189189
Associa ing speake a iables wi h English
p onuncia ion a ings in Spanish e ia y educa ion
ma eusz Pie aszek
Uni e sidad F ancisco de Vi o ia
Recei ed: 2023-01-10 / Accep ed: 2023-09-15
DOI: h ps://doi.o g/10.30827/po alin. i41.27083
ISSN pape edi ion: 1697-7467, ISSN digi al edi ion: 2695-8244
ABSTRACT: Al hough he use o English in he e e -changing con ex o bilingual e -
ia y educa ion has been ex ensi ely s udied o da e, p onuncia ion esea ch in he ield is
s ill sca ce. This pape aims a p o iding empi ical da a ega ding he hypo hesised impac
o bilingual educa ion p og ammes on p onuncia ion pe o mance ope a ionalised h ough
he no ions o in elligibili y, comp ehensibili y and o eign-accen edness. Fo his pu pose,
60 ad anced speake s o English om a Spanish uni e si y (en olled bo h in bilingual and
monolingual deg ee cou ses) we e eco ded and assessed by 330 in e na ional English
speake s om 30 di e en L1s. Bo h he speake s and he lis ene s comple ed a basic socio-
demog aphic ques ionnai e. The independen a iables we e ela ed o he pa icipan s’ (1)
epo ed exposu e o bilingual ins uc ion a he p e ious and cu en le els o educa ion,
as well as (2) hei o e all le el in English and (3) hei exposu e o English ab oad. The
indings showed ha bilingual educa ion was in no way associa ed wi h di e ences in p o-
nuncia ion a ings. Howe e , he speake s’ le el and hei expe ience ab oad we e co ela ed
wi h he comp ehensibili y and o eign-accen edness sco es. The esul s a e no only conse-
quen ial o he unde s anding o he (lack o ) impac o bilingual a iables on p onuncia ion,
bu hey also allow o examine he adequacy o he CEFR phonological desc ip o s.
Keywo ds: bilingual educa ion, EMI, le el, English p onuncia ion, in elligibili y
La asociación de las a iables del hablan e con la p onunciación en inglés en la uni e -
sidad española
RESUMEN: Aunque el uso del inglés en el con ex o cambian e de la educación supe io
bilingüe ha sido ex ensamen e es udiado, la in es igación sob e la p onunciación sigue sien-
do escasa. El obje i o de es e a ículo es p opo ciona da os empí icos sob e un hipo é ico
impac o de los p og amas bilingües en la p onunciación concebida median e las nociones de
in eligibilidad, comp ensibilidad y g ado de acen o ex anje o. Con es e in, 60 uni e si a-
ios españoles (ma iculados an o en g ados bilingües en monolingües) ue on g abados y
– pos e io men e – e aluados po 330 hablan es de inglés con 30 lenguas na i as di e en es.
También se ecogie on da os sociodemog á icos básicos sob e los pa icipan es. Las a ia-
bles independien es es án basadas en la in o mación p opo cionada po los hablan es ela i a
a (1) su pa icipación en p og amas bilingües an es de la uni e sidad y du an e su e apa
uni e si a ia, (2) su ni el de inglés y (3) su expe iencia en el ex anje o. Los análisis no han
demos ado asociación alguna en e la educación bilingüe y el desempeño en p onunciación.
Sin emba go, los esul ados mues an una co elación en e el ni el de inglés y las es ancias
en el ex anje o, po un lado, y la comp ensibilidad y el g ado de acen o ex anje o, po o o.
Los esul ados del es udio no solo son ele an es pa a en ende mejo el impac o de la edu-
cación bilingüe en la p onunciación de los alumnos (o la al a del mismo), sino que ambién
pe mi en examina la idoneidad de los desc ip o es onológicos del MCER.
Po a Lingua um, Issue 41, Janua y 2024 189-207
190
Po a Lingua um Issue 41, Janua y 2024
Palab as cla e: enseñanza bilingüe, EMI, ni el, p onunciación del inglés, in eligibilidad
1. In oduc Ion
English p onuncia ion is o en conside ed o be a di icul linguis ic skill o Spanish
lea ne s o English – an opinion which has o en made i s way o he Spanish media (c .
El Con idencial, 2014; Gal án, 2010; La Razón, 2019). I is also one o he i s aspec s
o language based on which imp essionis ic judgmen s a e made ega ding he speake s’
linguis ic compe ence (Cu illas Espinosa, 2017; Lippi-G een, 1994). As he exac measu e-
men o p onuncia ion skills may be a bu densome ask and he way i is cons ued may
p o e somewha elusi e, i has been one o he leas s udied (and augh ) aspec s o second
language acquisi ion o e en – as Gilbe (2010, p. 3) pu i – “ he o phan o ESL/EFL”.
This pape ocuses on he ela ionship be ween some cha ac e is ics o Spanish uni e si y
s uden s as p edic o s o hei p onuncia ion pe o mance in an educa ional con ex whe e he
p esence o English has been inc easing exponen ially o he pas wo decades (Diogo e
al., 2019; Fielden Bu ns e al., 2018; Foske , 2010; Ramos Ga cía, 2013; Ramos Ga cía &
Pa ón Vázquez, 2018). Said a iables comp ise he di e en con ex s in which he s uden s
ha e been exposed o English as well as hei le el. The da a p esen ed in his a icle a e
pa o a b oade esea ch p ojec in es iga ing he a iabili y o Spanish s uden s’ p onun-
cia ion sco es conside ing cha ac e is ics ela ed o he speake s, lis ene s and phonological
ea u es o he sampled speech.
