scieee Science in your language
[en] (orig)

Socio-Spatial Analysis of Inequality with SoRa-Service: A Geolinking Approach

Author: Rieche, Theodor; Ehrhardt, Denise; Jung, Alexander; Eichhorn, Sebastian; Jünger, Stefan; Zapilko, Benjamin; Goebel, Jan; Sikder, Sujit Kumar; Meinel, Gotthard
Publisher: Zenodo
DOI: 10.5281/zenodo.17638158
Source: https://zenodo.org/records/17638158/files/Rieche_et_al_2025_ILUS_SocioSpatial_Housing.pdf
Socio-Spa ial Analysis o Inequali y wi h
SoRa-Se ice: A Geolinking App oach
Theodo Rieche, Denise Eh ha d , Alexande Jung, Sebas ian Eichho n, S e an
Jünge , Benjamin Zapilko, Jan Goebel, Suji Kuma Sikde , Go ha d Meinel
Image c edi : D. Eh ha d
A e esiden s o single- amily
houses eally happie han hose
li ing in high-densi y housing?
→Ques ion add esses housing
equi y
→Such and simila ques ions
o en canno be answe ed
alone wi h a ailable s a is ical
da a
In o
07.11.2025 ILUS - In e na ional Land Use Symposium 2025
2
Sou ce: D. Eh ha d
Compa e: Ode ma , R., & S u ze , A. (2022). Does he D eam o Home Owne ship Res Upon Biased Belie s? A Tes Based on P edic ed and
Realized Li e Sa is ac ion. Jou nal o Happiness S udies. h ps://doi.o g/10.1007/s10902-022-00571-w
▪Resea ch ques ions
▪Da a linking wi h Geolinking Se ice SoRa
▪Da a
▪SOEP Su ey Da a
▪U ban S uc u e Type Da a
▪Me hodology
▪Resul s
▪Conclusion
▪Ou look
Agenda
07.11.2025 ILUS - In e na ional Land Use Symposium 2025
3
▪RQ1: How can su ey da a om he Socio-Economic Panel (SOEP) be
linked o small-scale U ban S uc u e Type da a and wha po en ial
does i ha e o socio-spa ial equi y esea ch in he con ex o housing
equi y?
▪RQ2: Wha cha ac e ises he esiden s o di e en ypes o u ban
s uc u es in Hesse? (An explo a o y app oach)
Resea ch ques ions
07.11.2025 ILUS - In e na ional Land Use Symposium 2025
4
▪Linking su ey and spa ial da a o socio-spa ial esea ch
▪Use in e ace: a “so a” package in R language
▪O e s p e-de ined linking me hods and da ase s
▪P i a e o public mode (inside o ou side a secu e oom o RDC)
▪P i acy-complian & FAIR p inciples
▪Cu en ly SOEP and IOER Moni o da a →ex endable o o he RDCs
▪O e ing syn he ic su ey da a o p epa e he linking ou side o RDC
New Geolinking Se ice SoRa
07.11.2025 ILUS - In e na ional Land Use Symposium 2025
5
Key ac s

New Geolinking Se ice SoRa
07.11.2025 ILUS - In e na ional Land Use Symposium 2025
6
Gene al wo k low o da a linking
Spa ial ea u e
Su ey a iable
▪Annual su eys (“wa es”) since 1984
▪Rep esen a i e sample o Ge many
(sample size: abou 20.000 households wi h
abou 30.000 indi iduals)
▪Limi ed access o spa ial e e ence o su ey
pa icipan s o a oid e-iden i ica ion
▪Real add ess coo dina es usable only on-si e
▪Fede al s a e le el usable wi h da a use ag eemen
▪Syn he ic SOEP S uc u al Da ase o es ing
Da ase 1: SOEP Su ey Da a
07.11.2025 ILUS - In e na ional Land Use Symposium 2025
7
Socio-Economic Panel (SOEP)
▪Use ul documen a ion:
▪SOEP Companion o panelda a.o g
Goebel, Jan e al. (2018). The Ge man Socio-Economic Panel (SOEP). Jah büche ü Na ionalökonomie und S a is ik. 239. 10.1515/jbns -2018-0022.
© DIW Be lin/Sand a Bohmann
▪100 m g idded da a wi h
U ban S uc u e Type (SST)
▪Spa ial ex en : Hesse
(Ge many)
▪Times amp: 2022
▪Da a collec ion me hod:
▪AI-based classi ica ion (XGBoos )
▪Based on o icial building da a (LoD2),
cadas al pa cels and emo e sensing
da a
▪O e all accu acy: 95 %
Da ase 2: U ban S uc u e Types in Hesse (SST)
07.11.2025 ILUS - In e na ional Land Use Symposium 2025
8
Kelle , Sina e al. (2025). En wicklung on Planungshil en ü Klimaschu z und Klimaanpassung in de äumlichen Gesam planung
mi els Fe ne kundung - Abschlussbe ich . Hessisches Minis e ium ü Wi scha , Ene gie, Ve keh , Wohnen und ländlichen Raum. 10.5445/IR/1000182412
High-densi y
buildings
Mul i- amily
houses
High- ise
buildings
Row
apa men
buildings
Ou buildings
Low- ise
buildings on
la ge a ea
Single- amily
houses
Pe ime e
block
▪Explo a i e da a analysis wi h 24 su ey a iables
▪Spa ial agg ega ion based on household le el
▪Including a weigh ing ac o o imp o e
ep esen a i eness
▪Simple poin - o- as e linking (me hod: “lookup”)
Me hodology
07.11.2025 ILUS - In e na ional Land Use Symposium 2025
9
Wo k low
Resul s
07.11.2025 ILUS - In e na ional Land Use Symposium 2025
16
Noda a also includes owne s
who we e no asked his ques ion.

