Resea ch Pape
This wo k is licensed unde a C ea i e Commons A ibu ion-NonComme cial 4.0 In e na ional License.
DOI: 10.5281/zenodo.17322626
Page 23
1. In oduc ion
Reading beha io o sociological books—de ined as sus ained
engagemen wi h ma e ials explo ing hemes like social s uc u e,
cul u al iden i y, inequali y, and communi y dynamics—does no
ope a e in isola ion. Ins ead, i is shaped by empo al pe sis ence
ac o s: long- e m in luences ha os e consis en in e ac ion wi h
sociological con en . These ac o s include epea ed eading
in e en ions, consis en esou ce access, endu ing eading habi s,
and en i onmen al suppo .
A bu geoning body o schola ship iden i ies con en in e es as a
pi o al media o in he ela ionship be ween hese empo al
pe sis ence ac o s and eading ou comes—including sus ained
engagemen wi h sociological li e a u e, mas e y o sociological
concep s, and li elong engagemen wi h disciplina y eading. Wi hin
his con ex , con en in e es deno es eade s’ subjec i e a ini y o
sociological hemes (e.g., acial equi y, demog aphic aging, he
socie al implica ions o echnology), cul u ally esonan con en
(e.g., ex s examining diaspo ic iden i ies o Indigenous social
sys ems), and ounda ional disciplina y opics (e.g., social
The Media ing Role o Con en In e es in he Rela ionship Be ween Tempo al
Pe sis ence Fac o s and Reading Beha io o Sociological Books
Jiexuan Liu
Nanjing No mal Uni e si y, Nanjing, Jiangsu, China
Co esponding Au ho : Jiexuan Liu
This s udy in es iga es he bo owing beha io o Ca ego y C (Sociology, Chinese Lib a y Classi ica ion) books a Nanjing
No mal Uni e si y (2016–2024), using a sample o ci cula ion da a co e ing 13,117 book i les, 13,344 eade s, and 59,198
loans, ia a mul i-me hod app oach. Fo heme analysis, subjec e ms om he 6XX ield (co e o CNMARC/UNIMARC
subjec desc ip ion) we e English- ansla ed, p ep ocessed ( okeniza ion, wo d equency coun ing), and seman ically
ec o ized. Voyan Tools isualized he op 10 subjec ca ego ies, 100 high- equency e ms, and c oss-pe iod clus e s,
e ealing eade s’ hema ic p e e ences. Linea eg ession ook “Leng h o Bo owing Yea s” and “Ca ego y C9” as
independen a iables, and “To al Bo owings” as he dependen a iable; model me ics con i med hese a iables as
signi ican di ec p edic o s. Media ion e ec analysis e i ied C9’s signi ican media ing ole be ween “Leng h o
Bo owing Yea s” and “To al Bo owings.” Decision ee analysis used “Renewal” as he dependen a iable, and “To al,”
“Gende ,” “Pe sis ence” as independen a iables, iden i ying “To al” and “Pe sis ence” as key enewal d i e s. This
s udy cla i ies sociological book bo owing’s empo al pa e ns, hema ic p e e ences, and in luencing ac o s, p o iding
empi ical suppo o lib a y esou ce alloca ion and se ice op imiza ion.
KEY WORDS:
Sociology Books; Theme Analysis; Linea Reg ession; Media ion E ec Analysis; Decision T ee Analysis;
Bo owing Beha io
Abb e ia ed Key Ti le: UAI J A s Humani Soc Sci
ISSN: 3048-7692 (Online)
Jou nal Homepage: h ps://uaipublishe .com/uaijahss/
Volume- 2 Issue- 10 (Oc obe ) 2025
F equency: Mon hly
Resea ch Pape
This wo k is licensed unde a C ea i e Commons A ibu ion-NonComme cial 4.0 In e na ional License.
DOI: 10.5281/zenodo.17322626
Page 24
s a i ica ion, ins i u ional sociology). E en con ex ual empo al
d i e s—such as book bans d awing a en ion o con o e sial
sociological con en o pandemic- ela ed conce ns abou socie al
esilience—spa k con en in e es , mo i a ing eade s o seek ou and
pe sis in eading ele an sociological books (Anan hak ishnan e
al., 2025; Buse a e al., 2024).
This esea ch ad ances he ield o sociological book bo owing
beha io s udies h ough ou key inno a ions: Fi s , i es ablishes a
mul i-me hod mechanis ic amewo k, in eg a ing empo al end
analyses ( ia ―Leng hs o Bo owing Yea s‖), linea eg essions ( o
di ec p edic o s o o al bo owings), and Boo s ap-based
media ion analyses ( o e eal C9’s media ing ole be ween
bo owing du a ions and o al bo owings), mo ing beyond
desc ip ions o explain why bo owing pa e ns eme ge. Second, i
adop s s anda dized me ada a-d i en hema ic analyses—using 6XX
ields wi h Voyan Tools’ seman ic ec o iza ion and c oss-pe iod
isualiza ions— o iden i y op hema ic ca ego ies, high- equency
e ms, and e olu iona y ends (e.g., he 2022–2024 shi o
― echnology in eg a ion‖), ensu ing objec i e, eplicable insigh s.
Thi d, i de elops in e p e able enewal p edic ions ia he CRT
algo i hm (a ―whi e-box‖ model) wi h p edic o s like ―To al
Bo owings‖ and ―Pe sis ence,‖ alida ing pe o mances ac oss
h ee pe iods (2016–2018, 2019–2021, 2022–2024) and highligh ing
he isks o excluding ime- a ying ac o s. Fou h, i con ex ualizes
indings wi h he CLC, de ailing he con en s o C Ca ego y sub-
ca ego ies (e.g., C0, C9) and linking C9 o bo owing/media ion
ou comes, explaining C9’s dominan bo owing olumes and
b idging abs ac da a wi h lib a y collec ion logics o ac ionable
esou ce alloca ions.
This s udy ca ies heo e ical, me hodological, and p ac ical alue
o lib a y and in o ma ion science, sociology o educa ion, and
eade beha io esea ch: heo e ically, i en iches he eade
beha io esea ch amewo k in sociology by iden i ying he dual
end o s able co e hemes (e.g., esea ch me hods, social s uc u e)
and e ol ing eme ging opics (e.g., echnology in eg a ion, aging,
eminis s udies) in eade s’ a en ion, and h ough media ion e ec
analysis, empi ically e i ies ha C9 (Sociology & Demog aphy)
pa ially media es he ela ionship be ween ―Leng h o Bo owing
Yea s (LBY)‖ and ―To al Bo owings (TB)‖, e ealing he di ec
and indi ec ( ia p e e ence o high-demand disciplines) impac s o
long- e m bo owing pe sis ence on o al bo owings and
o e coming p io o e eliance on desc ip i e co ela ions;
me hodologically, i p omo es s anda diza ion and in eg a ion by
adop ing s anda dized me ada a-d i en hema ic analysis (ensu ing
objec i i y and eplicabili y) and in eg a ing mul i-me hod
app oaches ( empo al end analysis, linea eg ession, media ion
analysis, CRT decision ees) o explo e di e en eade beha io
dimensions, while i s empo al alida ion o p edic ion models
(2016–2018, 2019–2021, 2022–2024) add esses he o e sigh o
igno ing ime- a ying ac o s (e.g., declining p edic i e ole o
―Gende ‖ la e ) in p e ious enewal p edic ion s udies; p ac ically, i
p o ides ac ionable guidance— o academic lib a ies, i aids
collec ion de elopmen by iden i ying C9 as he mos bo owed
subca ego y and mapping high- equency hemes o p io i ize high-
demand acquisi ions and op imize collec ions, and o highe
educa ion ins i u ions and sociological depa men s, i in o ms
cu iculum and eaching esou ce de elopmen by aligning ma e ials
wi h eade s’ hema ic p e e ences o connec academic inqui y wi h
eal-wo ld applica ions.
2. P e ious Resea ch
2.1 Con en In e es as a Media o o Engagemen ,
Mo i a ion, and Tempo al Pe sis ence
Tempo al pe sis ence ac o s like a ge ed eading in e en ions o
epea ed mo i a ional suppo ely on con en in e es o ansla e
in o sus ained sociological eading beha io . Fo ins ance, Chilu ya
e al. (2025) de eloped a social obo (―BookBo ‖) ha s imula es
ou h-g ade s uden s’ eading in e es h ough sociologically
ele an in e ac ions—including book con en discussions, cha ac e
analysis ied o social hemes, and pe sonalized ecommenda ions.
Thei indings show ha con en in e es , spa ked by he obo ’s
in e en ions, enhances eading engagemen , indi ec ly linking his
empo al pe sis ence ac o ( obo -media ed epea ed in e ac ion) o
consis en eading beha io . This aligns wi h Wu e al. (2023), who
ound ha o high school s uden s, eading mo i a ion (a empo al
d i e o pe sis ence) p edic s sa is ac ion wi h sociological books
ia sel -e icacy—and con en in e es ac s as a key d i e o sel -
e icacy, b idging mo i a ion o sus ained eading.
Con en in e es also add esses ba ie s o engagemen , pa icula ly
o eade s acing disciplina y o language challenges. Nausa e al.
(2024) s udied Colombian mas e ’s s uden s s uggling wi h access
o sociological esea ch (due o low English p o iciency and poo
eading skills) and ound ha ma ching eading con en o s uden s’
disciplina y in e es s (e.g., social inequali y, communi y
de elopmen ) elimina ed hese ba ie s. By aligning ma e ials wi h
p e-exis ing con en in e es , empo al ac o s like epea ed exposu e
o accessible ex s ansla ed in o consis en engagemen . Simila ly,
Hiebe (2023) emphasized ha eading amoun (a di ec measu e o
empo al pe sis ence) is c i ical o mas e ing sociological
disciplina y language and concep s—bu con en in e es ein o ces
his pe sis ence, ensu ing eade s main ain engagemen long enough
o build amilia i y wi h specialized hemes.
Cul u al ele ance eme ges as a sub-dimension o con en in e es
ha s ongly media es engagemen . Jones and Lynch (2023)
examined Black s uden s’ a e -school eading o A ican Ame ican
li e a u e ( ea u ing sociological hemes o ace, iden i y, and
sys emic opp ession) and ound ha con en in e es in cul u ally
esonan opics media es sus ained eading and comp ehension o
sociological concep s. Fo hese s uden s, empo al ac o s like
epea ed a e -school eading ime only led o meaning ul
engagemen when ma e ials e lec ed hei cul u al expe iences—
highligh ing con en in e es ’s ole in alida ing eade s’ iden i ies
and sus aining pe sis ence.
2.2 En i onmen al and Con ex ual Fac o s: Home, School,
and Digi al Pla o ms
The home li e acy en i onmen —de ined by ac o s like book
quan i y, pa en al in ol emen , and consis en access o sociological
ma e ials—ac s as a empo al pe sis ence ac o , wi h con en
in e es media ing i s impac on eading beha io . Mohammed e al.
(2023) s udied ca egi e -child eading dynamics and ound ha
sus ained ca egi e -child in e ac ion wi h sociological books
co ela es wi h child en’s li e acy de elopmen —bu his
ela ionship is ully media ed by con en in e es in social hemes.
Speci ically, home access o sociological books (a empo al
esou ce) only d i es engagemen i child en ind he con en (e.g.,
s o ies abou communi y, social oles) in e es ing.
Fo mul ilingual popula ions, HLE’s in luence is u he e ined by
con en in e es in cul u al and linguis ic con ex s. Aleksic and
Du us (2025) examined mul ilingual p eschoole s and ound ha
HLE ac o s (book quan i y, pa en al in e es in sociological hemes)
p edic eading skills—bu con en in e es in mul ilingual social
con ex s (e.g., s o ies abou cul u al di e si y, mul ilingual
communi y li e) media es his link. Fo hese child en, empo al
Resea ch Pape
This wo k is licensed unde a C ea i e Commons A ibu ion-NonComme cial 4.0 In e na ional License.