The measu emen o p onuncia ion has been ope a ionalised by means o h ee di e en
a iables: in elligibili y (INT), comp ehensibili y (COM) and o eign-accen edness (FA), each
ela ed o a di e en aspec o he pa icipan s’ phonological pe o mance. The i s one,
i.e., in elligibili y, can be bes de ined as he objec i e measu e o he segmen al decoding
o he speech o a pe son by a lis ene (Jenkins, 2000; Walke , 2010). On he o he hand,
he mo e subjec i e dimensions o comp ehensibili y and o eign-accen edness – as de ined
in s udies conduc ed by De wing and Mun o (2015) – e lec he a ings gi en by lis ene s
on Seman ic Di e en ial Scales (Dö nyei, 2007; Poljak 2019). While he o me gauges
he pe cei ed ease o unde s anding, he la e is conce ned wi h he pe cei ed deg ee o
di e gence om L1 speake s’ p onuncia ion. The in elligibili y a iable in his pape also
oughly ollows he way i is cons ued by De wing and Mun o (2009, 2015) in he sense
ha i in ol es eal unde s anding. Howe e , said unde s anding is educed o he decoding
o phonological componen by elimina ing as much seman ic in o ma ion as possible using
Seman ically Unp edic able Sen ences (SUS), as explained la e in he me hodology sec ion
(Benoî e al., 1996; Wang, 2007). This di e s om in elligibili y s udies whe e he a iable
is measu ed as he unde s anding o seman ically meaning ul ex s by means o ansc ip ions
o o he asks ocused on he meaning o he ex p esen ed in he speech samples (De wing
& Mun o, 1997, 2015). In ac , in elligibili y has been es ed h ough a a ie y o asks o
da e, including con en ques ions (mul iple choice, ue/ alse, speci ic in o ma ion), ans-
la ion, ansc ip ions (o meaning ul sen ences), SUS and eac ion imes (Gooskens, 2013).
Addi ionally, he e m in elligibili y has also been used o deno e wha is concep ualised as
comp ehensibili y in his s udy, i.e., he lis ene s’ opinions o how unde s andable hey ound a
speake a he han hei eal unde s anding o speech samples (Gooskens, 2013; Haugen, 1966).
191
Ma eusz Pie aszek Associa ing speake a iables wi h English p onuncia ion...
Focusing on he phonological componen aligns wi h he iews held by English as a
Lingua F anca esea che s (Jenkins, 2000; Walke , 2010). Thei concep o in elligibili y can
be aced back o ha o Smi h and Nelson (1985), o whom i is he i s s age o language
p ocessing based solely on phonological in o ma ion. Jenkins (2000, p. 82, 83, 90) no iced
ha e en p o icien ELF use s use mo e bo om-up s a egies when decoding meaning han
na i e speake s, who ely on con ex ual cues o a much g ea e ex en ( op-down s a egies).
This, in u n, may lead o inc eased di icul ies in communica ion based on phonological
misunde s andings as opposed o mo phosyn ac ic e o s. These obse a ions we e co obo-
a ed by u he ELF esea ch (De e ding, 2013; Field, 2005; Meie ko d, 2004) and in o med
he decision ega ding he concep ualisa ion o in elligibili y in his s udy as he a ge ed
lis ene sample he ein is comp ised by ELF speake s om all a ound he globe. Indeed,
many esea che s (inside and ou side he ELF pa adigm) ha e in es iga ed he impac o
segmen al ea u es on in elligibili y (Ba ass e al., 2020; Jenkins, 2000; Jeong e al., 2017;
Jo o e al., 2007; Le is, 2018; Minema su e al., 2011; O’Neal, 2015; Walke , 2010; Wang,
2007), and comp ehensibili y and accen edness (Ba ass, 2017; C ow he e al., 2015; Sai o
e al., 2016; T o imo ich & Isaacs, 2012) ou side he Spanish con ex .
As ega ds he a iables ela ed o he speake s’ cha ac e is ics conside ed he ein, a el-
a i ely high numbe o s udies ha e been conduc ed on he e ec s o bilingual p og ammes
a p e-uni e si y le els (e.g., Galla do del Pue o & Ma ínez Ad ián, 2013; Ruiz de Za obe,
2011). While ce ain s udies showed a posi i e e ec o CLIL (Con en and Language In e-
g a ed Lea ning) ins uc ion on p onuncia ion (Lasagabas e , 2008; Pé ez-Cañado, 2018; Rallo
Fab a & Juan-Ga au, 2010), o he s showed no such e ec (Dal on-Pu e , 2008; Galla do del
Pue o e al., 2009; Galla do del Pue o & Gómez Lacabex, 2017; Pé ez-Cañado & Lancas e ,
2017; Rallo Fab a & Jacob, 2015). The esul s o he esea ch being inconclusi e a bes ,
Me ino and Lasagabas e (2018, p. 21) s ill claimed ha “p onuncia ion happens o be he
leas a ec ed o such dimensions”, hus ei e a ing a simila s a emen by Dal on-Pu e (2011).
A he e ia y le el o educa ion, he e has been a dea h o s udies dealing wi h
language compe ence – le alone p onuncia ion. Rega ding phone ic pe o mance in EMI
(English Medium Ins uc ion), some e idence was ound o i s impac on o eign accen
in he Aus ian con ex (Rich e , 2015, p. 221) when compa ing EMI s uden s wi h hose
en olled on an English Language deg ee. The o me ou pe o med he la e as hei deg ee
o o eign accen diminished o an ex en signi ican ly g ea e han in he case o he Eng-
lish language deg ee s uden s. In Spain, a p elimina y s udy (Gómez Lacabex & Galla do
del Pue o, 2021, p. 133) wi h 12 EMI s uden s epo ed he pa icipan s’ comp ehensibili y
o be a ed a a 4.9/9 and in elligibili y - 0.43/1. Howe e , he esea ch does no explici ly
compa e EMI and non-EMI s uden s and, he e o e, one canno asce ain whe he he EMI
a iable is a wo k he e all o he ac o s being equal.
Apa om he di e en ypes o exposu e o English in domes ic con ex s, ano he
a iable ha is aken in o accoun in his esea ch is he speake s’ expe ience li ing ab oad.
