Cross Validated Asked by Nikolay Bogdanov on December 23, 2021
I am working on research in R using a dataset from the German Comparative Candidate Survey (a survey of political candidates during election campaigns). I would like to run a regression with a dichotomous dependent variable (if an MP should follow the voters position on a certain issue or the party position) and mostly ordinal independent variables – e.g. immigrants should be required to adapt to the customs of the country (strongly agree, agree, neither, disagree, strongly disagree) or politics should abstain from intervening in the economy (strongly agree, agree, neither, disagree, strongly disagree) etc. Last but not least, I have a few standard demographic independent variables – gender, age & subjective size of hometown. Nevertheless, I’m not sure which type of regression to use. My intuition tells me that the logistic regression fits best. Am I correct?
Note: I have not yet done the data cleansing – variable recoding, removing missings etc.
This is how the head of the dataset looks like:
t0 t1 t2 t3 t4 t5 t6 t7 t8 t9 a1 a2
1 3,20E+12 Germany 5 2009 -1 -1 999999 mail Question not asked or data not available Q not asked 6 2007
2 3,20E+12 Germany 8 2009 -1 -1 999999 mail Question not asked or data not available Q not asked 1 1970
3 3,20E+12 Germany 9 2009 -1 -1 999999 mail Question not asked or data not available Q not asked 4 1974
4 3,20E+12 Germany 10 2009 -1 -1 999999 mail Question not asked or data not available Q not asked 2 1982
a3a a3b1 a3b2 a3b3 a4a1 a4a2 a4a3 a4a4 a4a5 a4a6 a4b1
1 no -1 -1 -1 Q not asked Q not asked Q not asked Q not asked Q not asked Q not asked Q not asked
2 no -1 -1 -1 Q not asked Q not asked Q not asked Q not asked Q not asked Q not asked Q not asked
3 no -1 -1 -1 Q not asked Q not asked Q not asked Q not asked Q not asked Q not asked Q not asked
4 no -1 -1 -1 Q not asked Q not asked Q not asked Q not asked Q not asked Q not asked Q not asked
a4b2 a4b3 a4b4 a4b5 a4b6 a5a a5b a6a a6b a7a1 a7a2 a7a3 a7a4 a7a5
1 Q not asked Q not asked Q not asked Q not asked Q not asked no 97 no 97 yes no yes no no
2 Q not asked Q not asked Q not asked Q not asked Q not asked no 97 no 97 yes no yes yes no
3 Q not asked Q not asked Q not asked Q not asked Q not asked no 97 no 97 no no no yes no
4 Q not asked Q not asked Q not asked Q not asked Q not asked no 97 no 97 yes no no no no
a7a6 a7a7 a7a8 a7a9 a7b1 a7b2 a7b3 a7b4 a7b5 a7b6 a7b7 a7b8 a7b9 a8a a8b a8c a9a a9b a9c a9d a9e
1 no no no Q not asked 1 997 4 997 997 997 997 997 -1 0 0 0 0 -1 0 0 0
2 no yes no Q not asked 999 997 999 999 997 997 999 997 -1 12 10 0 99 -1 99 99 12
3 yes no yes Q not asked 997 997 997 2 997 4 997 5 -1 20 99 99 99 -1 99 99 20
4 yes yes no Q not asked 2 997 997 997 997 4 4 997 -1 0 0 0 0 -1 0 0 0
a9f a9g a9h a10a a10b a11a1 a11a2 a11a3 a11a4 a11b1 a11b2 a11b3 a11b4 a12
1 0 0 0 no answer 99 no answer no answer no answer no answer -1 -1 -1 -1 4