DOI: 10.5281/zenodo.17322626
Page 25
pe sis ence (e.g., daily eading ime wi h ca egi e s) ansla es in o
skill de elopmen only when ma e ials align wi h hei linguis ic and
cul u al in e es s.
2.3 Digi al and Technological Pla o ms: Gami ica ion and
Algo i hmic Recommenda ions
Digi al ools ep esen a mode n class o empo al pe sis ence
ac o s, as hey enable epea ed, scalable access o sociological
con en —and con en in e es media es hei e ec i eness. Yang e
al. (2023) de eloped gami ied po olios o s uden s, ea u ing goal-
se ing and gen e ewa ds ocused on sociological gen es (e.g.,
social jus ice, u ban s udies). Thei esul s show ha con en in e es
media es gami ied ac i i ies (a empo al ac o o epea ed, ewa d-
d i en in e ac ion) and pe sis en explo a ion o sociological
books—wi h in e es ed s uden s mo e likely o main ain engagemen
beyond ini ial gami ica ion p omp s.
Algo i hmic ecommenda ion sys ems, which p o ide ongoing
access o ailo ed con en , also ely on con en in e es o d i e
pe sis ence. T an e al. (2023) s udied social ela ions/ a ing-based
algo i hms ha ecommend sociological books and ound ha
con en in e es media es algo i hm-d i en sugges ions (a empo al
ac o o con inuous, pe sonalized exposu e) and consis en eading.
Reade s who ecei ed ecommenda ions aligned wi h hei in e es s
(e.g., aging s udies o hose conce ned wi h communi y dynamics)
we e a mo e likely o sus ain engagemen han hose ecei ing
gene ic sugges ions—con i ming con en in e es ’s ole in
ansla ing algo i hmic suppo in o long- e m beha io .
2.4 Con en In e es , Special Popula ions, and Li elong
Sociological Reading
Ea ly exposu e o sociological con en and con en in e es lay he
g oundwo k o li elong eading beha io , wi h empo al ac o s like
pa en al eading o cul u al esou ce access media ed by in e es .
Shigemasu e al. (2024) s udied in an -pa en eading and ound ha
pa en al eading o sociological- hemed books (e.g., s o ies abou
social s uc u es, communi y oles) o oddle s imp o es child en’s
ea ly comp ehension o social concep s—and ea ly con en in e es
in hese opics p edic s li elong engagemen wi h sociological books.
Fo hese child en, empo al ac o s like epea ed pa en al eading
only shape long- e m habi s i hey spa k in e es in social hemes
du ing ea ly de elopmen .
Indigenous child en ep esen a popula ion whe e cul u al ele ance
in con en in e es is pa icula ly c i ical. Sian u i and Hu i (2024)
examined decolonized pic u e books (in eg a ing Indigenous cul u al
expe iences and sociological hemes like cul u al iden i y,
communi y go e nance) o Indigenous child en and ound ha
con en ele ance ( ied o cul u al backg ound) media es epea ed
eading and sus ained engagemen . Fo hese child en, empo al
ac o s like egula access o cul u ally speci ic books ansla e in o
pe sis ence only when he con en e lec s hei li ed expe iences—
unde sco ing he need o cul u ally esponsi e con en o ac i a e
in e es .
Fo olde adul s, con en in e es media es empo al ac o s like
lib a y access o leisu e eading ime and engagemen wi h
sociological books—o en wi h spillo e e ec s on well-being.
O ani e al. (2025) s udied Japanese olde adul s and ound ha
lib a y book quan i y (a empo al ac o o consis en esou ce
access) educes unc ional disabili y isk—bu his ela ionship is
media ed by con en in e es in aging and communi y hemes. Olde
adul s who accessed sociological books aligned wi h hei in e es s
(e.g., s udies on in e gene a ional ela ionships, communi y suppo )
main ained mo e consis en eading habi s, d i ing heal h bene i s.
S i as a a, P. and S i as a a, M. (2023) ound ha Indian elde ly
adul s who ead sociological books on aging and in e gene a ional
dynamics expe ience educed loneliness—bu only i hey ha e
con en in e es in hese opics. Fo hese eade s, empo al ac o s
like daily leisu e eading ime only imp o e well-being when pai ed
wi h in e es -aligned con en .
2.5 C oss-Ou come Links: Con en In e es , Tempo al
Pe sis ence, and Beyond Reading Beha io
Con en in e es does no only media e empo al ac o s and eading
beha io —i also b idges sus ained sociological eading o b oade
ou comes like cogni i e unc ion, emo ional well-being, and cul u al
pa icipa ion.
Ea ly con en in e es in sociological hemes media es he link
be ween childhood empo al ac o s (e.g., book access) and la e
cogni i e heal h. Tani e al. (2024) ound ha a book- ich childhood
(a empo al ac o o sus ained ea ly esou ce access) os e s con en
in e es in sociological con en —and his in e es media es he
ela ionship be ween childhood book en i onmen and la e
cogni i e unc ion, pa icula ly o olde adul s wi h low educa ion.
Fo hese indi iduals, ea ly exposu e o sociological books only
p ese ed cogni i e unc ion i i spa ked in e es , ensu ing hey
main ained eading habi s in o adul hood.
Du ing pe iods o s ess, con en in e es in sociological hemes
media es empo al pe sis ence (e.g., sus ained eading) and
emo ional esilience. Buse a e al. (2024) s udied I alian uni e si y
s uden s du ing he pandemic and ound ha sus ained eading o
sociological books on communi y esilience mi iga es anxie y—bu
his e ec is media ed by con en in e es . S uden s who we e
in e es ed in communi y- ocused hemes main ained mo e consis en
eading habi s, leading o g ea e emo ional bene i s.
His o ical and egional s udies con i m ha con en in e es has long
media ed empo al pe sis ence ac o s and sociological eading
beha io , wi h con ex -speci ic hemes shaping engagemen .
Aga ono a (2023) examined ea ly 20 h-cen u y Russian am
passenge s, who ead sociological books ( ocused on u ban issues
like wo ke s ikes, communi y dynamics) du ing daily commu es.
The s udy ound ha con en in e es in hese con ex - ele an
hemes media ed egula am ime (a empo al ac o o daily,
epea ed eading oppo uni ies) and sociological book choice—
passenge s consis en ly selec ed ma e ials aligned wi h hei in e es s
in u ban social issues, ansla ing commu e ime in o sus ained
engagemen .
Simila ly, Ga cia-Gonzalez e al. (2023) s udied his o ical CRA
(Reading P omo ion Cen e s) p og ams in Chilean school lib a ies,
which his o ically ocused on social jus ice (a co e sociological
heme). They ound ha con en in e es media es lib a y access (a
empo al ac o o sus ained esou ce a ailabili y) and s uden
engagemen — hough his media ion has diminished in mode n
imes, as p og ams now p io i ize academic ou comes o e social
jus ice hemes. This his o ical shi highligh s how con en in e es ’s
media ing ole depends on alignmen be ween empo al ac o s (e.g.,
lib a y p og amming) and eade s’ in e es s in con ex ually ele an
sociological hemes.
Exis ing s udies on sociological book bo owing beha io s in
lib a ies ha e no able limi a ions. Fi s , hey o e ly ely on single-
dimensional analyses (e.g., desc ip i e s a is ics, simple
co ela ions), ailing o explo e he empo al dynamics o
bo owings (e.g., he pe sis ence o long- e m eading in e es s) o
he mechanis ic ela ionships be ween in luencing ac o s (e.g., he
indi ec e ec s o bo owing du a ions ia speci ic book ca ego ies).
Resea ch Pape
This wo k is licensed unde a C ea i e Commons A ibu ion-NonComme cial 4.0 In e na ional License.
DOI: 10.5281/zenodo.17322626
Page 26
Second, hey inadequa ely in eg a e me ada a and hema ic
analyses— a ely using s anda dized me ada a (e.g.,
CNMARC/UNIMARC 6XX subjec e m ields) and ins ead elying
on subjec i e i le classi ica ions o ee- ex keywo ds, which lead
o inconsis en subjec delinea ions and he inabili y o cap u e ine-
g ained hema ic clus e s o c oss-pe iod in e es ends. Thi d, hey
pay limi ed a en ion o media ion mechanisms and in e p e able
enewal p edic ions; ew s udies examine indi ec a iable pa hways,
and enewal p edic ions o en use opaque ―black-box‖ algo i hms o
exclude ime- a ying ac o s, educing gene alizabili y. Fou h, hey
lack con ex ualiza ion wi h s anda dized classi ica ion sys ems (e.g.,
he Chinese Lib a y Classi ica ion, CLC), ailing o cla i y how sub-
ca ego ies (e.g., C9) di e in in luencing bo owings o why speci ic
sub-ca ego ies domina e bo owing olumes.
The p ima y pu pose o his s udy is o sys ema ically explo e he
cha ac e is ics, dynamics, and in luencing mechanisms o
sociological book bo owing beha io a Nanjing No mal
Uni e si y, wi h ou speci ic objec i es:
Fi s , o iden i y he hema ic cha ac e is ics o sociological books
a o ed by eade s, and analyze he empo al e olu ion o high-
equency subjec e ms ac oss h ee pe iods (2016–2018, 2019–
2021, 2022–2024)—cla i ying bo h he s able co e hemes and
phased eme ging opics in eade s’ disciplina y in e es s.
Second, o unpack he key ac o s d i ing eade s’ o al sociological
book bo owings, and e i y he media ing ole o C9 (CLC
subca ego y ―Sociology and Demog aphy‖) in he ela ionship
be ween ―Leng h o Bo owing Yea s‖ and ―To al Bo owings‖,
e ealing he di ec and indi ec mechanisms shaping bo owing
ou comes.
Thi d, o de elop an in e p e able p edic ion model o eade s’ book
enewal beha io using he Classi ica ion and Reg ession T ees
(CRT) algo i hm, and alida e he model’s pe o mance ac oss
di e en ime pe iods—assessing how ac o s like ―To al
Bo owings,‖ ―Gende ,‖ and ―Pe sis ence‖ in luence enewal
decisions and he isks o excluding ime- a ying p edic o s.
Fou h, o ansla e he abo e analy ical indings in o ac ionable
insigh s o academic lib a ies and highe educa ion ins i u ions—
p o iding e idence-based guidance o op imizing sociological book
collec ion alloca ion, e ining eade se ices, and aligning
eaching/lea ning esou ces wi h eade s’ e ol ing disciplina y
needs.
3. Da a and Me hod
The esea ch sample consis s o he ci cula ion da a o ca ego y C
(Sociology) books om Nanjing No mal Uni e si y be ween 2016
and 2024, co e ing 13,344 eade s, 13,117 ypes o books, and a
o al loan olume o 59,198.
3.1 Book Theme Analysis
Field 6XX cons i u es he co e ield g oup o desc ibing he subjec
con en o documen s in bo h CNMARC (China Machine-Readable
Ca aloging Fo ma ) and UNIMARC (Uni e sal Machine-Readable
Ca aloging Fo ma ). I s p ima y unc ion is o se e as a "subjec
access poin ," enabling h ee key capabili ies: (1) accu a ely
e ealing he documen ’s subjec ma e wi h g ea e g anula i y
han classi ica ion ields; (2) suppo ing mul i-dimensional e ie al
(encompassing disciplina y subjec s, pe sonal names, geog aphical
loca ions, ins i u ions, and documen i les); and (3) acili a ing
c oss-classi ica ion associa ion (agg ega ing documen s on he same
subjec ha may all unde di e en classi ica ion numbe s). Use s
can sea ch h ough lib a y ca alog sys ems ia "subjec e ms".
Publishe s and lib a ies use Field 6XX o uni y subjec desc ip ions
du ing ca aloging, a oiding me ada a inconsis encies caused by ee-
ex ocabula y.