Flege (1988) sugges ed ha accen s we e la gely se led a e a one-yea s ay ab oad. De wing
and Mun o (2015, p. 40-43) p o ide an o e iew o he s udies dealing wi h he e ec o he
leng h o esidence in na i e imme sion con ex s. Ou o he 13 su eyed pape s, 9 p o ide
e idence in a ou o such e ec s, including 3 ou o he 4 s udies dealing speci ically wi h
accen . A he same ime, esea ch in o s ays ab oad om he Spanish pe spec i e indica es
192
Po a Lingua um Issue 41, Janua y 2024
ha e en s ays as sho as 3 mon hs can lead o a milde o eign accen in bo h child en and
adul s (Muñoz & Llanes, 2014, p. 442). O he s udies sugges ed he e was no li le o no
e ec o a sho s udy ab oad on ce ain phonological ea u es, such as owel leng h o Voice
Onse Time (A ello & La a, 2014). Addi ionally, s udying ab oad had no signi ican impac
on he deg ee o accen edness a ed by non-na i e lis ene s (A ello e al., 2012). Juan-Ga-
au (2014) poin s ou ha while s udying ab oad may be ad an ageous o he de elopmen
o o al accu acy, i is no always he case and speake - ela ed a iables (e.g. p e-depa u e
le el o mo i a ion) may o en be associa ed wi h he end esul . A ecen li e a u e e iew
o he s udy ab oad expe ience esea ch (Moo e e al. 2021, p.1) highligh s ha in spi e o
he numbe o published a icles, “ he e idence emains inconclusi e”.
Finally, i is only logical ha p o iciency le el, one o he independen a iables con-
side ed in his esea ch, should bea on all aspec s o p onuncia ion – including phone ic
pe o mance. Language p o iciency has been p o ed o be a de e mining ac o in p edic ing
many di e ences in accen edness, comp ehensibili y and in elligibili y a ings (Ba ass, 2017;
De wing & Mun o, 1997; Sai o e al., 2016). Since he publica ion o he Common Eu ope-
an F amewo k o Re e ence o Languages (Council o Eu ope, 2001), a emp s ha e been
made a a sys ema ic desc ip ion o all aspec s o language skills and usage o eaching and
assessmen pu poses. Howe e , as shown in Table 1, he o iginal desc ip ion o phonological
con ol was sca ce. No only did i con ain highly imp essionis ic labels (clea , na u al, some
e o ) bu i also used he e ms na i e speake and o eign accen wi hou speci ying hei
exac meaning o sugges ing ways o ope a ionalising hem in p ac ice.
Table 1. Phonological con ol desc ip o s: CEFR 2001
C2 As C1
C1 Can a y in ona ion and place sen ence s ess co ec ly in o de o exp ess ine shades o meaning.
B2 Has acqui ed a clea , na u al, p onuncia ion and in ona ion.
B1 P onuncia ion is clea ly in elligible e en i a o eign accen is some imes e iden and occasional mis-
p onuncia ions occu .
A2 P onuncia ion is gene ally clea enough o be unde s ood despi e a no iceable o eign accen , bu con-
e sa ional pa ne s will need o ask o epe i ion om ime o ime.
A1 P onuncia ion o a e y limi ed epe oi e o lea n wo ds and ph ases can be unde s ood wi h some
e o by na i e speake used o dealing wi h speake s o his/he language g oup.
No e: Council o Eu ope, 2001, p. 117.
In a la e documen (Council o Eu ope, 2018), i was acknowledged ha he phono-
logical desc ip o s we e no speci ic enough in hei 2001 o m, which was e en dubbed
“un ealis ic” (p. 134). Also, he in elligibili y p inciple (c . Le is, 2005) was applied as
he phonological desc ip o s in he o iginal CEFR we e he only ones o make a e e ence
o a na i e model, albei indi ec ly h ough he exp ession o eign accen . In he CEFR
companion (2018, p. 134), in elligibili y is inally de ined as “accessibili y o meaning o
lis ene s, co e ing also he lis ene s’ pe cei ed di icul y in unde s anding (no mally e e ed
o as comp ehensibili y)”. Fu he in he documen (p. 135), i is added ha he in elligibili y
desc ip o s accoun o “ he ex en o in luence om o he languages”, “con ol o sounds”
193
Ma eusz Pie aszek Associa ing speake a iables wi h English p onuncia ion...
and “con ol o p osodic ea u es” making i e en ha de o pinpoin . I can, howe e , be
assumed ha he in elligibili y cons uc in he CEFR is ac ually – o a la ge ex en and as
s a ed in he documen – comp ehensibili y he way i is unde s ood in his esea ch and
p e ious s udies (De wing & Mun o, 2015), especially because no objec i e measu emen
me hods a e sugges ed, and only imp essionis ic desc ip o s a e o e ed. Qui e in e es ingly,
e e ences o accen (al hough no o eign accen anymo e) a e p esen a all le els.
Table 2. Phonological con ol desc ip o s: CEFR Companion 2018
C2 […] In elligibili y and e ec i e con eyance o and enhancemen o meaning a e no a ec ed in any
way by ea u es o accen ha may be e ained om o he language(s).
C1
Can employ he ull ange o phonological ea u es in he a ge language wi h su icien con ol o ensu e
in elligibili y h oughou . […] some ea u es o accen e ained om o he language(s) may be no icea-
ble, bu hey do no a ec in elligibili y.
B2 […] accen ends o be in luenced by o he language(s) he/she speaks, bu has li le o no e ec on in el-
ligibili y.
B1 P onuncia ion is gene ally in elligible.
A2
P onuncia ion is gene ally clea enough o be unde s ood. […] A s ong in luence om o he lan-
guage(s) he/she speaks on s ess, hy hm and in ona ion may a ec in elligibili y, equi ing collabo a ion
om in e locu o s […]
A1 P onuncia ion o a e y limi ed epe oi e o lea n wo ds and ph ases can be unde s ood wi h some
e o by in e locu o s.
