesti 243 .714 38 .634

*GB
prtesti 189 .754 285 .805
prtesti 189 .754 333 .790
prtesti 189 .754 58 .704

*SI
prtesti 205 .928 161 .931
prtesti 205 .928 279 .930
prtesti 205 .928 28 .956



*************************
*Data quality
**************************

********nonresponse wave 1

gen numbermissingsw1=0 if w1disp_status==31 | w1disp_status==32
*we only focus here on the ones who finished the survey (since breakoff is already considered before)

forvalues x=1/61 {
replace numbermissingsw1=numbermissingsw1+1 if w1q`x'==.
}

tab numbermissingsw1

replace numbermissingsw1=numbermissingsw1+1 if w1q62a==. & w1q62b==.

forvalues x=1/15 {
replace numbermissingsw1=numbermissingsw1+1 if w1q63a_a`x'==.
}

forvalues x=64/82 {
replace numbermissingsw1=numbermissingsw1+1 if w1q`x'==.
}

replace numbermissingsw1=numbermissingsw1+1 if w1q83_filter==.

replace numbermissingsw1=numbermissingsw1+1 if w1q83_filter==1 & w1q83==.

forvalues x=84/96 {
replace numbermissingsw1=numbermissingsw1+1 if w1q`x'==.
}

replace numbermissingsw1=numbermissingsw1+1 if w1q97_a1==0 & w1q97_a2==0 & w1q97_a3==0 & w1q97_a4==0 & ///
 w1q97_a5==0 & w1q97_a6==0 & w1q97_a7==0 & w1q97_a8==0
 
replace numbermissingsw1=numbermissingsw1+1 if w1q98a==.
replace numbermissingsw1=numbermissingsw1+1 if w1q98b==. & w1q98a==1
replace numbermissingsw1=numbermissingsw1+1 if w1q98c==. & w1q98a==1

replace numbermissingsw1=numbermissingsw1+1 if w1q99_a1==0 & w1q99_a2==0 & w1q99_a3==0 & w1q99_a4==0 & ///
 w1q99_a5==0 & w1q99_a6==0 & w1q99_a7==0 & w1q99_a8==0 & w1q99_a9==0 & w1q99_a10==0 & w1q99_a11==0
 
tab numbermissingsw1

*number of question answered:
*since focus on the ones who finished the number is the same except for filters
*in total, there are a maximum of 115, out of these, 3 are asked only to some people depending on filters: w1q83 w1q98b and w1q98c
*so we generate number questions=115 and then we substract if they did not get the 3 with filters
gen numberquestions=115 if w1disp_status==31 | w1disp_status==32

replace numberquestions=numberquestions-1 if w1q83_filter==2
replace numberquestions=numberquestions-1 if w1q98a==2

tab numberquestions

gen rateofmissingsw1=numbermissingsw1*100/numberquestions

by cntry: sum rateofmissingsw1
by cntry: sum rateofmissingsw1 if silent==1
by cntry: sum rateofmissingsw1 if boomers==1
by cntry: sum rateofmissingsw1 if genX==1
by cntry: sum rateofmissingsw1 if millennials==1

*EE
prtesti  212 .02 278 .036
prtesti  212 .02 176 .046
prtesti  212 .02 38 .050

*SI
prtesti 142 .018 222 .036
prtesti 142 .018 135 .049
prtesti 142 .018 23 .091

*UK
prtesti 127 .009 254 .02
prtesti 127 .009 227 .017
prtesti 127 .009 44 .03


********primacy effects wave 1

sum

gen numberselectfirstw1=0 if w1disp_status==31 | w1disp_status==32

forvalues x=1/61 {
replace numberselectfirstw1=numberselectfirstw1+1 if w1q`x'==1
}

replace numberselectfirstw1=numberselectfirstw1+1 if w1q62a==1 | w1q62b==1

forvalues x=1/15 {
replace numberselectfirstw1=numberselectfirstw1+1 if w1q63a_a`x'==1
}

forvalues x=64/82 {
replace numberselectfirstw1=numberselectfirstw1+1 if w1q`x'==1
}

replace numberselectfirstw1=numberselectfirstw1+1 if w1q83_filter==1

replace numberselectfirstw1=numberselectfirstw1+1 if w1q83==1

forvalues x=84/96 {
replace numberselectfirstw1=numberselectfirstw1+1 if w1q`x'==1
}

replace numberselectfirstw1=numberselectfirstw1+1 if w1q97_a1==1 
 
replace numberselectfirstw1=numberselectfirstw1+1 if w1q98a==1
replace numberselectfirstw1=numberselectfirstw1+1 if w1q98b==1
replace numberselectfirstw1=numberselectfirstw1+1 if w1q98c==1

replace numbermissingsw1=numbermissingsw1+1 if w1q99_a1==1 

tab numberselectfirstw1

gen rateofselectfirstw1=numberselectfirstw1*100/numberquestions

sum rateofselectfirstw1

by cntry: sum rateofselectfirstw1
by cntry: sum rateofselectfirstw1 if silent==1
by cntry: sum rateofselectfirstw1 if boomers==1
by cntry: sum rateofselectfirstw1 if genX==1
by cntry: sum rateofselectfirstw1 if millennials==1


*EE
prtesti  212 .228 278 .221
prtesti  212 .228 176 .234
prtesti  212 .228 38 .274

*SI
prtesti 142 .217 222 .241
prtesti 142 .217 135 .232
prtesti 142 .217 23 .250

*UK
prtesti 124 .246 254 .239
prtesti 124 .246 227 .246
prtesti 124 .246 44 .251


********************************
*Regressions
*******************************


gen EE=0
replace  EE=1 if cntry=="EE"

gen SI=0
replace SI=1 if cntry=="SI"

gen GB=0
replace GB=1 if cntry=="GB"

gen univ=.
replace univ=1 if eisced==6 | eisced==7
replace univ=0 if eisced==1 | eisced==2 | eisced==3 | eisced==4 | eisced==5
tab uni

reg PR genX boomers silent SR DR LR gndr univ pdwrk ctzcntr EE SI, beta
stdBeta, se

reg BR genX boomers silent SR DR LR gndr univ pdwrk ctzcntr EE SI, beta
stdBeta, se

reg rateofmissingsw1 genX boomers silent SR DR LR gndr univ pdwrk ctzcntr EE SI, beta
stdBeta, se

reg rateofselectfirstw1 genX boomers silent SR DR LR gndr univ pdwrk ctzcntr EE SI, beta
stdBeta, se

reg DR genX boomers silent SR gndr univ pdwrk ctzcntr EE SI, beta
stdBeta, se

reg LR genX boomers silent SR gndr univ pdwrk ctzcntr EE SI, beta
stdBeta, se

********************************
*Appendices
*******************************

*Appendix 1

sort cntry
by cntry: tab gndr
by cntry: tab univ
by cntry: tab pdwrk
by cntry: tab ctzcntr

*Appendix 2 (only main effects without control)

reg PR genX boomers silent, beta
stdBeta, se

reg BR genX boomers silent, beta
stdBeta, se

reg rateofmissingsw1 genX boomers silent, beta
stdBeta, se

reg rateofselectfirstw1 genX boomers silent, beta
stdBeta, se

reg DR genX boomers silent, beta
stdBeta, se

reg LR genX boomers silent, beta
stdBeta, se