The isualiza ion is gene a ed ia Voyan Tools, a web-based
pla o m dedica ed o digi al ex analysis. I s de elopmen elies on
a se ies o in e ela ed analy ical p ocedu es. Fi s , he subjec e ms
o books in Ca ego y C bo owed by eade s a e ansla ed in o
English and inpu in o he ex box. Ini ially, he inpu digi al ex s
unde go p ep ocessing, which in ol es okeniza ion ( o segmen he
ex in o indi idual wo ds) and wo d equency coun ing.
Subsequen ly, seman ic ec o iza ion echniques—such as wo d
embedding o opic modeling—a e applied o con e hese wo ds
in o high-dimensional seman ic ec o s. These ec o s cap u e he
seman ic ela ionships be ween wo ds, wi h wo ds o simila
meanings posi ioned in close p oximi y wi hin he ec o space.
Th ough ex analysis, he op 10 subjec e m ca ego ies a e mined,
and he op 100 high- equency subjec e ms as well as subjec e m
clus e s (including he op 50 high- equency e ms) ac oss h ee
ime pe iods a e isually p esen ed.
Figu e 1. Resea ch design
Resea ch Pape
This wo k is licensed unde a C ea i e Commons A ibu ion-NonComme cial 4.0 In e na ional License.
DOI: 10.5281/zenodo.17322626
Page 27
Figu e 1 illus a es he esea ch design cen e ed on book bo owing
da a om 2016 o 2024. I i s del es in o he empo al pe sis ence
o bo owing beha io , u ilizing a bina y indica o (assigning 1 o
indi idual book bo owing and 0 o he wise) and examining he
"Leng h o he Bo owing Yea s". Then, i analyzes he in e es
endency o bo owing beha io ia a line cha . Finally, o
p edic ing s uden s' sociological book bo owing beha io , wo
app oaches a e employed: a linea model wi h "To al Bo owings"
as he dependen a iable and "Leng h o he Bo owing Yea s" and
"C9" as independen a iables; and a media ion e ec analysis
whe e "C9" media es he ela ionship be ween "Leng h o he
Bo owing Yea s" and "To al Bo owings".
Compa ing hese wo algo i hms, he linea model o e s a
s aigh o wa d way o assess he di ec impac s o bo owing
du a ion and "C9" on o al bo owings, being simple and
in e p e able. In con as , he media ion e ec analysis unco e s he
indi ec pa hway h ough which "C9" ansmi s he in luence o
bo owing yea s on o al bo owings, p o iding a mo e nuanced
unde s anding. The signi icance lies in ha he linea model helps
iden i y key di ec p edic o s, while he media ion analysis en iches
he comp ehension o he unde lying mechanisms, oge he o e ing
a comp ehensi e iew o ac o s shaping s uden s' book - bo owing
beha io and aiding in mo e a ge ed in e en ions o esou ce
alloca ion in educa ional se ings.
Ca ego y C in he Chinese Lib a y Classi ica ion (CLC), i led
"Gene al Wo ks o Social Sciences", is a classi ica ion summa y o
basic and uni e sal heo ies and me hods in he ield o social
sciences. I s seconda y ca ego ies ollow a logical sys em o " om
basic heo y o p ac ical applica ion, and om he whole o he pa ".
The speci ic ca ego ies and hei co e con en s a e as ollows:
The seconda y ca ego y wi h he classi ica ion code C0 is "Theo ies
and Me hodologies o Social Sciences", which co e s he basic
heo ies o social sciences (such as he objec s, na u e, and sys em o
social sciences), esea ch me hods (such as su ey me hods and
s a is ical me hods), and he his o y o academic hough and i s
schools.
The seconda y ca ego y C1 is "Cu en Si ua ion and O e iew o
Social Sciences", encompassing he o e all de elopmen s a us o
social sciences in a ious coun ies/ egions, esea ch ends,
academic ins i u ions (e.g., academies o social sciences, uni e si y
esea ch ins i u es), academic o ganiza ions, academic jou nals, and
li e a u e e iews.
The seconda y ca ego y C2 is "Ins i u ions and O ganiza ions o
Social Sciences", which includes he o ganiza ional s uc u e,
unc ions, ac i i ies, and di ec o ies o specialized ins i u ions in he
ield o social sciences (e.g., social science esea ch ins i u es, policy
esea ch cen e s) and academic o ganiza ions (e.g., sociology
socie ies, economics associa ions).
The seconda y ca ego y C3 is "Resea ch Me hods o Social
Sciences", ocusing on speci ic me hods and ools o social science
esea ch, including quan i a i e esea ch (e.g., da a analysis, model
cons uc ion), quali a i e esea ch (e.g., in e iews, case s udies),
in e disciplina y esea ch me hods, and esea ch design.
The seconda y ca ego y C4 is "Educa ion and Popula iza ion o
Social Sciences", co e ing he educa ion sys em in he ield o social
sciences (e.g., he se up o social science majo s in uni e si ies,
cu iculum and eaching ma e ials), popula science ac i i ies (e.g.,
he p omo ion o social science knowledge, public educa ion
p og ams), and ela ed esou ces.
The seconda y ca ego y C5 is "Se ies, Collec ed Wo ks, and Se ial
Publica ions o Social Sciences", which includes collec ed li e a u e
in he ield o social sciences, such as se ies (e.g., "F on ie Se ies o
Social Sciences"), collec ed pape s, yea books (e.g., Yea book o
Chinese Social Sciences), and se ial publica ions like jou nals.
The seconda y ca ego y C6 is "Re e ence Books o Social Sciences",
e e ing o e e ence ools in he ield o social sciences, such as
dic iona ies (e.g., Dic iona y o Social Sciences), encyclopedias
(e.g., Encyclopedia o China: Social Sciences Volume), handbooks,
indexes, and ca alogs o e ie al and e e ence pu poses.
The seconda y ca ego y C7 is "In o ma ion Re ie al Tools o Social
Sciences", speci ically ools used o e ie ing social science
li e a u e, including guides o li e a u e da abases, e ie al
s a egies, ci a ion indexes (e.g., ools ela ed o he Social Sciences
Ci a ion Index - SSCI), and handbooks on li e a u e classi ica ion
and indexing.
The seconda y ca ego y C8 is "S a is ics", which in ol es s a is ical
heo ies and me hods applicable in he ield o social sciences,
including desc ip i e s a is ics, in e en ial s a is ics, eg ession
analysis, sampling su eys, he applica ion o s a is ical so wa e
(e.g., SPSS, S a a), and specialized s a is ical echniques in a ious
social science disciplines (e.g., economics, sociology).
The seconda y ca ego y C9 is "Sociology and Demog aphy".
Sociology he e co e s a eas such as social s uc u e, social
s a i ica ion, social mobili y, social o ganiza ions, and social issues;
demog aphy includes popula ion size, s uc u e, dis ibu ion, e ili y
and aging, and popula ion policies, as well as ela ed sub- ields (e.g.,
medical sociology, u al sociology). I can be seen om Figu e 1
ha he bo owing olume o C9 books is signi ican ly highe han
ha o books in o he seconda y ca ego ies.
3.2 Linea Reg ession Analysis
In he op menu ba o SPSS, selec Analyze → Reg ession →
Linea . This opens he ―Linea Reg ession‖ main dialog box, whe e
a iables and model pa ame e s a e con igu ed. "To al Bo owings"
is he dependen a iable being p edic ed. "Leng h o he Bo owing
Yea s" and "C9" a e he dependen a iables. "Fo wa d S epwise"
me hod is used. This me hod s a s wi h no p edic o s and i e a i ely
adds he mos s a is ically signi ican one a each s ep un il no mo e
signi ican p edic o s emain, aiming o build an op imal model
e icien ly.
The ―Model Summa y‖ able se es as he co e ou pu o his
o wa d s epwise linea eg ession analysis, and i s indica o s mus
be sys ema ically in e p e ed o e alua e he alidi y, explana o y
powe , and p edic i e accu acy o he op imal model gene a ed.
Fi s , R ( he mul iple co ela ion coe icien ) quan i ies he s eng h
o he linea ela ionship be ween he p edic ed alues o he
dependen a iable ―To al Bo owings (TB)‖ and i s obse ed
alues; a alue close o 1 indica es a s onge consis ency be ween
he model’s p edic ions and ac ual da a, e lec ing a mo e eliable
p edic i e end.
Second, R Squa e (R²) ep esen s he p opo ion o a iance in TB
ha can be explained by he independen a iables (IVs)—―Leng h
o he Bo owing Yea s (LBY)‖ and/o ―C9‖—included in he inal
model. Fo ins ance, an R² alue o 0.35 means 35% o he
a iabili y in TB is accoun ed o by he IVs, while he emaining
65% is a ibu ed o unmeasu ed ac o s o andom a ia ion; a
highe R² gene ally indica es ha he model cap u es mo e o he
key d i e s o TB.
Thi d, Adjus ed R Squa e is a co ec ed e sion o R² ha adjus s o
Resea ch Pape
This wo k is licensed unde a C ea i e Commons A ibu ion-NonComme cial 4.0 In e na ional License.
DOI: 10.5281/zenodo.17322626
Page 28
he numbe o IVs in he model. Unlike R², which inhe en ly
inc eases (o emains unchanged) when addi ional a iables a e
added—e en i hose a iables a e non-signi ican and do no
meaning ully imp o e model i —Adjus ed R Squa e penalizes he
inclusion o edundan p edic o s. This makes i a mo e conse a i e
and accu a e measu e o model pe o mance, especially when
compa ing models wi h di e en numbe s o IVs, as i a oids
o e es ima ing he model’s explana o y powe .
Fou h, R Squa e Change e lec s he inc emen al inc ease in R²
when a new IV is added o he model du ing each s ep o he
o wa d selec ion p ocess. A s a is ically signi ican R Squa e
Change ( ypically de e mined by a p- alue < 0.05) a a speci ic s ep
indica es ha he newly added IV con ibu es unique, meaning ul
explana o y powe o he model. Fo example, a signi ican R Squa e
Change a S ep 1 would mean he i s IV (ei he LBY o C9,
depending on s a is ical signi icance) explains a subs an ial amoun
o addi ional a iance in TB ha was no cap u ed by he in e cep -
only model.
Finally, S anda d E o o he Es ima e (SEE) measu es he a e age
dis ance be ween he obse ed TB alues and he alues p edic ed
by he model. I quan i ies he magni ude o p edic ion e o : a
smalle SEE indica es ha he model’s p edic ions a e mo e closely
aligned wi h ac ual TB alues, signi ying highe p edic i e accu acy.
Con e sely, a la ge SEE sugges s g ea e a iabili y be ween
p edic ed and obse ed alues, indica ing po en ial limi a ions in he
model’s abili y o o ecas TB eliably.
3.3 Media ion E ec Analysis
In he P ocess dialog box, speci y "Leng h o Bo owing Yea s
(LBY)" as he independen a iable (X), "Media o Va iable C9" as
he media o a iable (M), and "To al Bo owings (TB)" as he
dependen a iable (Y). Use he Boo s ap me hod o es ima e he
indi ec e ec because i does no ely on he no mal dis ibu ion
assump ion. In he "Boo s ap Samples" op ion, se he numbe o
esampling imes 5000. In he "Con idence Le el" op ion, se he
con idence in e al, usually 95%. A e comple ing he abo e
se ings, click "OK" o un he analysis. SPSS will ou pu a se ies o
esul s, including pa h coe icien s, signi icance le els, R-squa ed
alues, and Boo s ap con idence in e als.
E alua e he o e all i o he model h ough he R-squa ed alue.
The R-squa ed alue ep esen s he p opo ion o he a iance o he
dependen a iable explained by he independen a iable and he
media o a iable. A la ge R-squa ed alue indica es a be e i o
he model. Check he signi icance le els (p- alues) o he pa h
coe icien s o X→M, M→Y, and X→Y. I he p- alue is less han
0.05, i indica es ha he co esponding pa h is signi ican , ha is,
he e is a signi ican linea ela ionship be ween he a iables. Focus
on he poin es ima e and 95% con idence in e al o he indi ec
e ec . I he 95% con idence in e al does no include 0, i indica es
ha he media ion e ec is signi ican , ha is, he independen
a iable X has an impac on he dependen a iable Y h ough he
media o a iable M.