No e: Council o Eu ope, 2018, p. 136.
The new desc ip o s (Table 2) show clea p og ession in in elligibili y (as de ined by he
CEFR Companion 2018 and oughly equi alen o comp ehensibili y in his s udy (De wing
& Mun o, 2015). Full p o iciency – whe e he e a e no in elligibili y di icul ies caused by
phonological con ol – is achie ed a ound le el C1. Indeed, he e is li le (i any) di e ence
be ween he e e ence o “in elligibili y” in he CEFR C1 and C2 desc ip o s. As discussed
la e on in his pape , he da a he ein shed new ligh on he applicabili y o hese desc ip o s
a leas in he Spanish con ex o his s udy.
The esea ch ques ions (RQs) add essed in his pape a e ela ed o he a o emen ioned
speake cha ac e is ics which ha e been hypo hesised o bea a po en ial ela ionship o he
lea ne s’ p onuncia ion pe o mance.
RQ1: Wha associa ion is he e be ween he speake s’ le el o English and hei COM-
FA-INT sco es?
RQ2: Wha is he e ec o he ype o exposu e o English (p e-uni e si y CLIL, EMI
and s ays ab oad) on he pa icipan s’ p onuncia ion sco es?
2. Me hodology
A he ini ial s age o his esea ch, 60 Spanish speake s o English we e eco ded a a
Spanish uni e si y (including 34 males and 26 emales) a e ob aining hei in o med con-

194
Po a Lingua um Issue 41, Janua y 2024
sen h ough a signed o m. Sociodemog aphic da a we e also collec ed. The pa icipa ion in
he s udy was olun a y and was no used o he pu poses o cou se assessmen . 32 o he
speake s we e en olled on English Medium Ins uc ion (EMI) cou ses while he emaining
28 we e monolingual deg ee s uden s (non-EMI). 39 had a ended bilingual schools p io o
uni e si y while 21 had no such expe ience. They we e all in hei 2nd, 3 d o 4 h yea , and
came om 5 di e en deg ees. Thei a e age age was 21.2. (Min = 19, Max = 26, SD =
1.44) and hey all had a leas a B2 le el in English acco ding o ins i u ional equi emen s
and placemen es s. Howe e , only hose who explici ly epo ed holding a ce i ied B2
(o highe ) le el o English we e conside ed in he analyses ega ding le el (n = 50) la e
in his pape while he emaining 10 we e excluded. In his pa o he sample, he e we e
27 B2, 15 C1, and 8 C2 speake s. Rega ding hei expe ience li ing ab oad (“Do you ha e
expe ience li ing ab oad”), 15 speake s epo ed ha ing none, while 45 answe ed “yes”. The
amoun o ime spen ab oad (“How much ime did you spend ab oad?”) anged om 0 o
6 yea s wi h an a e age si ua ed a 1.16 yea s. 21 ou o 45 s uden s spen less han a yea
ab oad, 7 we e away o a yea and he emaining 17, o mo e han a yea .
Be o e ga he ing he da a, he ins umen s we e alida ed by wo independen expe s
and pilo ed wi h a g oup o 28 s uden s wi h sa is ac o y esul s. The esea ch ools included:
(1) a se ies o 40 Seman ically Unp edic able Sen ences (hence o h SUS, see APPENDIX)
(Benoî e al., 1996; Wang, 2007), (2) an o iginal Elici a ion Pa ag aph (EP, see APPENDIX),
(3) a sociodemog aphic (backg ound) ques ionnai e, and (4) an a i udinal ques ionnai e (no
add essed in his pape ). The SUS we e hough o be an app op ia e ool as he limi ed
seman ic in o ma ion hey con ain would o ce he lis ene s o in e p e phone ic in o ma ion
wi hou eso ing o he seman ic con ex (c . Kang e al. 2018).
The eco dings ook place in emp y uni e si y class ooms and las ed a ound 5 minu es pe
speake . The pa icipan s had ime o ead he w i en p omp s be o e he eco ding and hey
could epea any ou pu hey we e no sa is ied wi h. A e he eco ding and he collec ion o
ques ionnai e da a, he inal online in elligibili y, comp ehensibili y and o eign-accen edness
es s we e se up and adminis e ed by means o an online pla o m (Limesu ey Gmbh.,
2019) o 330 lis ene in o man s om nea ly 30 L1 backg ounds, including L1 Spanish and
L1 English speake s. The lis ene s we e ec ui ed h ough an online sel -selec ion p ocess
and olun ee ed o pa icipa e in he esea ch. The e we e 206 (62.4%) emales and 120
(36.4%) males in he lis ene sample whe eas 4 (1.2%) people did no iden i y wi h ei he o
he wo gende s. The a e age age o he lis ene s was 30 (Min = 14, Max = 70, SD = 11.1).
Rega ding hei L1s, 33% we e na i e speake s o English 25.2% we e na i e speake s o
Spanish and 20.3% we e Polish1. Among he people whose mo he ongue was no English,
94% epo ed o ha e a B2 o highe le el in he language.
Each lis ene a ed a andom selec ion o i e speake s assigned by he pla o m. The
in elligibili y pa o he es consis ed in he o hog aphic ansc ip ion o audi o y inpu
ex ac ed om he speake s’ SUS, while comp ehensibili y and o eign-accen edness we e
measu ed using 6-poin seman ic di e en ial scales based on 25-30-wo d exce p s om he EP.
Only he ends o he scales we e labelled in he comp ehensibili y and o eign-accen edness
1 All he L1s epo ed in he s udy we e: A abic, Basque, Can onese, Ca alan, C oa , Danish, Du ch, English,
Finnish, F ench, Ilocano, I alian, Li huanian, Manda in, Nepali, Pe sian, Polish, Po uguese, Romanian, Russian,
Se bian, Spanish, Swedish, Tagalog, Tamil, Tu kish, U du, Vie namese.