Resul s
07.11.2025 ILUS - In e na ional Land Use Symposium 2025
17
High-densi y buildings 7,67
Mul i- amily houses 7,84
Row apa men buildings 7,30
Pe ime e block 7,50
Single- amily houses 8,07
Big ci y 7,59
Medium-sized own 7,82
La ge small own 8,07
Small own 8,10
Ru al municipali y 7,78
▪The SOEP households we e mos ly assigned o esiden ial building ypes and a e well
dis ibu ed ac oss municipali y ypes (excep u al municipali ies).
▪Dwellings wi h many ooms a e mo e likely o be ound in single- and mul i- amily
houses, as well as ou side o big ci ies.
▪Residen s o single- amily houses, small owns and u al municipali ies change hei
homes less equen ly.
▪Ne household income is highes in single- and mul i- amily houses, as well as in u al
municipali ies.
▪Social housing is o en loca ed in ow apa men buildings in big ci ies.
▪Sa is ac ion wi h dwelling is gene ally qui e high and ela i ely consis en (wi h sligh
peaks o single- amily houses, small owns and la ge small owns).
Conclusion
07.11.2025 ILUS - In e na ional Land Use Symposium 2025
18
RQ2
▪Times amps o selec ed da ase s (2022 s. 2020)
▪Sample Size and s a is ical con iden iali y (e.g. spa ial agg ega ion o sha e o noda a)
▪Agg ega ion be ween SOEP households and SOEP indi iduals
▪Weigh ing ac o o households and indi iduals in su ey da a
▪Posi ional accu acy in 100 m g id cell (building oo p in s. add ess) & pa ly missing
co e age o add esses
▪Unce ain ies in classi ica ion o U ban S uc u es Types
Conclusion
07.11.2025 ILUS - In e na ional Land Use Symposium 2025
19
RQ2 - Limi a ions
▪Geolinking Se ice SoRa was used o link an ex e nal spa ial da ase (a IOER D esden)
o highly p o ec ed SOEP su ey da a
▪Demons a ed, how SOEP da a and small-scale U ban S uc u e Types can be linked
▪Spa ial ypologies can help o agg ega e su ey da a o he equi ed pu pose
▪SOEP Su ey Da a co e s a ious opics & o e s eal add ess coo dina es
▪SOEP Su ey Da a a e highly p o ec ed o a oid e-iden i ica ion o su ey pa icipan s
▪Linkage can be p epa ed om ou side he SOEP secu e oom using syn he ic da a, o
sa e wo king ime on-si e
Conclusion
07.11.2025 ILUS - In e na ional Land Use Symposium 2025
20
RQ1
▪Va ious esea ch ques ions can now be answe ed wi h he Geolinking Se ice SoRa
▪Fo example in he ield o housing/en i onmen /mobili y equi y
▪Using su ey da a om whole Ge many could be use ul (sample size)
▪Linking u he spa ial da ase s, e.g. based on densi y o accessibili y indica o s
▪Include neighbou hood o su ey pa icipan s (ci cle, isoch ones e c.)
Tes pe iod wi h Geolinking Se ice SoRa s a s in Janua y 2026
A e you in e es ed o en ich you spa ial model wi h indi idual pe spec i es,
beha iou , o pe cep ions?
Ou look
07.11.2025 ILUS - In e na ional Land Use Symposium 2025
21

www.ioe .de/en
Thank you
o you
a en ion!
Theodo Rieche
[email p o ec ed]
www.ioe .de
U ban S uc u e Types in Ma bu g / Hesse, Google Maps
Geolinking Se ice SoRa
[email p o ec ed]
www.so a-se ice.o g
A e you in e es ed in linking you spa ial
science models wi h social science su ey
da a?→Tes pe iod s a s in Janua y 2026 !
Resul s
23
Resul s
24
High-densi y buildings 1969
Mul i- amily houses 1953
Row apa men buildings 1965
Pe ime e block 1945
Single- amily houses 1968
Big ci y 1960
Medium-sized own 1972
La ge small own 1964
Small own 1963
Ru al municipali y 1958
Resul s
25
High-densi y buildings 85,93
Mul i- amily houses 114,45
Row apa men buildings 63,78
Pe ime e block 71,17
Single- amily houses 114,45
Big ci y 84,93
Medium-sized own 102,92
La ge small own 106,62
Small own 124,54
Ru al municipali y 135,68
Resul s
32
High-densi y buildings 7,14
Mul i- amily houses 7,29
Row apa men buildings 7,52
Pe ime e block 7,51
Single- amily houses 7,23
Big ci y 7,24
Medium-sized own 7,17
La ge small own 7,43
Small own 7,20
Ru al municipali y 7,33

Resul s
33
Resul s
34
Resul s
35
Resul s
36
Resul s
37
Noda a also includes owne s
who we e no asked his ques ion.

Resul s
38