3.4 Decision T ee Analysis
The dependen a iable is "Renewal", and he independen a iables
include "To al", "Gende ", and "Pe sis ence". C oss - alida ion is
used, wi h he maximum ee dep h se o 5, he minimum numbe o
cases in he pa en node being 100, and he minimum numbe o
cases in he child node being 50. SPSS uses he "CRT"
(Classi ica ion and Reg ession T ees) algo i hm, which is sui able
o bo h ca ego ical and con inuous dependen a iables. The
decision ee model gene a es co e analy ical da a ac oss model
pe o mance, node s uc u e, and a iable impo ance.
The analy ical da a gene a ed by he a o emen ioned decision ee
model collec i ely add ess h ee c i ical esea ch ques ions ha
unde pin he model’s alidi y and p ac ical u ili y o "Renewal"
p edic ion. Fi s , he model’s pe o mance me ics—including
o e all accu acy, c oss- alida ion accu acy, and class-speci ic
indica o s such as sensi i i y and p ecision—p o ide empi ical
e idence o answe whe he he model can eliably p edic
"Renewal" ou comes and gene alize o unseen da a. Second, he
s uc u al da a o he ee’s nodes (e.g., oo node benchma ks, spli
ules o in e media e nodes, and "Renewal" a e dis ibu ions in
e minal nodes) elucida e he sequen ial and condi ional mechanisms
h ough which independen a iables (i.e., "To al", "Gende ", and
"Pe sis ence") d i e "Renewal" beha io , ende ing he model’s
p edic i e logic in e p e able a he han opaque. Thi d, he ela i e
impo ance sco es assigned o each independen a iable cla i y
which ac o s exe he mos subs an ial in luence on "Renewal"
(e.g., "To al" ypically eme ging as he p ima y d i e , ollowed by
"Pe sis ence"), he eby p io i izing a iables o ac ionable analysis.
Toge he , hese a ibu es ele a e he model beyond a me e
p edic i e ool: by esol ing he h ee co e ques ions, i se es as a
obus guide o da a-d i en decision-making in p ac ical
scena ios— o ins ance, enabling he design o a ge ed ma ke ing
s a egies ha ocus on use s wi h high "To al" alues and s ong
"Pe sis ence", who a e empi ically mo e likely o exhibi "Renewal"
beha io .
4. Resul
4.1 Top 10 Topics
4.1.1Sociology Co e Disciplines
As he ounda ional b anches o sociology, hese co e disciplines
cons uc he heo e ical amewo k and esea ch pa adigm o he
en i e discipline, and p o ide essen ial me hodological and
analy ical ools o explo ing social phenomena: Ins i u ional
sociology, sociological de ec ion p ocedu es esea ch, Ame ican
social s udies, u al s udies, o ganiza ional di e ences, e c.
4.1.2 Li e a u e, A & Philosophy
This sub- opic in eg a es h ee in e ela ed ields—li e a u e, a , and
philosophy—explo ing he exp ession o human hough s, emo ions,
and alues h ough li e a y wo ks, a is ic p ac ices, and
philosophical e lec ions. I co e s bo h heo e ical esea ch and
p ac ical o ms, e lec ing he di e si y o human cul u al c ea ion
and spi i ual explo a ion, including philosophical analysis o no els,
wo ld li e a u e, social li e a u e, p ose, philosophy, callig aphy
binding, legal s o ies, e c.
4.1.3 Highe & Voca ional Educa ion
This heme ocuses on he de elopmen , e o m, and p ac ical
applica ion o highe educa ion and oca ional educa ion sys ems,
aiming o b idge he gap be ween educa ion and social needs, and
cul i a e alen s wi h p o essional compe ence and comp ehensi e
quali y. I co e s bo h heo e ical explo a ion o educa ional models
and p ac ical esea ch on eaching me hods, including highe
educa ion in an h opology, oca ional s udies, mode n educa ion,
seconda y educa ion, sec e a ial s udies, e c.
4.1.4 Academic Exchange
This heme emphasizes he impo ance o academic communica ion
and knowledge sha ing in p omo ing academic p og ess, ocusing on
he o ms, channels, and impac s o academic exchange, as well as
he dissemina ion and inno a ion o academic achie emen s. I
co e s bo h adi ional and mode n academic exchange me hods,
e lec ing he global and in e disciplina y ends o cu en academic
Resea ch Pape
This wo k is licensed unde a C ea i e Commons A ibu ion-NonComme cial 4.0 In e na ional License.
DOI: 10.5281/zenodo.17322626
Page 29
esea ch, including academic con e ences, academic wo ks,
con empo a y hough , pos mode nis s udies, applica ion
echnologies, e c.
4.1.5 Specialized Disciplines
This heme co e s a a ie y o specialized disciplines ha play
impo an oles in social de elopmen and academic esea ch, each
wi h i s unique heo e ical sys em and esea ch me hods, p o iding
p o essional suppo o sol ing speci ic social p oblems and
p omo ing in e disciplina y de elopmen , including economic
geog aphy, popula ion heo y, managemen science, de elopmen al
design, de elopmen policy, ins i u ional ime managemen , e c.
4.1.6 Cul u al & E hnic Resea ch
This heme ocuses on he explo a ion o cul u al di e si y and
e hnic cha ac e is ics, s udying he o ma ion, e olu ion, and mu ual
in luence o di e en cul u es and e hnic g oups, aiming o p omo e
cul u al he i age, e hnic ha mony, and c oss-cul u al unde s anding.
I co e s bo h heo e ical esea ch on cul u e and e hnici y and
p ac ical esea ch on cul u al p o ec ion, including Chinese cul u al
s udies, e hnic s udies, cul u e, ancien issues, na ional s udies, e c.
4.1.7 Na ional & Regional S udies
This heme ocuses on he s udy o di e en coun ies and egions
a ound he wo ld, explo ing hei social, economic, cul u al, and
poli ical cha ac e is ics, as well as he di e ences and connec ions
be ween hem. I aims o p o ide a comp ehensi e unde s anding o
global egional de elopmen pa e ns and p omo e in e na ional
coope a ion and egional s abili y, including China, Uni ed S a es,
Wes e n coun ies, Japan, Sou h Ko ea, p o ince, u ban, sou he n
egions, e c.
4.1.8 Resea ch Me hods & S a is ical Tools
This heme p o ides essen ial me hodological suppo o academic
esea ch, ocusing on he de elopmen , applica ion, and inno a ion
o esea ch me hods and s a is ical ools, aiming o imp o e he
scien i ic, accu acy, and eliabili y o esea ch esul s. I co e s bo h
quali a i e and quan i a i e esea ch me hods, as well as he
p ac ical applica ion o s a is ical so wa e, including su ey
s a is ics, esea ch s a is ics, sampling and o de con ol, SPSS,
consume ques ionnai es, e alua ion s udies, so wa e, e c.
4.1.9 Psychology & Social Secu i y
This heme combines he s udy o psychology and social secu i y,
explo ing he psychological ac o s a ec ing indi idual well-being
and social s abili y, as well as he ole o social secu i y sys ems in
p o ec ing indi idual igh s and p omo ing social equi y. I co e s
bo h heo e ical esea ch and p ac ical policy analysis, including
poli ical psychology, esea ch on ela ionships, public, his o ical
ne wo ks, cons uc ion s udies, campus, secu i y, isk, lie, wel a e
his o y, e c.
4.1.10 Sexual & Gende S udies
This heme ocuses on he explo a ion o sexual beha io , sexual
cul u e, gende iden i y, and gende equali y, challenging adi ional
gende no ms and p omo ing he unde s anding and espec o
di e se gende and sexual o ien a ions. I co e s bo h heo e ical
esea ch on gende and sexuali y and p ac ical ad ocacy o gende
equi y, e lec ing he p og ess o social concep s and he pu sui o
human igh s, including sexual s udies, in oduc ion o sexology,
homosexuali y s udies, women’s changes, e c.
4.2 Top 100 Subjec Wo ds
Among he op 100 high- equency subjec wo ds in he h ee ime
pe iods o 2016-2018, 2019-2021, and 2022-2024 (see Figu es 2, 3,
and 4), he e a e 86 wo ds ha appea in all h ee ime pe iods
(accoun ing o 86%). Among hese 86 wo ds:
The e a e 27 wo ds wi h a o al equency o ≥ 1000, which a e:
esea ch, social, china, sociology, educa ion, highe , ins i u ions,
managemen , science, s udies, s a is ical, analysis, psychology,
me hods, his o y, mode n, an h opology, o ganiza ional, public,
ela ions, wo k, s a is ics, na ional, u ban, leade ship, applica ion,
sociological.
The e a e 59 wo ds wi h a o al equency o ＜ 1000, which a e:
wo ld, human, cul u al, beha io , heo y, popula ion, su ey,
so wa e, issues, cul u e, coun ies, hough , ela ionship, amily,
in e pe sonal, wes e n, philosophy, u al, li e a u e, women's, sexual,
in oduc ion, g oup, wo ks, knowledge, p ocedu es, li e, Chinese,
gende , de elopmen , impac , psychological, sexology, ma iage,
choice, basic, ca ee , policy, opinion, ela ionships, Uni ed S a es,
in e ne , humani ies, lo e, decision, ime, Japan, SPSS, li e a y,
alen , ancien , communi y, models, p o ince, a , me hodology,
an hologies, F ance, na ion, beha io al, sys ems, Qing.
Figu e 2. Top 100 subjec wo ds (2016-2018)
Resea ch Pape
This wo k is licensed unde a C ea i e Commons A ibu ion-NonComme cial 4.0 In e na ional License.
DOI: 10.5281/zenodo.17322626
Page 30
Figu e 3. Top 100 subjec wo ds (2019-2021)
Figu e 4. Top 100 subjec wo ds (2022-2024)
The op 100 subjec high- equency wo ds ac oss he h ee ime
pe iods (2016–2018, 2019–2021, and 2022–2024) e eals a s ong
concen a ion o c oss-pe iod co e hemes in sociology books. This
high deg ee o o e lap indica es ha eade s’ a en ion o co e
sociological hemes has emained s able o e nea ly a decade,
p ima ily co e ing esea ch me hods and pa adigms, social s uc u e
and ins i u ional analysis, educa ion and social de elopmen ,
in e disciplina y connec ions, cul u al and beha io al s udies, social
g oups and iden i y, and global and egional con ex s.
Among he high- equency wo ds ac oss he h ee ime pe iods o
2016-2018, 2019-2021, and 2022-2024, he e a e 31 wo ds ha
appea in only one pe iod. Single-pe iod unique wo ds e eal a shi
om "p ac ical specializa ion" (2016–2018) and "educa ional ex
ocus" (2019–2021) o " echnology in eg a ion, social issues, and
opinion dynamics" (2022–2024). This end e lec s he adap abili y
o sociology o social changes and he expanding scope o eade
in e es in esponse o con empo a y challenges:
11 wo ds appea ing only in 2016-2018 (11%): These wo ds a e
unique o he 2016-2018 pe iod and did no appea in he subsequen
wo pe iods. They mainly consis o specialized e minology in
speci ic ields and low- equency wo ds, co e ing a eas such as
sec e a ial wo k, leisu e, and consul ing, such as "sec e a ial" (82
occu ences), "leisu e" (39 occu ences), and "consul ing" (42
occu ences). They also include gene al domain wo ds like
"in o ma ion" (47 occu ences) and "cases" (47 occu ences).
5 wo ds appea ing only in 2019-2021 (5%): The unique wo ds in
his pe iod ocus on educa ional ins i u ions and ex ypes, wi h co e
educa ion- ela ed ocabula y such as "uni e si ies" (110
occu ences) and "colleges" (110 occu ences). The e a e also ex -
ela ed wo ds like "essays" (38 occu ences), cha ac e - ela ed wo ds
like "cha ac e " (61 occu ences), and change- ela ed wo ds like
"changes" (28 occu ences). O e all, hey e lec he pe iod's phased
ocus on educa ional scena ios and ex analysis.