195
Ma eusz Pie aszek Associa ing speake a iables wi h English p onuncia ion...
i ems o es . Each lis ene answe ed i e ques ions e alua ing he speake s’ comp ehensi-
bili y by comple ing he sen ence “The speake is ___ o unde s and” on a scale om (1)
e y di icul o (6) e y easy. The i e i ems assessing accen edness eques ed he lis ene s
o “ a e he speake ’s deg ee o o eign accen ” on a scale om (1) na i e-like o (6) e y
o eign. Each lis ene lis ened o and was asked o ansc ibe 20 di e en SUS, i.e., 4 SUS
o each o he 5 speake s andomly selec ed o ha lis ene by he pla o m. The same
SUS (p onounced by di e en speake s) we e ne e p esen in one lis ene ’s es . The o de
o he 30 i ems (20 SUS ansc ip ions and 10 scala i ems) in he es we e andomised by
he pla o m. No ime limi o he comple ion o he su ey was se and he a e age ime
necessa y o comple e he ask was 21 minu es.
The 6,600 SUS ansc ibed by he lis ene s we e con e ed in o nume ic in elligibili y
alues using a sco ing p o ocol de ised o he pu poses o he s udy. The o al numbe o
poin s a speake could ob ain om each lis ene on he SUS es was 30. The e was a o al
o 15 wo ds in he 4 SUS coming om each speake , each bea ing an equal weigh o 2
poin s. The ull 2 poin s we e allo ed o hose ansc ip ions which we e ully co ec , and
one poin was allo ed i one mis ake was p esen in a wo d, i.e., when a phoneme was no
ecognised co ec ly. E iden spelling mis akes o ypos (e.g., *bough o bough ) no esul ing
in o he exis ing wo ds and homophones (e.g., weak o week o cell o sell) we e ea ed
as co ec ansc ip ions. O he wise, poin s we e deduc ed. The a e age numbe o lis ene s
who a ed an indi idual speake was 27.5. Each indi idual speake ’s COM and FA sco es
we e ob ained by calcula ing he means o he poin s assigned by he lis ene s who a ed
hei speech. The speake ’s INT a ing was he a e age o hei lis ene s’ sco es on he SUS
ansc ip ion ask. The di e en magni udes o he scales used o COM/FA (Max = 6) and INT
(Max = 30) we e no p oblema ic as no c oss- a iable s a is ics we e calcula ed o his s udy
excep o co ela ions o which he ange o he scale is i ele an . I mean compa isons
we e necessa y, he sco es can be easily scaled by di iding he in elligibili y a ings by 5.
The ollowing independen a iables we e hypo hesised o bea a ela ionship wi h he
h ee dependen a iables in he s udy (i.e., in elligibili y, comp ehensibili y and o eign-ac-
cen edness).
a) English le el (RQ 1)
b) p e-uni e si y exposu e o CLIL (RQ 2)
c) uni e si y exposu e o EMI (RQ 2)
d) s aying ab oad (RQ 2)
The da a was based on he speake cha ac e is ics ex ac ed om he ques ionnai es
accompanying he eco dings. All s a is ics we e conduc ed using IBM SPSS e sion 26.0
(IBM Co p., 2020).
3. esul s
The indings o he s a is ical es s a e p esen ed in his sec ion o e e y independen
a iable: English le el, CLIL and EMI ins uc ion, and s aying ab oad. The sco es used in
he s a is ical es s co espond o he means calcula ed by a e aging all in elligibili y, com-
196
Po a Lingua um Issue 41, Janua y 2024
p ehensibili y and o eign-accen edness sco es ob ained om all he lis ene s who e alua ed
each speake . The desc ip i e s a is ics ega ding he COM, FA and INT alues in each
g oup a e p esen ed in ables. The co ela ion and pai wise compa ison e ec sizes we e
in e p e ed sensu Cohen (1988), while he ANOVA e ec sizes we e calcula ed ollowing
Tomczak and Tomczak (2014).
3.1. English le el
The le el o English compe ence as sel - epo ed by 50 speake s anged om a CEFR
B2 h ough C1 o C22. A one-way ANOVA was conduc ed o check whe he INT, COM
and FA di e ed ac oss le els. The es demons a ed he e was a signi ican di e ence in
he o e all comp ehensibili y depending on he speake s’ le el wi h a la ge e ec size3, F(2,
47) = 4.345, p = .019, η2 = .16. Pos -hoc analyses using he HSD Tukey es (p = .014)
showed ha he sco es we e signi ican ly highe o C2 le el s uden s han o B2 s uden s.
Howe e , he e was no signi ican di e ence (p = .609) ac oss le els B2 and C1 o C1 and
C2 (p = .122). Simila ly, o eign-accen edness di e ed signi ican ly amongs speake s a
di e en le els, he e ec size o le el being la ge, F(2, 47) = 6.950, p = .002, η2 = .228.
Again, pos -hoc HSD Tukey es esul s e ealed a s a is ically signi ican di e ence (p =
.002) only be ween le els B2 and C2 bu no be ween C1 and C2 (p = .078) o B2 and C1
(p = .260). Unlike comp ehensibili y and o eign-accen edness, in elligibili y did no di e
ac oss le els, F(2, 47) = 1.284, p = .286.
Table 3. COM, FA and INT by le el: desc ip i e s a is ics
COM FA INT
B2 M = 3.87, SD = .74, M = 4.4, SD = .80 M = 21.18, SD = 3.25
C1 M = 4.09, SD = .62, M = 3.98, SD = .86 M = 21.34, SD = 2.8
C2 M = 4.71, SD = .73 M = 3.18, SD = .81 M = 23.07, SD = 2.13
Apa om he a o emen ioned e ec sizes, non-pa ame ic co ela ions we e also cal-
cula ed be ween le el and he h ee dependen a iables. As can be seen in Table 4, le el
was ound o be posi i ely co ela ed wi h COM ( s = .348, p = .013) and nega i ely wi h
FA ( s = -.453, p = 001), bu no wi h INT ( s = .152, p = .293). Jus like in he case o
e ec sizes, he co ela ion is s onge be ween le el and o eign-accen edness han be ween
le el and comp ehensibili y.