2 99 99 99 no answer 99 no answer no answer no answer no answer -1 -1 -1 -1 2
3 99 99 99 no answer 99 no answer no answer no answer no answer -1 -1 -1 -1 10
4 0 0 0 no answer 99 no answer no answer no answer no answer -1 -1 -1 -1 5
b1a b1b b1c b2 b3a b3b b4 b5a1 b5a2 b5a3
1 Q not asked no answer no answer Q not asked Q not asked Q not asked 320 no answer no answer no answer
2 Q not asked no answer no answer Q not asked Q not asked Q not asked 999 no answer no answer no answer
3 Q not asked no answer no answer Q not asked Q not asked Q not asked 80 no answer no answer no answer
4 Q not asked no answer no answer Q not asked Q not asked Q not asked 60 no answer no answer no answer
b5a4 b5a5 b5a6 b5a7 b5a8 b5a9 b5a10 b5b1 b5b2 b5b3
1 no answer no answer no answer no answer no answer no answer no answer no answer no answer no answer
2 no answer no answer no answer no answer no answer no answer no answer no answer no answer no answer
3 no answer no answer no answer no answer no answer no answer no answer no answer no answer no answer
4 no answer no answer no answer no answer no answer no answer no answer no answer no answer no answer
b5b4 b5b5 b5b6 b5b7 b6a1 b6a2 b6a3 b6a4 b6a5 b6a6 b6a7 b6b1
1 no answer no answer no answer no answer no no no no no no yes not applicable
2 no answer no answer no answer no answer no no no no no no no not applicable
3 no answer no answer no answer no answer no no no no no no no not applicable
4 no answer no answer no answer no answer no yes no yes no no no not applicable
b6b2 b6b3 b6b4 b6b5 b6b6 b6b7
1 not applicable not applicable not applicable not applicable not applicable not very important
2 not applicable not applicable not applicable not applicable not applicable not applicable
3 not applicable not applicable not applicable not applicable not applicable not applicable
4 fairly important not applicable very important not applicable not applicable not applicable
b6c1 b6c2 b6c3 b6c4 b6c5 b6c6 b6c7 b7 b8a
1 Q not asked Q not asked Q not asked Q not asked Q not asked Q not asked Q not asked 1 no
2 Q not asked Q not asked Q not asked Q not asked Q not asked Q not asked Q not asked 99 no answer
3 Q not asked Q not asked Q not asked Q not asked Q not asked Q not asked Q not asked 0,6666667 yes
4 Q not asked Q not asked Q not asked Q not asked Q not asked Q not asked Q not asked 1 no
b8b
1 9996 no (further) issue mentioned
2 9996 no (further) issue mentioned
3 Regenerative Energie
4 9996 no (further) issue mentioned
b8c
1 -1 Q not asked
2 -1 Q not asked
3 -1 Q not asked
4 -1 Q not asked
b8d
1 -1 Q not asked
2 -1 Q not asked
3 -1 Q not asked
4 -1 Q not asked
b8e
1 -1 Q not asked
2 -1 Q not asked
3 -1 Q not asked
4 -1 Q not asked
b8f
1 -1 Q not asked
2 -1 Q not asked
3 -1 Q not asked
4 -1 Q not asked
b9a b9b b9c b9d b9e b9f b9g b10a b10b b10c b10d b10e b11a1 b11a2 b11b
1 no yes yes yes no yes no no answer no answer no answer no answer no answer 3 1 -1