15 wo ds appea ing only in 2022-2024 (15%): The unique wo ds
in his pe iod a e he mos nume ous and co e a wide ange o
ields, including eme ging a eas such as echnology (" echnology",
19 occu ences), social issues ("aging", 19 occu ences; " eminis ",
18 occu ences), and indi idual- ela ed e ms ("indi idual", 23
occu ences; "you h", 18 occu ences). They also include newly
eme ging ocuses like "ne wo k" (18 occu ences), "e hics‖ (16
occu ences), and opinions ("c i icism", 19 occu ences;
"commen s", 22 occu ences). O e all, hey e lec he addi ion and
ex ension o ecen esea ch hemes.
4.3 Theme Clus e
Resea ch Pape
This wo k is licensed unde a C ea i e Commons A ibu ion-NonComme cial 4.0 In e na ional License.
DOI: 10.5281/zenodo.17322626
Page 31
Figu es 5, 6, and 7 iden i y he hema ic wo d clus e s and he op 50
high- equency wo ds. The cyan - colo ed "Bin" in he igu es
ep esen s clus e ing. In he con ex o sociological book opic
analysis, hese bins g oup oge he e ms ha ha e s ong hema ic
connec ions. In di e en ime pe iods (2016 - 2018, 2019 - 2021,
2022 - 2024), he cyan bins agg ega e sociological concep s and
opics ha sha e simila esea ch ocuses o seman ic ela ionships,
helping o isualize he clus e ing s uc u e and e olu ion o
sociological esea ch hemes o e ime.
Figu e 5. Theme clus e s (2016-2018)
2016 – 2018 co e hemes: Topics clus e a ound in e pe sonal
ela ionships, amily, cul u al an h opology, and social s uc u e
(e idenced by e ms like "in e pe sonal", " amily", "cul u al
an h opology", "social s uc u e"). The e's also a ocus on gende
s udies (wi h "women's" p esen ).
Scope: Themes a e ela i ely adi ional, co e ing ounda ional
sociological a eas like in e pe sonal dynamics and cul u al
explo a ion wi hin a socie al con ex .
Figu e 6. Theme clus e s (2019-2021)
2019 – 2021 shi in ocus: The clus e cen e s shi owa d mode n
social issues (e.g., "mode n", "u baniza ion", "o ganiza ions") and
p ac ical sociology applica ions ( e ms like "p ocedu es",
"communica ion", "public policy").
New emphases: Eme gence o opics ela ed o u ban de elopmen ,
o ganiza ional beha io , and sociological me hods in eal - wo ld
se ings, e lec ing a mo e owa d applied sociology esponding o
con empo a y socie al changes.
Resea ch Pape
This wo k is licensed unde a C ea i e Commons A ibu ion-NonComme cial 4.0 In e na ional License.
DOI: 10.5281/zenodo.17322626
Page 38
in Koso o cu icula (add essing eal-wo ld cul u al cohesion issues)
u he embodies his phase’s shi o p ac icali y.
The 2022–2024 phase ea u es h ee-dimensional in eg a ion:
s eng hening disciplina y heo e ical dep h, highligh ing
in e disciplina y in e sec ions ( ia high- equency e ms linked o
psychology, philosophy, and s a is ics), and ocusing on
global/mac o-social challenges ( ia e ms ela ed o na ional
dynamics and ecology). This in eg a ion is e lec ed in di e se
eading p ac ices: Feng e al. (2025)’s s udy o diabe ic/hype ensi e
pa ien s (2022–2024) ound ha sociological books on heal h equi y
(in eg a ing sociology wi h heal hca e, an in e disciplina y ocus)
educed anxie y, wi h con en in e es media ing sus ained eading—
aligning wi h he phase’s in e disciplina y emphasis. Abimbola
(2025)’s ocus on clima e change eading (a global mac o-challenge)
also i s his phase, as guided eading on en i onmen al
sus ainabili y (a sociological heme) enhanced li e acy ia con en
in e es , demons a ing he in eg a ion o mac o-issues and
heo e ical dep h.
Sociology’s esea ch ocus ollows a dis inc e olu iona y
ajec o y— om adi ional ounda ion o p ac ical eali y, and
u he o h ee-dimensional in eg a ion o heo y, in e disciplina i y,
and mac o-issues—d i en by social changes and e lec ing he
discipline’s adap abili y. This e olu ion is no andom bu ep esen s
ac i e esponse o s age-speci ic social needs: 2016–2018’s
ounda ional ocus buil disciplina y consensus, 2019–2021’s
p ac ical shi add essed u gen socie al p oblems, and 2022–2024’s
in eg a ion esponded o complex global challenges. This ajec o y
is mi o ed in long- e m eading ou comes: Lewis e al. (2023) ound
ha Eu opean adul s who accessed sociological books in childhood
(2016–2018’s ounda ional phase) de eloped ea ly con en in e es ,
which la e media ed educed cogni i e impai men isk—showing
how ea ly ounda ional ocus laid he g oundwo k o la e
engagemen wi h p ac ical and in eg a ed hemes.
Sociological heme clus e ing embodies he discipline’s
de elopmen al logic: expanding om simpli ied o di e si ied
esea ch scopes, and upg ading om s a ic social desc ip ion o
dynamic pa icipa ion. The shi om single disciplina y ields o
in e disciplina y collabo a ion (e.g., sociology + heal hca e in 2022–
2024) e lec s bounda y expansion, while he mo e om passi e
phenomenon desc ip ion o ac i e p oblem-sol ing (e.g., 2019–
2021’s educa ional in e en ions) ep esen s unc ional upg ading.
This logic is e iden in communi y eading sys ems: book sys ems
e ol ed om p o iding basic sociological ex s (2016–2018) o
p ac ical ma e ials (2019–2021) and inally o in e disciplina y,
mac o- hemed books (2022–2024), wi h con en in e es media ing
sus ained use—con i ming ha he discipline’s de elopmen al logic
di ec ly shapes eading esou ce p o ision and engagemen .
Ul ima ely, his e olu ion posi ions sociology as a p agma ic
discipline aligned wi h eal-wo ld needs, balancing heo e ical igo
wi h p ac ical impac .
5.4 Compa ison o Di e ences Be ween Linea Reg ession
Model and Media ion E ec Model
5.4.1Di e ences in Model Posi ioning and Resea ch
Objec i es
Bo h models ake "To al Bo owings (TB)" as he dependen
a iable. The linea eg ession model is an associa ion desc ip i e
model. I s co e objec i e is only o quan i y he s eng h o linea
associa ion be ween each p edic o a iable and o al bo owings,
answe ing ques ions such as "which a iables a ec o al
bo owings" and " o wha ex en do hey a ec i ", wi hou in ol ing
he in luence mechanism be ween a iables. Fo example, i can
only de e mine ha Leng h o Bo owing Yea s (LBY) and C9 bo h
a ec o al bo owings bu canno explain whe he LBY a ec s o al
bo owings di ec ly o indi ec ly h ough o he a iables.
The media ion e ec model is a mechanism-explana o y model. I s
co e objec i e ocuses on "how LBY a ec s o al bo owings". By
in oducing he media ing a iable C9, i cla i ies he ansmission
pa h o "LBY→C9→To al Bo owings", decomposes he di ec
e ec (LBY a ec s o al bo owings wi hou going h ough C9) and
indi ec e ec (LBY a ec s o al bo owings indi ec ly h ough C9),
and inally answe s "wha is he speci ic mechanism by which LBY
a ec s o al bo owings", a he han s aying a he supe icial
judgmen o "whe he an impac exis s".
5.4.2Di e ences in Va iable Composi ion and Rela ionship
Assump ions
The a iable composi ion o he linea eg ession model is b oade ,
including mul iple p edic o a iables. Howe e , i only analyzes he
di ec associa ion be ween each p edic o a iable and o al
bo owings and does no de ine he causal o de be ween a iables.
I s ela ionship assump ions a e also ela i ely simple, only equi ing
a linea ela ionship be ween p edic o a iables and o al
bo owings, wi hou in ol ing assump ions abou "whe he he e is
in e media e ansmission be ween a iables". Fo example, i does
no p ese ha "LBY a ec s C9 i s and hen a ec s o al
bo owings"; ins ead, i ea s C9 as a p edic o a iable pa allel o
LBY.
The a iable composi ion o he media ion e ec model has s ic
de ini ions, which mus include h ee ypes o a iables: independen
a iable (LBY, X), media ing a iable (C9, M), and dependen
a iable (To al Bo owings, Y). Mo eo e , i clea ly se s he causal
o de as "X→M→Y". I s ela ionship assump ions a e mo e
complex: in addi ion o sa is ying he linea ela ionship, i also
needs o assume ha LBY can signi ican ly a ec C9 and C9 can
signi ican ly a ec o al bo owings. A he same ime, i de aul s o
no e e se causali y (e.g., C9 does no a ec LBY in u n). Th ough
a clea causal chain, i ocuses on ansmission mechanism a he
han pa allel associa ion.
5.4.3 Di e ences in Model Fi ing E ec and Explana o y
Powe
The i ing e ec o he linea eg ession model is a a mode a e
le el. Acco ding o he esul s, he model p edic ion accu acy is
62.1%. Addi ionally, he e is a p oblem o non-no mal esiduals ( he
esidual his og am does no ma ch he no mal cu e, and only he
mean o 0.000 and s anda d de ia ion o 1.077 mee some
assump ions). Combining wi h he o al e ec model R²=0.3091
con aining only LBY in he media ion e ec model, i can be
in e ed ha e en i he linea eg ession model includes C9, i s
o e all explana o y powe is s ill a lowe han ha o he media ion
e ec model.
The i ing e ec o he media ion e ec model is a an ex emely
high le el, and he goodness o i is e i ied om mul iple
dimensions. Fi s ly, he eg ession model o "LBY→C9" has an R²
o 0.2871 (LBY can explain 28.71% o he a ia ion in C9), wi h an
F- alue o 5,373.1762 (p<0.001), indica ing ha he model is highly
signi ican o e all. Secondly, he eg ession model o "LBY +
C9→To al Bo owings" has an R² o 0.8923 ( he wo a iables
oge he explain 89.23% o he a ia ion in o al bo owings), wi h
an F- alue o 55,254.5725 (p<0.001), and i s explana o y powe is
a highe han ha o he linea eg ession model. Finally, al hough
esiduals a e no epo ed sepa a ely, he Boo s ap sampling (5,000
epe i ions) is used o a oid he impac o non-no mal esiduals on
Resea ch Pape
This wo k is licensed unde a C ea i e Commons A ibu ion-NonComme cial 4.0 In e na ional License.
DOI: 10.5281/zenodo.17322626
Page 39
he conclusions, making he model mo e obus .
5.4.4 Di e ences in Co e Conclusions and E ec
Quan i ica ion
The conclusions o he linea eg ession model only s ay a p edic o
impo ance anking and supe icial associa ion. I can only
de e mine ha LBY is he mos impo an p edic o a iable (wi h
he longes ba ), and C9 has lowe impo ance (wi h a sho e ba ),
bu canno quan i y he indi ec impac o LBY on o al bo owings
h ough C9. I s e ec indica o s a e also ela i ely simple: i can
only ob ain he o al associa ion coe icien o each p edic o
a iable on o al bo owings (e.g., he coe icien o LBY on o al
bo owings may be close o 6.9115), bu canno decompose he
p opo ion o di ec impac and indi ec impac , esul ing in a lack o
dep h in he conclusions.