Table 4. Non-pa ame ic co ela ions: le el and dependen a iables
COM FA INT
Le el s = .348*4 , p = .013 s = -.454**, p = .001 s =.152, p = .293
2 As men ioned ea lie in he me hodology sec ion, only hose speake s who epo ed o be holde s o language
ce i ica es, i.e., in he es s including he le el a iable, he sample size was 50, no 60.
3 η2= .01 smaLL, η2 = .06 medium, η2 = .16 La ge (Tomczak & Tomczak 2014).
4 * is used o deno e p- alues below .05, while ** is used o alues below .01.
197
Ma eusz Pie aszek Associa ing speake a iables wi h English p onuncia ion...
3.2. Bilingual educa ion: CLIL and EMI p og ammes
Rega ding he exposu e o he speake sample o CLIL a p e ious educa ion s ages, no
s a is ically signi ican di e ences we e ound in o e all dependen a iable means be ween
he g oups o s uden s p e iously exposed o CLIL (CLIL g oup) and hose wi h no such
exposu e (non-CLIL g oup). The analyses did no indica e s a is ically signi ican di e ences
in he COM o he s uden s in bo h g oups, (58) = -.84 , p = .402. Likewise, FA did no
di e signi ican ly be ween he wo g oups, (58) = 1.47, p = .146, and nei he did INT,
(58) = -.17, p = .863.
Table 5. COM, FA and INT by medium o ins uc ion: desc ip i e s a is ics
COM FA INT
CLIL M = 4.14, SD = .70, M = 4.01, SD = .91 M = 21.73, SD = 3.16
non-CLIL M = 3.97, SD = .57, M = 4.35, SD = .74 M = 21.59, SD = 2.41
EMI M = 4.01, SD = .75 M = 4.01, SD = .75 M = 21.86, SD = 3.01
non-EMI M = 4.11, SD = .71 M = 4.12, SD = .92 M = 22.36, SD = 3.02
Simila ly, none o he di e ences in o e all INT, COM and FA means be ween he
EMI and non-EMI s uden g oups we e ound o be s a is ically signi ican . An independen
sample - es did no show di e ences in he comp ehensibili y o he sampled EMI s uden s
and non-EMI s uden s, (58) = .260 , p = .796. Mo eo e , no signi ican di e ences we e
de ec ed in an independen sample - es be ween he FA sco es o in bo h g oups, (58) =
-.076 , p = .940. Finally, he INT o he EMI sample was no s a is ically di e en om
ha o he non-EMI sample, (58) = -.493, p = .624.
3.3. S aying ab oad
The s uden s’ expe ience ab oad was also hypo hesised o be po en ially associa ed wi h
hei phonological pe o mance. Howe e , no s a is ically signi ican di e ences be ween g oups
we e suppo ed by he esul s o - es s. COM p o ed no o di e signi ican ly be ween
hose wi h expe ience ab oad (g oup A) and hose wi hou i (g oup B), (58) = -.691, p =
.493. Nei he did he le el o FA a y signi ican ly be ween g oups, (58) = .351, p = .727.
Finally, INT le els we e no signi ican ly di e en , (58) = -.074, p = .941.
Table 6. COM, FA and INT by leng h o s aying ab oad: desc ip i e s a is ics
COM FA INT
g oup A (YES) M = 4.12, SD = .69 M = 4.11, SD = .81 M = 21.70, SD = 2.7,
g oup B (NO) M = 3.97, SD = .86 M = 4.20, SD = 1.05, M = 21.64, SD = 3.53
Howe e , despi e he lack o g oup di e ences, ce ain co ela ions we e de ec ed as
desc ibed below. The ime spen ab oad was posi i ely co ela ed wi h COM ( = .402, p =
.001) and nega i ely co ela ed wi h FA ( = -.315, p = .001). Ne e heless, he INT le el
204
Po a Lingua um Issue 41, Janua y 2024
Galla do del Pue o, F., & Ma ínez Ad ián, M. (2013). ¿Es más e ec i o el ap endizaje de la
lengua ex anje a en un con ex o AICLE? Resul ados de la in es igación en España. Pad es
y Maes os, 263, 25-28.
Galla do del Pue o, F., Gómez Lacabex, E., & Ga cía Lecumbe i, M. L. (2009). Tes ing he
e ec i eness o con en and language in eg a ed lea ning: he assessmen o English p o-
nuncia ion. In Y. Ruiz de Za obe, & R. M. Jiménez Ca alán (Eds.), Con en and language
in eg a ed lea ning. E idence om esea ch om Eu ope (pp. 63-80). Mul ilingual Ma e s.
Gal án, F. (2010). Po qué a los españoles se nos da mal el inglés. El País (9 July 2010). h ps://
elpais.com/dia io/2010/07/19/opinion/1279490411_850215.h ml
Gilbe , J. B. (2010). P onuncia ion as o phan: wha can be done? Speak Ou ! IATEFL P onun-
cia ion Special Newsle e , 43, 3-7.
Gómez Lacabex, E., & Galla do del Pue o, F. (2021). P onuncia ion in EMI. A p elimia y s udy
o Spanish uni e si y s uden s’ in elligibili y and comp ehensibili y. In D. Lasagabas e ,
& A. Doiz (Eds.), Global pe spec i es on language aspec s and eache de elopmen in
English medium Ins uc ion (pp. 126-144). Rou ledge, Taylo & F ancis.
Gooskens, C. (2013). Expe imen al me hods o measu ing in elligibili y o closely ela ed lan-
guage a ie ies. In R. Bayley, R. Came on & C. Lucas (Eds.), The Ox o d Handbook o
Sociolinguis ics (pp. 195-213). Ox o d Uni e si y P ess.