2 no no no no no no no no answer no answer no answer no answer no answer 0 0 -1
3 no no no no no yes no no answer no answer no answer no answer no answer 0 0 -1
4 no no yes no no no no no answer no answer no answer no answer no answer 0 0 -1
b11c1 b11c2 b11d b12a
1 Q not asked -1 -1 no
2 Q not asked -1 -1 no answer
3 Q not asked -1 -1 no
4 Q not asked -1 -1 no
b12b
1
2
3
4
b13 b14a b14b b14c b15 b16
1 13000 100 0 0 9 no answer
2 0 0 0 0 9 no answer
3 500 0 0 100 9 no answer
4 1000 0 0 100 9 no answer
c1a
1 Krieg
2 soziale Gerechtigkeit
3 Finanzkrise
4 Demographie
c1b
1 soziale Ungerechtigkeit
2 Friedenpolitik
3 Bildung
4 Wirtschaftskrise
c1c
1 Krieg
2 Umweltpol.
3 Natur u. Umwelt
4 Umweltschutz
c2a c2b c2c c2d c2e
1 disagree strongly disagree strongly agree strongly disagree strongly agree
2 disagree disagree strongly disagree strongly disagree neither
3 agree agree neither strongly disagree strongly disagree
4 strongly agree neither strongly agree neither strongly disagree
c2f c2g c2h c2i c2j c2k
1 strongly disagree strongly agree strongly agree strongly agree no answer strongly agree
2 neither strongly agree agree strongly disagree agree strongly disagree
3 neither strongly agree disagree neither agree agree
4 disagree strongly disagree strongly disagree strongly disagree agree strongly disagree
c2l c2m c2n c3 c3a c4 c5 c6 c7
1 strongly disagree strongly agree no answer 0 -1 a good thing 10 not at all satisfied Q not asked
2 disagree strongly disagree disagree 3 -1 a good thing 7 not very satisfied Q not asked
3 agree strongly agree disagree 7 -1 a good thing 4 not very satisfied Q not asked
4 strongly agree strongly agree neither 5 -1 a good thing 10 fairly satisfied Q not asked
c7a c7b c7c c7d c7e c7f c7g c7h c7i c7k d1 d2 d3a d3b d3c d3d d3e d3f
1 -1 -1 -1 -1 -1 -1 -1 -1 -1 -1 not very satisfied not very well 3 -1 5 1 2nd rank 4
2 -1 -1 -1 -1 -1 -1 -1 -1 -1 -1 fairly satisfied fairly well 9 -1 1 9 no answer 1
3 -1 -1 -1 -1 -1 -1 -1 -1 -1 -1 not very satisfied fairly well 9 -1 1 2 no answer 3
4 -1 -1 -1 -1 -1 -1 -1 -1 -1 -1 very satisfied very well 4 -1 2 3 5th rank 1
d4a d4b d4c d5a d5b d5c d5d
1 no answer no answer no answer no answer no answer no answer no answer
2 voter opinion own opinion own opinion agree agree neither agree
3 party position own opinion own opinion agree agree disagree disagree
4 voter opinion own opinion own opinion strongly agree strongly disagree agree strongly agree
d5e d5f d5g d5h d6a d6b
1 no answer no answer strongly disagree strongly agree strongly agree strongly disagree
2 agree agree strongly agree agree agree disagree
3 neither disagree agree agree agree neither
4 strongly disagree disagree strongly agree strongly disagree strongly disagree strongly agree
d6c d7a d7b d7c d7d e1 e2 e3 e3a
1 neither no answer Q not asked no answer no answer male 1961 in [country] -1
2 agree agree Q not asked agree strongly agree male 1945 in [country] -1
3 agree agree Q not asked neither agree male 1945 in [country] -1
4 strongly disagree strongly agree Q not asked disagree disagree female 1961 in [country] -1
e4 e4a e4b e6a e6b e7 e7a e8
1 large town or city -1 48 -1 Q not asked other, not in labour force -1 -1
2 small or middle-sized town -1 63 -1 Q not asked employed full-time [32 hrs or more per week] -1 -1
3 rural area of village -1 64 -1 Q not asked retired -1 -1
4 rural area of village -1 48 -1 Q not asked employed part-time [15-32 hrs] -1 -1
e9 e10 e11 e12a
1 never None married or living as married Q not asked
2 never None married or living as married Q not asked
3 2 or 3 times a month Protestant, no denomination given married or living as married Q not asked
4 a number of times a year Roman Catholic married or living as married Q not asked
e12b e12c e13 e14 e15a e15b
1 Q not asked Q not asked -1 -1 -1 -1
2 Q not asked Q not asked -1 -1 -1 -1
3 Q not asked Q not asked -1 -1 -1 -1
4 Q not asked Q not asked -1 -1 -1 -1
[ reached 'max' / getOption("max.print") -- omitted 2 rows ]
My variables:
I would use logistic regression. In order to make sure this is the correct model you'll need to make sure you satisfy a few assumptions which are outlined here.
https://www.statisticssolutions.com/assumptions-of-logistic-regression/
Answered by Harry on December 23, 2021
Get help from others!
Recent Answers
Recent Questions
© 2024 TransWikia.com. All rights reserved. Sites we Love: PCI Database, UKBizDB, Menu Kuliner, Sharing RPP