The conclusions o he media ion e ec model ealize e ec
decomposi ion and mechanism cla i ica ion. Th ough he calcula ion
o pa h coe icien s, i quan i ies h ee ypes o key e ec s: he
indi ec e ec (LBY→C9→To al Bo owings) is 5.4816 ( he
coe icien o LBY on C9) × 1.0990 ( he coe icien o C9 on o al
bo owings) ≈ 6.0244, accoun ing o 87.16% o he o al e ec
(6.9115) and se ing as he co e impac pa h; he di ec e ec
(LBY→To al Bo owings) is 0.8871, accoun ing o 12.84% o he
o al e ec , wi h a weak bu signi ican impac ; a he same ime, i
e i ies ha " o al e ec = di ec e ec + indi ec e ec " (6.9115 =
0.8871 + 6.0244). Finally, i clea ly con i ms ha "mo e han 87%
o he impac o LBY on o al bo owings is achie ed h ough C9",
a he han he linea eg ession model’s conclusion ha LBY
di ec ly domina es he impac .
5.4.5 Di e ences in Tes ing Logic and Me hods
The es ing logic o he linea eg ession model is ela i ely simple.
I only uses he - es o de e mine he signi icance o he coe icien s
o each p edic o a iable (e.g., de e mining whe he he coe icien s
o LBY and C9 a e signi ican ly non-ze o) and uses he esidual
his og am o es whe he esiduals con o m o basic assump ions
such as no mal dis ibu ion. This ype o es can only e i y
"whe he an associa ion exis s be ween a iables" bu canno e i y
he causal pa h be ween a iables, which easily con uses associa ion
wi h causali y. Fo example, i canno p o e whe he LBY is he
cause and C9 is he e ec o C9 is he cause and LBY is he e ec .
The es ing logic o he media ion e ec model is mo e hie a chical
and igo ous. The i s s ep is o e i y he p e equisi es o he
media ion e ec h ough he s epwise es : i s , es ha he
coe icien s o "LBY→C9", "C9→To al Bo owings", and
"LBY→To al Bo owings ( o al e ec )" a e all signi ican (in he
esul s, he p- alues o he h ee a e all <0.001, mee ing he
p e equisi es). The second s ep is o di ec ly es ima e he 95%
con idence in e al o he indi ec e ec h ough he "Boo s ap es "
(5,000 epea ed samplings); i he in e al does no con ain 0, he
media ion e ec is signi ican . In he esul s, he Boo s ap
con idence in e als o "LBY→C9", "C9→To al Bo owings", and
"LBY→To al Bo owings (di ec e ec )" all do no con ain 0. This
no only a oids he de ec o low powe in he s epwise es bu also
s eng hens he causal explana o y powe o he conclusions h ough
he e i ica ion o a clea causal pa h, a he han s aying a he
associa ion le el.
5.4.6 Di e ences in P ac ical Applica ion Value
The p ac ical alue o he linea eg ession model is ela i ely
limi ed. I can only be used o p edic ing o al bo owings o
iden i ying key p edic o a iables. Fo example, based on he
conclusion ha "LBY is he mos impo an p edic o a iable", i
can be concluded ha " o a ec o al bo owings, p io i y should be
gi en o LBY", bu i canno guide "how o speci ically a ec o al
bo owings h ough LBY" o answe "whe he C9 can be egula ed
o indi ec ly a ec o al bo owings", lacking ac ionable in e en ion
di ec ions.
The p ac ical alue o he media ion e ec model is signi ican ly
imp o ed. I can no only p edic bu also p o ide a ge ed
in e en ion plans. Based on he mechanism ha LBY a ec s o al
bo owings h ough C9, wo co e in e en ion ideas can be de i ed:
i s , i o al bo owings a e o be inc eased h ough LBY, he co e
ocus is o inc ease he le el o C9 by ex ending LBY ( o each 1-
uni inc ease in LBY, C9 inc eases by 5.4816 uni s, which in u n
inc eases o al bo owings by 6.0244 uni s); second, i o al
bo owings a e o be egula ed mo e e icien ly, C9 can be di ec ly
in e ened ( o each 1-uni inc ease in C9, o al bo owings inc ease
by 1.0990 uni s) wi hou elying solely on LBY. Such conclusions
upg ade om iden i ying in luencing ac o s o p o iding solu ions,
and ha e mo e speci ic guiding signi icance o policy o mula ion
o managemen decision-making.
5.5 CRT Model Cha ac e is ics, P edic ion Pe o mance
Di e ences, and Risk Va ia ion T ends o Renewal
Beha io
5.5.1 Model Cons uc ion and Cha ac e is ics o Va iables
In luencing Renewal Beha io
The p edic ion models o enewal beha io in 2016–2018 and
2019–2021 bo h inco po a ed h ee independen a iables: "To al",
"Gende ", and "Pe sis ence", indica ing hese ac o s we e key o
shaping enewal beha io du ing hese pe iods. By 2022–2024, he
model e ained only "To al" and "Pe sis ence" while excluding
"Gende ", sugges ing gende ’s impac on enewal beha io
weakened and i no longe unc ioned as an e ec i e p edic i e
a iable. This shi aligns wi h b oade ends in sociological eading
esea ch, whe e demog aphic ac o s (like gende ) a e inc easingly
o e shadowed by con en - ela ed and en i onmen al d i e s o
engagemen . In he con ex o eading en i onmen s ha os e
enewal beha io (e.g., sus ained engagemen wi h sociological
ma e ials), a iables ied o con en ele ance (e.g., ma ching
ma e ials o eade in e es s) and pe sis ence (e.g., long- e m access
o esou ces) p o e mo e in luen ial han gende , mi o ing he CRT
model’s a iable adjus men .
The 2016–2018 model had he mos complex s uc u e (29 nodes, 15
e minal nodes, dep h o 5), while he 2019–2021 and 2022–2024
models simpli ied o 19 nodes, 10 e minal nodes, and a dep h o 4.
This simpli ica ion e lec s a g owing concen a ion in he
mechanism in luencing enewal beha io : la e pe iods equi ed
ewe a iables and s uc u al laye s o p edic engagemen , as co e
d i e s (e.g., con en in e es -media ed pe sis ence) became mo e
dis inc . This end is suppo ed by s udies on eading pe sis ence,
such as O'B ien e al. (2024)’s esea ch on bilingual home li e acy
en i onmen s (HLE), whe e "child in e es " eme ged as a key,
s eamlined ac o media ing HLE esou ces and sus ained eading—
educing he need o complex demog aphic o con ex ual a iables
o explain long- e m engagemen (a pa allel o enewal beha io ).
Addi ionally, Yu e al. (2024) ound ha au ho i a i e pa en s’
p o ision o in e es -aligned sociological books di ec ly media es
sus ained eading, u he demons a ing how ocused, con en -
d i en ac o s simpli y he p edic ion o pe sis en engagemen (and
by ex ension, enewal beha io ).
Ac oss all h ee pe iods, models used he CRT g owing me hod and
c oss- alida ion wi h uni ied pa ame e s (maximum ee dep h = 5,
minimum pa en node cases = 100, minimum child node cases = 50),
Resea ch Pape
This wo k is licensed unde a C ea i e Commons A ibu ion-NonComme cial 4.0 In e na ional License.
DOI: 10.5281/zenodo.17322626
Page 40
ensu ing compa abili y and elimina ing pa ame e in e e ence. This
me hodological consis ency is c i ical o linking model indings o
eal-wo ld eading dynamics, whe e s able, eplicable ac o s (e.g.,
access o in e es -ma ched con en ) consis en ly d i e enewal
beha io . Fo example, Webbe e al. (2025)’s s udy o school
lib a ians—who cu a e sociologically ele an books e lec ing
s uden s’ li es—shows ha consis en cu a ion p ac ices (pa allel o
he model’s uni ied pa ame e s) lead o p edic able s uden
engagemen , as con en in e es eliably media es cu a ion e o s
and sus ained in e ac ion wi h ma e ials.
5.5.2 Phased Di e ences in Renewal Beha io P edic ion
Pe o mance
The co ec p edic ion a e o "non- enewal" beha io emained
abo e 96% ac oss pe iods (96.9% in 2016–2018, 96.6% in 2019–
2021, 96.5% in 2022–2024), wi h minimal luc ua ion. This s abili y
sugges s "non- enewal" is d i en by consis en , easily iden i iable
ba ie s—o en ied o poo alignmen be ween eade needs and
eading esou ces. In sociological eading esea ch, such ba ie s a e
well-documen ed: A oyo-Vazquez e al. (2023) ound ha only
21.0% o ecommended sociological monog aphs o Spanish social
sciences deg ees a e digi al, and poo accessibili y o in e es -
aligned con en b eaks he media ion o con en in e es be ween
pe sis ence ac o s and eading beha io —di ec ly leading o non-
enewal. Simila ly, G ams (2025) no ed ha poo accessibili y o
academic ebooks (e.g., lacking ex sea ch o anno a ion ea u es)
weakens con en in e es , educing long- e m engagemen and
inc easing non- enewal isk. These s udies con i m ha non-
enewal’s s abili y in he CRT model e lec s eal-wo ld, pe sis en
ba ie s o accessible, in e es -ma ched eading.
By con as , " enewal" (Ca ego y 1) p edic ion pe o mance was
weak and ollowed a " ise- hen- all" end: 28.3% (2016–2018),
31.0% (2019–2021), and 29.3% (2022–2024). This ola ili y
indica es enewal beha io depends on dynamic, con ex -dependen
ac o s—many o which a e ied o con en in e es and
en i onmen al suppo , which shi o e ime. Fo example, in 2019–
2021 (a pe iod o imp o ed enewal p edic ion), sociological eading
in e en ions inc easingly ocused on p ac ical, in e es -d i en
con en . By 2022–2024, enewal p edic ion declined, possibly due o
he ise o di e se, agmen ed eading con ex s (e.g., digi al s.
physical ma e ials) ha in oduce new a iables (e.g., echnology
amilia i y) no cap u ed by he model. Rod iguez-Gallego e al.
(2024)’s s udy o Book ube communi ies—whe e digi al
engagemen media es sociological eading ia con en in e es —
highligh s how eme ging con ex s add complexi y o enewal
beha io , making i ha de o ixed a iables o p edic .
The o e all co ec a e mi o ed " enewal" p edic ion ends: 84.7%
(2016–2018), 86.5% (2019–2021), 82.8% (2022–2024), con i ming
ha enewal beha io is he co e bo leneck o model e ec i eness.
This aligns wi h sociological eading esea ch, whe e sus ained
engagemen ( enewal) is he mos a iable ou come—dependen on
ac o s like con en ele ance, esou ce access, and en i onmen al
suppo . Fo ins ance, Dino ola and Tes a (2025) ound ha public
lib a ies’ ansmedia shel design enhances se endipi ous access o
sociological con en , and con en in e es media es exposu e o
eading beha io —bu his e ec a ies by lib a y layou and eade
demog aphics, leading o inconsis en enewal. Simila ly, Ba es e
al. (2023) no ed ha sociological heme in e es media es
ins uc ional design and ac i e eading o heal h science s uden s,
bu his media ion weakens i hemes do no align wi h s uden s’
e ol ing ca ee in e es s— u he explaining why enewal (and hus
o e all model pe o mance) luc ua es.
5.5.3 Risk Changes in Renewal Beha io P edic ion and
Po en ial Causes
Model isk ollowed a " all- hen- ise" end, in e sely co ela ed
wi h o e all p edic ion pe o mance: mode a e isk in 2016–2018
( esubs i u ion = 0.153, c oss- alida ion = 0.157), lowes in 2019–
2021 (0.135, 0.138), and highes in 2022–2024 (0.172, 0.177). This
in e se ela ionship con i ms ha isk p ima ily s ems om uns able
enewal p edic ion—a pa e n suppo ed by eal-wo ld eading
dynamics, whe e enewal depends on mu able ac o s like con en
in e es and esou ce a ailabili y. Fo example, in 2019–2021 (low
isk), lib a ies and schools expanded access o in e es -aligned
sociological ma e ials. By 2022–2024 (high isk), ac o s like
limi ed digi al access o sociological books and shi ing eade
p e e ences (e.g., owa d sho - o m digi al con en ) in oduced
unp edic abili y in o enewal beha io , inc easing model isk.