Haugen, E. (1966). Semicommunica ion: The language gap in Scandina ia. Sociological Inqui y
36, 280–97.
Hellekjæ , G. O. (2010). Language ma e s: assessing lec u e comp ehension in No wegian Eng-
lish-medium highe educa ion, Language use and language lea ning in CLIL class ooms
(pp. 233-258). John Benjamins.
IBM Co p. (2020). IBM SPSS [compu e so wa e]. IBM.
Jenkins, J. (2000). The phonology o English as an in e na ional language. Ox o d Uni e si y P ess.
Jeong, H. J., Tho én, B., & O hman, J. (2017). Mu ual in elligibili y o Malay- and Swedish-accen ed
English: an expe imen al s udy. Indonesian Jou nal o Applied Linguis ics, 7(1), 43-53.
Jo o, A., Nagase, Y., & Funa su, S. (2007). E ec o VOT on in elligibili y o English oiceless
s ops p oduced by na i e speake s o Japanese [Con e ence P esen a ion]. PTLC2007:
Phone ics Teaching and Lea ning Con en ion, London, Uni ed Kingdom. h p://www.phon.
ucl.ac.uk/p lc/p oceedings/p lcpape _31e.pd
Juan-Ga au, M. (2014). O al accu acy g ow h a e o mal ins uc ion and s udy ab oad: onse
le el, con ac ac o s and long- e m e ec s. In Pé ez-Vidal, Ca men (ed.), Language ac-
quisi ion in s udy ab oad and o mal ins uc ion con ex s (pp. 87-109). John Benjamins.
Kach u, B. B. (1985). S anda ds, codi ica ion and sociolinguis ic ealism: The English language
in he ou e ci cle. In R. Qui k, & H. Widdowson (Eds.), English in he wo ld: eaching
and lea ning he language and he li e a u e. Camb idge Uni e si y P ess.
Kang, O., Thomson, R. I., & Mo an, M. (2018). Empi ical app oaches o measu ing he in elli-
gibili y o di e en a ie ies o English in p edic ing lis ene comp ehension. Language
Lea ning, 68(1), 115-146. doi:10.1111/lang.12270
La Razón. (2019). Los españoles son los eu opeos que peo hablan inglés. La Razón (13 July
2019). h ps://www.la azon.es/ amilia/los-espanoles-son-los-eu opeos-que-peo -hablan-in-
gles-PO24191608/
Lasagabas e , D. (2008). Fo eign language compe ence in con en and language in eg a ed cou ses.
The Open Applied Linguis ics Jou nal, 1(1), 30-41.

205
Ma eusz Pie aszek Associa ing speake a iables wi h English p onuncia ion...
Le is, J. M. (2018). In elligibili y, o al communica ion, and he eaching o p onuncia ion. Cam-
b idge Uni e si y P ess.
LimeSu ey Gmbh. (2019). LimeSu ey: An Open Sou ce Su ey Tool. h ps:// www.limesu ey.o g.
Lippi-G een, R. (1994). Accen , s anda d language ideology, and disc imina o y p e ex in he
cou s. Language in Socie y, 23(2), 163-198.
Meie ko d, C. (2004). Syn ac ic a ia ion in in e ac ions ac oss in e na ional Englishes. English
Wo ld-Wide, 25(1), 109-132.
Me ino, J. A., & Lasagabas e , D. (2018). The e ec o con en and language in eg a ed lea ning
p og ammes’ in ensi y on English p o iciency: A longi udinal s udy. In e na ional Jou nal
o Applied Linguis ics, 28(1), 18-30.
Minema su, N., Okabe, K., Ogaki, K., & Hi ose, K. (2011). Measu emen o objec i e in elli-
gibili y o Japanese accen ed English using ERJ (English Read by Japanese) Da abase,
In e speech P oceedings, Flo ence, I aly. (pp. 1481-1484).
Moo e, I., To ge son, C., & Beckmann, N. (2021). Sys ema ic e iew measu ing he e icacy o
s udy ab oad in unde g adua e language lea ne s on linguis ic p o iciency gains. Re iew
o Educa ion, 9(3).
Muñoz, C., & Llanes, À. (2014). S udy ab oad and changes in deg ee o o eign accen in child en
and adul s. The Mode n Language Jou nal, 98(1), 432-449. h ps://doi:10.1111/j.1540-
4781.2014.12059.x
O’Neal, G. (2015). In e ac ional in elligibili y: he ela ionship be ween consonan modi ica ion
and p onuncia ion in elligibili y in English as a lingua anca in Japan. Asian Englishes,
17(3), 222–239.
O o, A., & Es ada, J. L. (2019). Towa ds an unde s anding o CLIL assessmen p ac ices in a
Eu opean con ex : main assessmen ools and he ole o language in con en subjec s.
CLIL Jou nal o Inno a ion and Resea ch in Plu ilingual and Plu icul u al Educa ion,
2(1), 31-42. h ps:// doi:10.5565/ e /clil.11
Pé ez Cañado, M. L., & Lancas e , N. K. (2017). The e ec s o CLIL on o al comp ehension and
p oduc ion: a longi udinal case s udy. Language, Cul u e, and Cu iculum, 30(3), 300-316.
Pé ez-Cañado, M. L. (2018). CLIL and educa ional le el: A longi udinal s udy on he impac o
CLIL on language ou comes. Po a Lingua um, 29, 51-70.
Poljak, L. (2019). T ends in accen edness and comp ehensibili y esea ch wi h espec o L2
speech a ings. SFU Educa ional Re iew, 12(2), 36-63.
Rallo Fab a, L., & Jacob, K. (2015). Does CLIL enhance o al skills? Fluency and p onuncia ion e o s
by Spanish-Ca alan lea ne s o English. In M. Juan-Ga au, & M. Salaza Nogue a (Eds.). Con-
en -based language lea ning in mul ilingual educa ional en i onmen s (pp. 163-177). Sp inge .