The 2022–2024 isk inc ease may be linked o he exclusion o
"Gende " and educed a iable dimensions, which weakened he
model’s explana o y powe . Howe e , enewal beha io in his
pe iod likely depends on uninco po a ed a iables ied o con en
and con ex — ac o s mo e in luen ial han gende . Addi ionally, he
ise o in e disciplina y sociological eading (e.g., linking sociology
o heal hca e o en i onmen al s udies; Di ga e al., 2024;
Abimbola, 2025) in oduces new d i e s o enewal ha he model’s
simpli ied a iables ail o accoun o — u he inc easing isk.
Conclusion
This s udy sys ema ically analyzed he hema ic cha ac e is ics o
sociological books, he dynamics o high- equency subjec wo ds
ac oss ime, and he ac o s in luencing bo owing and enewal
beha io s, p o iding comp ehensi e insigh s in o he eade
a en ion ends and beha io al mechanisms in sociological book
consump ion.
Fi s , ega ding hema ic con en , he op 10 opics in sociological
books e lec a balanced s uc u e o ounda ional disciplines,
in e disciplina y in eg a ion, and p ac ical o ien a ion. Co e
sociology disciplines (e.g., ins i u ional sociology, u al s udies)
o m he heo e ical backbone, while in e disciplina y hemes (e.g.,
Li e a u e, A & Philosophy; Psychology & Social Secu i y) expand
he esea ch scope by linking sociology o humani ies and social
wel a e. P ac ical hemes such as Highe & Voca ional Educa ion
and Academic Exchange u he highligh he discipline’s
esponsi eness o eal-wo ld needs, b idging academic inqui y wi h
social de elopmen .
Second, he analysis o op 100 high- equency subjec wo ds ac oss
h ee pe iods (2016–2018, 2019–2021, 2022–2024) e eals wo key
ends: empo al s abili y o co e hemes and phased e olu ion o
eme ging opics. Eigh y-six c oss-pe iod wo ds (86%)—including
― esea ch,‖ ―social,‖ ―sociology,‖ and ―educa ion‖—con i m ha
eade s’ a en ion o ounda ional sociological domains (e.g.,
esea ch me hods, social s uc u e, educa ion) has emained
consis en o e nea ly a decade. In con as , pe iod-unique wo ds
e lec adap i e shi s: om ―p ac ical specializa ion‖ (e.g.,
―sec e a ial,‖ ―consul ing‖) in 2016–2018 and ―educa ional ex
ocus‖ (e.g., ―uni e si ies,‖ ―essays‖) in 2019–2021 o ― echnology
in eg a ion and con empo a y social issues‖ (e.g., ― echnology,‖
―aging,‖ ― eminis ‖) in 2022–2024. This e olu ion unde sco es
sociology’s abili y o align wi h socie al changes and expanding
eade in e es s.
Thi d, heme clus e analysis u he illus a es he p og essi e
expansion o sociological inqui y. The 2016–2018 clus e s cen e ed
on adi ional ounda ional opics (e.g., in e pe sonal ela ionships,
Resea ch Pape
This wo k is licensed unde a C ea i e Commons A ibu ion-NonComme cial 4.0 In e na ional License.
DOI: 10.5281/zenodo.17322626
Page 41
cul u al an h opology); 2019–2021 shi ed o applied mode n issues
(e.g., u baniza ion, public policy); and 2022–2024 in eg a ed
heo e ical dep h, in e disciplina y collabo a ion (e.g., psychology,
s a is ics), and global challenges (e.g., ecological conce ns). This
ajec o y e lec s a shi om s a ic desc ip ion o social phenomena
o dynamic, holis ic engagemen wi h complex con empo a y issues.
Fou h, linea eg ession and media ion e ec analyses cla i ied he
d i e s o ―To al Bo owings (TB).‖ The linea model explained
62.1% o TB a iance, wi h ―Leng h o Bo owing Yea s (LBY)‖
iden i ied as he mos in luen ial p edic o . Media ion analysis
(PROCESS Model 4, N=13,344) con i med ha ―Media o Va iable
C9‖ plays a signi ican pa ial media ing ole in he LBY→TB
ela ionship: LBY explains 28.71% o C9 a iance, and LBY + C9
oge he explain 89.23% o TB a iance ( a exceeding he 30.91%
explained by LBY alone). The indi ec e ec (6.0244) and di ec
e ec (0.8871) we e bo h highly signi ican (p<0.001), wi h
Boo s ap 95% CIs excluding 0, e i ying he obus ness o he
media ion mechanism.
Finally, CRT (Classi ica ion and Reg ession T ee) analysis o
enewal beha io e ealed phased di e ences in p edic ion
pe o mance. Ac oss all pe iods, he model exhibi ed excellen
accu acy in p edic ing ―Non- enewal‖ (96.5%–96.9%) bu weak
pe o mance in p edic ing ―Renewal‖ (28.3%–31.0%). The 2019–
2021 model (19 nodes, 4 dep h) showed he lowes isk
( esubs i u ion isk=0.135) and highes o e all co ec a e (86.5%),
while he 2022–2024 model (excluding ―Gende ‖ as a p edic o ) saw
inc eased isk (0.172) and dec eased accu acy (82.8%). This
sugges s ha enewal beha io is shaped by complex, ime- a ying
ac o s, and excluding po en ial p edic o s may unde mine model
e ec i eness.
Despi e i s con ibu ions, his s udy has se e al limi a ions ha
wa an a en ion in u u e esea ch. This s udy ocused on high-
equency subjec wo ds and b oad hemes bu did no explo e he
dep h o hema ic con en (e.g., how speci ic sub- hemes o ―gende
s udies‖ e ol e) o eade demog aphics (e.g., whe he s uden s.
p o essional eade s di e in hema ic p e e ences). The media ion
analysis con i med C9’s ole bu did no de ine i s heo e ical o
ope a ional meaning (e.g., whe he C9 ep esen s ― eading
equency,‖ ―con en ele ance,‖ o ― eade sa is ac ion‖). Cla i ying
C9’s concep ual de ini ion and alida ing i wi h addi ional da ase s
would s eng hen he gene alizabili y o he media ion mechanism.
The 2022–2024 CRT model excluded ―Gende ‖ and s ill ailed o
p edic ―Renewal‖ e ec i ely, indica ing unmeasu ed a iables
(e.g., digi al s. physical book o ma , seasonal bo owing ends, o
eade engagemen wi h speci ic hemes) may in luence enewal.
The s udy analyzed agg ega e da a wi hou compa ing hema ic o
beha io al di e ences ac oss egions (e.g., u ban s. u al lib a ies)
o coun ies. C oss- egional s udies could e eal how cul u al,
economic, o educa ional con ex s shape sociological book
consump ion. While he s udy compa ed h ee ime pe iods, i did
no adop a con inuous longi udinal design o cap u e eal- ime shi s
in eade beha io (e.g., how e en s like he COVID-19 pandemic
impac ed bo owing pa e ns). Fu u e esea ch should add ess hese
limi a ions by in eg a ing mul i-me hod (quali a i e + quan i a i e)
designs, expanding p edic o a iables, and inco po a ing con ex ual
and longi udinal da a. This will u he ad ance ou unde s anding o
sociological book consump ion and p o ide e idence-based
guidance o lib a y collec ion de elopmen and eade se ice
op imiza ion.
Funding S a emen
This esea ch was unded by he 2025 Teaching Re o m Resea ch
P ojec o Jiangsu Uni e si y Lib a y Wo king Commi ee: Resea ch
on Demand Iden i ica ion and Resou ce Adap a ion Based on
Reade s' Reading Beha io s, China (2025JTYB39).
Re e ences
1. Abimbola, M. O. (2025). Clima e change li e acy o
young eade s h ough guided eading: A case s udy o
Soa ing C ew Reading Club. Jou nal o Poe y
The apy. h ps://doi.o g/10.1080/08893675.2025.2486259
2. Aga ono a, A. B. (2023). U ban e e yday li e and
elec i ica ion in con ex o uel c ises in i s qua e o
20 h cen u y. Nauchnyi Dialog, 12(8), 348–
371. h ps://doi.o g/10.24224/2227-1295-2023-12-8-348-
371
3. Aleksic, G., & Du us, N. M. (2025). Home li e acy
en i onmen o mul ilingual p eschool child en. Jou nal o
Educa ional Psychology, 117(5), 717–
736. h ps://doi.o g/10.1037/edu0000938
4. Anan hak ishnan, U. M., Basa a aj, N., Ka megam, S. R.,
Sen, A., & Smi h, M. D. (2025). Book Bans in Ame ican
Lib a ies: Impac o Poli ics on Inclusi e Con en
Consump ion. Ma ke ing Science, 44(4), 933–953.
h ps://doi.o g/10.1287/mksc.2024.0716
5. A oyo-Vazquez, N., Al i e-Diez, M. L., Rod iguez-
B a o, B., & Gonzalez-Talad iz, L. de C. (2023). E-books
in Spanish uni e si y lib a ies: An analysis o online social
sciences deg ees. Jou nal o Academic Lib a ianship,
49(6), A icle
102809. h ps://doi.o g/10.1016/j.acalib.2023.102809
6. Ba es, S., Fishe , K., Wa son, J., Fai , K., O wig, T., &
Walke , D. (2023). S uc u al compe ence and equi y-
minded in e p o essional educa ion: A common eading
app oach o lea ning. Educa ion Sciences, 13(3), A icle
287. h ps://doi.o g/10.3390/educsci13030287
7. Buse a, G., Campolo, M. G., & Pana ello, D. (2024).
Changes in daily li e habi s du ing COVID-19 and hei
ansi o y and pe manen e ec s on I alian uni e si y
s uden s' anxie y le el. Applied Resea ch in Quali y o
Li e, 19(4), 1711–1733. h ps://doi.o g/10.1007/s11482-
024-10304-1
8. Chilu ya, E. M., A ola, M., Se e insson, S., Bylund, A.
M., S enliden, L., Mo aza i, A., & Ziemke, T. (2025).
The BookBo p ojec : Concep ual design o a social obo
acili a ing eading mo i a ion. In A. Fols ad, S.
Papadopoulos, T. A aujo, E. L. Law, E. Luge , S. Hobe ,
& P. B. B and zaeg (Eds.), Cha bo s and Human-Cen e ed
AI, Con e sa ions 2024 (Vol. 15545, pp. 132–149). Cham,
Swi ze land: Sp inge . h ps://doi.o g/10.1007/978-3-031-
88045-2_9
9. Dino ola, S., & Tes a, R. (2025). T ansmedia lib a y shel
expe ience: Inno a i e esea ch app oaches. Jou nal o
Lib a ianship and In o ma ion
Science. h ps://doi.o g/10.1177/09610006241310908
10. Di ga , E., Be se, S., Bo , N. A., & Tosun, B. (2024). The
e ec o concep mapping in nu sing educa ion on c i ical
hinking mo i a ion: A pa h analysis. Nu se Educa ion
Resea ch Pape
This wo k is licensed unde a C ea i e Commons A ibu ion-NonComme cial 4.0 In e na ional License.
DOI: 10.5281/zenodo.17322626
Page 42
Today, 143,
106386. h ps://doi.o g/10.1016/j.ned .2024.106386
11. Feng, Z., Liu, H., Zhang, W., Liu, Y., Xiong, N., Chen,
W., Yang, J., Wu, X., & Dai, W. (2025). Associa ions o
li es yle beha io s wi h anxie y and dep ession in Type 2
Diabe es Melli us pa ien s wi h hype ension: A c oss-
sec ional s udy. Psychia ic
Qua e ly. h ps://doi.o g/10.1007/s11126-025-10136-8
12. Ga cia-Gonzalez, M., Velazquez-Bu gos, R., Veliz, S.,
Oya zun, J. D., & Saona, I. (2023). Discu si e luc ua ions
o he school lib a y p og am in Chile: A c i ical policy
analysis. Lib i-In e na ional Jou nal o Lib a ies and
In o ma ion S udies, 73(3), 199–
208. h ps://doi.o g/10.1515/lib i-2023-0005
13. Gonzalez, J. E., Kim, H., Ande son, J., & Polla d-
Du odola, S. (2024). The e ec s o a science and social
s udies con en ich sha ed eading in e en ion on he
ocabula y lea ning o p eschool dual language
lea ne s. Ea ly Childhood Resea ch Qua e ly, 66, 34–
47. h ps://doi.o g/10.1016/j.ec esq.2023.08.011
14. G ams, K. (2025). An inc ease in academic ebook
p e e ences: A decade compa ison o ebook use e sus
non-use. E idence Based Lib a y and In o ma ion
P ac ice, 20(2), 92–
95. h ps://doi.o g/10.18438/eblip30732
15. Hiebe , E. H. (2023). Thinking h ough esea ch and he
science o eading. Phi Del a Kappan, 105(2), 37–
41. h ps://doi.o g/10.1177/00317217231205940
16. Hiidenmaa, P., Lindh, I., Linko, M., Suomalainen, R., &
Tossa ainen, T. (2024). Reading cul u e as sha ed e hos:
A s udy o Finnish sel -iden i ied eade s. Poe ics, 105,
A icle
101912. h ps://doi.o g/10.1016/j.poe ic.2024.101912
17. Jones, B., & Lynch, J. (2023). Teache s' and Black
s uden s' iews on he inco po a ion o A ican Ame ican
child en's li e a u e in an a e -school book club:
Collabo a i e and cul u ally based lea ning. Li e acy,
57(3), 209–220. h ps://doi.o g/10.1111/li .12326
18. Kad iu-Ci aku, S., & Malaj-Abdyli, F. (2024). Li e a u e
in p ima y schools in Koso o, i s a is ic and educa ional
cha ac e . Eu asian Jou nal o Educa ional Resea ch,
(110), 281–296. h ps://doi.o g/10.14689/eje .2024.110.17
19. Lee, J., Co ie, B., Da is, K. E., Wa en, C., & Zhang, T.
(2023). P omising p ac ices o p omo e physical ac i i y
and heal hy ea ing among Head S a ca egi e s and hei
child en. Ame ican Jou nal o Li es yle
Medicine. h ps://doi.o g/10.1177/15598276231221512
20. Le usalimskiy, Y. Y., Ie usalimskaya, S. Y., & Ha i ono ,
G. P. (2023). Pa ish lib a ies o he Ya osla l diocese in
he second hal o he 19 h–ea ly 20 h cen u ies. Ves nik
Tomskogo Gosuda s ennogo Uni e si e a Is o iya-Tomsk
S a e Uni e si y Jou nal o His o y, (85), 39–
46. h ps://doi.o g/10.17223/19988613/85/5
21. Lewis, N. A., Yoneda, T., Melis, R. J. F., M oczek, D. K.,
Ho e , S. M., & Muniz-Te e a, G. (2023). A ailabili y o
cogni i e esou ces in ea ly li e p edic s ansi ions
be ween cogni i e s a es in middle and olde adul s om
Eu ope. Inno a ion in Aging, 7(9),
igad124. h ps://doi.o g/10.1093/ge oni/igad124
22. Li, H., Majumda , R., & Oga a, H. (2025). Sel -di ec ed
ex ensi e eading wi h social suppo : E ec on eading
and lea ning pe o mance o high and low English
p o iciency s uden s. Resea ch and P ac ice in Technology
Enhanced Lea ning, 20, A icle
25. h ps://doi.o g/10.58459/ p el.2025.20025
23. Mohammed, S., A aya, A., & Abuka i, A. S. (2023).
Reading, singing, and s o y elling: The impac o
ca egi e -child in e ac ion and child access o books and
p eschool on ea ly childhood de elopmen in
Ghana. Scien i ic Repo s, 13(1), A icle
13751. h ps://doi.o g/10.1038/s41598-023-38439-5
24. Nausa, R., Zi ko ic, J., & Sichko, L. (2024). Needs
analysis and design o a mas e 's le el academic eading
cou se in English. P o ile-Issues in Teache s P o essional
De elopmen , 26(2), 89–
106. h ps://doi.o g/10.15446/p o ile. 26n2.110905
25. Nenonen, T., Kou onen, A., Ha kanen, T., & Ma elin, T.
(2025). Cul u al ac i i ies and all-cause mo ali y among
Finnish adul s: A 19-yea ollow-up. BMC Public Heal h,
25(1), 1906. h ps://doi.o g/10.1186/s12889-025-23032-4
26. O'B ien, B. A., A shad, A., & Ng, S. C. (2024). Bilingual
amilies and he home li e acy en i onmen : An
examina ion o English and e hnic language ac i i ies and
ou comes. Jou nal o Resea ch in Reading, 47(4), 580–
600. h ps://doi.o g/10.1111/1467-9817.12465
27. O ani, S., Sa o, K., & Kondo, N. (2025). Public lib a ies
and unc ional disabili y: A coho s udy o Japanese olde
adul s. SSM-Popula ion Heal h, 29,
101762. h ps://doi.o g/10.1016/j.ssmph.2025.101762
28. Polla d-Du odola, S. D., Gonzalez, J. E., Saenz, L., &
Da is, H. S. (2024). The sus ainabili y o con en en iched
sha ed book eading ocabula y p ac ices and p eschool
eme gen bilinguals. In e na ional Jou nal o Bilingual
Educa ion and Bilingualism, 27(3), 315–
331. h ps://doi.o g/10.1080/13670050.2023.2174372
29. Rod iguez-Gallego, M. R., De-Cecilia-Rod iguez, C.,
Domene-Ma os, S., & O donez-Sie a, R. (2024).
Book ube as a li e a y p ac ice in seconda y educa ion: A
sys ema ic e iew. Ocnos-Jou nal o Reading Resea ch,
23(1), A icle
e22549099. h ps://doi.o g/10.18239/ocnos_2024.23.1.405
30. Shen, X., Wang, J., Chen, J., Zhang, H., Shen, S., & Zhao,
X. (2024). Rela ionship be ween pa icipa ion in leisu e
ac i i ies and he main enance o success ul aging in olde
Chinese adul s: A 4-yea longi udinal s udy. BMC
Ge ia ics, 24(1), A icle
989. h ps://doi.o g/10.1186/s12877-024-05574-1
31. Shigemasu, K., Kono, M., Ikemo o, S., & Akabayashi, H.
(2024). Causal e ec o pa en al eading on la e
de elopmen o child en: Demons a ing a Bayesian
app oach. B i ish Jou nal o De elopmen al Psychology,
42(3), 293–304. h ps://doi.o g/10.1111/bjdp.12482
32. Sian u i, M., & Hu i , A. A. (2024). ―I wan o ead his
book again!‖ Decolonizing child en's li e a u e o suppo
Resea ch Pape
This wo k is licensed unde a C ea i e Commons A ibu ion-NonComme cial 4.0 In e na ional License.
DOI: 10.5281/zenodo.17322626
Page 43
indigenous child en in eading and ma hema ics
lea ning. Jou nal o In e cul u al S udies, 45(2), 338–
362. h ps://doi.o g/10.1080/07256868.2023.2247345
33. Si isena, M., Lhussie , M., Kane , E., Wea n, A., G ay, J.,
James, R., & Redga e, S. (2024). ―The book's a
con e sa ion s a e ‖: A ealis explo a ion o he
salu ogenic po en ial o eading o pleasu e. Medical
Humani ies, 50(3), 504–
512. h ps://doi.o g/10.1136/medhum-2023-012880
34. S i as a a, P., & S i as a a, M. (2023). P e alence and
co ela es o loneliness in he la e li e—insigh s om
Longi udinal Ageing S udy in India (LASI) Wa e-
1. Indian Jou nal o Psychia y, 65(9), 914–
921. h ps://doi.o g/10.4103/indianjpsychia y.indianjpsyc
hia y_594_22
35. Sun, H., & Xi, Y. (2025). Reading oo s o esilience: The
ole o childhood eading en i onmen s in closing he
educa ional gap in la e -li e cogni i e
unc ioning. Jou nals o Ge on ology Se ies B-
Psychological Sciences and Social Sciences, 80(8), A icle
gba 128. h ps://doi.o g/10.1093/ge onb/gba 128
36. Tani, Y., Kawaha a, T., Sugiha a, G., Machida, M.,
Amagasa, S., Mu ayama, H., Inoue, S., Fujiwa a, T., &
Shobugawa, Y. (2024). Childhood book a ailabili y helps
o p ese e cogni i e unc ion in olde adul s wi h low
educa ion: Resul s om he NEIGE s udy. Jou nals o
Ge on ology Se ies B-Psychological Sciences and Social
Sciences, 79(6), A icle
gbae052. h ps://doi.o g/10.1093/ge onb/gbae052
37. T an, T. T., Snasel, V., & Nguyen, L. T. (2023).
Combining social ela ions and in e ac ion da a in
ecommende sys em wi h g aph con olu ion collabo a i e
il e ing. IEEE Access, 11, 139759–
139770. h ps://doi.o g/10.1109/ACCESS.2023.3340209
38. T o suk, I. (2023). A book on economics o sociological
eading. Jou nal o Economic Sociology-Ekonomicheskaya
So siologiya, 24(1), 133–
148. h ps://doi.o g/10.17323/1726-3247-2023-1-133-148
39. Wang, Z., & Deng, R. (2024). S udy on use in en ions o
a sus ainable communi y idle book ecycling and eading
se ice sys em om he pe spec i e o alue co-c ea ion:
A design p oposal and alida ion based on Chinese
communi y esiden s. En i onmen De elopmen and
Sus ainabili y. h ps://doi.o g/10.1007/s10668-024-05781-
1
40. Webbe , C., Bohan, K., Collins, H., Ewens, J., Wilkinson,
K., & McGeown, S. (2025). Beyond books: High school
lib a ians as champions o pupil inclusi i y, au onomy,
and eade de elopmen . Jou nal o Lib a ianship and
In o ma ion Science, 57(3), 674–
686. h ps://doi.o g/10.1177/09610006241230101
41. Wes e eld, M. F., Malone, S. A., Clendon, S., Bowen, R.,
Hayley, G., & Payn e , J. (2024). The home li e acy
en i onmen o school-age au is ic child en wi h high
suppo needs. Jou nal o Applied Resea ch in In ellec ual
Disabili ies, 37(5),
e13284. h ps://doi.o g/10.1111/ja .13284
42. Wu, R., Liu, F., Si u, X., & Huang, W. (2023). ―Reading a
book, by a gain in you wi ‖—How o cul i a e high
school s uden s' sus ainable eading: A mul i- ac o
analysis. Heliyon, 9(12), A icle
e23333. h ps://doi.o g/10.1016/j.heliyon.2023.e23333
43. Yang, H.-T., Liao, C.-Y., Syu, C.-M., & Chan, T.-W.
(2023). Explo ing he impac on s uden eading
p e e ences in gami ied eading po olio. In 31s
In e na ional Con e ence on Compu e s in Educa ion
(ICCE 2023), Vol. I (pp. 704–706). Singapo e:
IEEE. h ps://doi.o g/10.1109/ICCE58068.2023.00111
44. Yu, L., Huang, J., Liu, P. D., Yeung, S. S.-S., Lin, D.,
Cheung, H., & Tong, X. (2024). How pa en ing s yles
a ec he de elopmen o language skills and eading
comp ehension in p ima y school s uden s. B i ish Jou nal
o Educa ional Psychology, 94(4), 1245–
1270. h ps://doi.o g/10.1111/bjep.12718
45. Zheng, W. (2024). Reading ecommenda ion echnology
in digi al lib a ies based on eade s' social ela ionships
and eade s' in e es s. In e na ional Jou nal o Ad anced
Compu e Science and Applica ions, 15(7), 689–
697. h ps://doi.o g/10.17223/15617793/505/2