Rallo Fab a, L., & Juan-Ga au, M. (2010). In elligibili y and o eign accen edness in a con-
ex -and-language-in eg a ed-lea ning (CLIL) se ing [Con e ence Pape ]. 6 h In e na ional
Symposium on he Acquisi ion o Second Language Speech, New Sounds 2010 (pp. 373-378).
Ramos Ga cía, A. M. (2013). Highe educa ion bilingual p og ammes in Spain. Po a Lingua um:
Re is a In e nacional de Didác ica de las Lenguas Ex anje as, 19(19), 101-112. h ps://
doi:10.30827/Digibug.29632
Ramos Ga cía, A. M., & Pa ón Vázquez, V. (2018). The linguis ic in e na ionaliza ion o highe
educa ion: a s udy on he p esence o language policies and bilingual s udies in Spanish
uni e si ies. Po a Lingua um: Re is a In e nacional de Didác ica de las Lenguas Ex an-
je as, Monog á ico III, 31-46.
206
Po a Lingua um Issue 41, Janua y 2024
Rich e , K. (2015). English-medium ins uc ion and p onuncia ion a longi udinal case s udy o
Aus ian business s uden s a he Uni e si y o Applied Sciences Vienna. [PhD Disse a-
ion]. Uni e si y o Vienna.
Ruiz de Za obe, Y. (2007). CLIL in a bilingual communi y: simila i ies and di e ences. Vienna
English Wo king Pape s, 16(3), 47-52.
Ruiz de Za obe, Y. (2011). Which language compe encies bene i om CLIL? An insigh in o
applied linguis ics esea ch. In Y. Ruiz de Za obe, J. M. Sie a & F. Galla do del Pue o
(Eds.). Con en and o eign language in eg a ed lea ning: con ibu ions o mul iliingualism
in Eu opean con ex s (pp. 129-154). Pe e Lang.
Sai o, K., T o imo ich, P., & Isaacs, T. (2016). Second language speech p oduc ion: In es iga ing
linguis ic co ela es o comp ehensibili y and accen edness o lea ne s a di e en abili y
le els. Applied Psycholinguis ics 37(2), 217-240.
Smi h, L. E., & Nelson, C. L. (1985). In e na ional in elligibili y o English: di ec ions and e-
sou ces. Wo ld Englishes, 4(3), 333-342.
Tomczak, M., & Tomczak, E. (2014). The need o epo e ec size es ima es e isi ed. An o e iew
o some ecommended measu es o e ec size. T ends in Spo Sciences, 1(21), 19-25.
T o imo ich, P., & Isaacs, T. (2012). Disen angling accen om comp ehensibili y. Bilingualism,
15(4), 905-916.
Volín, J., Ska ni zl, R., & Hende son, A. (2018). Pe cep ual impac o o eign-accen ed speech.
In J. Volín, & R. Ska ni zl (Eds.), The p onuncia ion o English by speake s o o he
languages (pp. 73-94). Camb idge Schola s Publishing.
Walke , R. (2010). Teaching he p onuncia ion o English as a lingua anca. Ox o d Uni e si y
P ess.
Wang, H. (2007). English as a lingua anca: Mu ual in elligibili y o Chinese, Du ch and Ame -
ican speake s o English. LOT.
7. AppendIx
Seman ically Unp edic able Sen ences (SUS) (adap ed om Wang, 2007)
1. The s a e sang by he long week.
2. The man lay h ough he wide wa .
3. The day hung o he g ea nigh .
4. The yea smiled h ough he young head.
5. The ime b ea hes wi h he high side.
6. The way an o he ho oom.
7. The hing hung om he small line.
8. The g ass lied on he blue nigh .
9. The school s ayed o he new ube.
10. The hand ell o he high o m.
11. The eal ield made he o e.
12. The whi e home go he a .
13. The clea iend b ough he g ound.
14. The black an held he zone.
15. The whole mon h bans he ai .
16. The hin job go he oad.
17. The poo e hi he ax.
18. The sho ield said he you h.
19. The ull home ook he e m.
20. The whi e sense a e he zoo.
207
Ma eusz Pie aszek Associa ing speake a iables wi h English p onuncia ion...
21. Use he game o he hai .
22. Ask he ial and he ee.
23. Lea e he spo and he hough .
24. Call he club and he g ow h.
25. Tu n he lo e o he es .
26. Spend he sale o he nose.
27. S a he age o he p ice.
28. Show he plan o he sound.
29. S op he s ock and he lis .
30. Li e he spo and he und.
31. When does he cha ge like he la e plane?
32. Whe e does he band sell he low se ?
33. Why does he cell like he deep leng h?
34. When does he gun like he deep bed?
35. Why does he ange wa ch he ine es ?
36. When does he sign lead he ed oo ?
37. How does he oice plan he cold ea ?
38. How does he chance send he deep oo ?
39. Why does he juice bea he jump ade?
40. How does he cloud wa ch he low ex ?
Elici a ion pa ag aph (o iginally designed o his esea ch)
The sun was ising slowly and he bi ds we e singing. The iew om he op o he hill
was amazing. Susan u ned ound and s epped hea ily in o he ki chen. She ook a plas ic
cup and illed i wi h juice. Then, she sa down on a chai and s a ed hinking wha ood
she would cook o lunch. Suddenly, he dog, Zoe, jumped on o he , spilling he d ink. She
h ew he emp y cup away in o a small was e bin. She hadn’ been eeling good o yea s.
He job as a uni e si y nu se didn’ gi e he pleasu e and she ha ed s a ing in he ea ly
mo ning. A he age, i wasn’ easy o deal wi h people. As she couldn’ hea well, she was
ge ing used o being on he own. She decided o skip wo k, s ay in and enjoy he i s day
o sp ing. Then she emembe ed i was Sa u day anyway.