P ac ice Pape
Recommended ci a ion: Fishe , R., Nai , D. J., Ma io, J. V., & A yampa, S. (2025).
Re iewing Re iews: Using Mul is age Pee Re iews o P o ide Feedback and
Imp o e S uden Lea ning. In Kangaslampi, R., Langie, G., Jä inen, H.-M., &
Nagy, B. (Eds.), SEFI 53 d Annual Con e ence. Eu opean Socie y o Enginee ing
Educa ion (SEFI), Tampe e, Finland. DOI: 10.5281/zenodo.17631505.
This Con e ence Pape is b ough o you o open access by he 53 d Annual Con e ence
o he Eu opean Socie y o Enginee ing Educa ion (SEFI) a Tampe e Uni e si y in
Tampe e, Finland. This wo k is licensed unde a C ea i e Commons
A ibu ion-NonComme cial-Sha e Alike 4.0 In e na ional License.
Re iewing Re iews: Using mul is age pee e iews o p o ide
eedback and imp o e s uden lea ning
R.M. Fishe a, D. J. Nai a,1, J. Videla Ma ió a, S. A yampa a
a School o Ci il and En i onmen al Enginee ing, UNSW Sydney
Con e ence Key A eas: Enginee ing skills, p o essional skills, and ans e sal skills
Keywo ds: Feedback, Re lec ion, AI analysis, Me a- e iews, Pee Re iew
ABSTRACT
A undamen al skill in enginee ing p ac ice is p o iding and ecei ing eedback.
Howe e , his is no always explici ly inco po a ed in cu iculums. In his s udy, ‘me a-
e iew’ asks we e inco po a ed in o ou cou ses wi hin ci il and en i onmen al
enginee ing. While he co e s ages o he me a- e iew p ocess we e consis en (pee
e iew– e iew o eedback – e lec ion and co ec ion), a ia ions included di e en
coho s, discipline a eas, assignmen ypes, eedback ypes and ma king s uc u es.
The p ocess was e alua ed using he eedback i sel , s uden pe cep ions ia su eys
and eache e lec ions. Thema ic analysis o he eedback using gene a i e AI
iden i ied key o e all hemes. Feedback mos app ecia ed by pee s was ha which
was speci ic, ac ionable, cons uc i e and had a p o essional o suppo i e one. The
in luence o expe ise o hose p o iding eedback showed highe echnical
engagemen and c i ique om hose mo e amilia wi h he opics, and mo e s uc u al
and p esen a ion/audience based sugges ions om hose less amilia . Howe e , bo h
p o ided use ul cons uc i e eedback, and di e ences in assigned g ades we e no
signi ican . Pos ac i i y e alua ion showed signi ican imp o emen s in assignmen
g ades, wi h posi i e s uden pe cep ions. While acili a ion o he second laye o he
e iew p ocess was challenging, ou indings show ha ‘me a- e iew’ ype asks can
help suppo s uden s in lea ning wha e ec i e eedback is.
1 Co esponding Au ho
D.J Nai
d.jayakuma nai @unsw.edu.au
1 INTRODUCTION
Feedback suppo s s uden s o deepen hei unde s anding, e lec on hei
pe o mance, and imp o e. In he educa ion con ex , eedback can be de ined as he
in o ma ion p o ided by an agen (e.g. eache , pee , sel ) ega ding aspec s o one's
pe o mance (Ha ie, 2007).O e he las wo decades, he e has been inc easing
in e es in unde s anding how e ec i e eedback ope a es, speci ically in highe
educa ion con ex s.
P e ious esea ch has iden i ied key cha ac e is ics o e ec i e eedback in highe
educa ion. Li and De Luca (2012) ound ha s uden s p io i ise imely eedback as a
c i ical componen . Dawson e al. (2019) p o ided empi ical e idence om wo
Aus alian uni e si ies showing ha s uden s alue high-quali y eedback commen s
(84%) ha a e usable, de ailed, speci ic, and pe sonalised, enabling hem o ake
ac ionable s eps o imp o emen . Howe e , hei s udy e ealed a gap on wha
makes eedback e ec i e: while bo h s uden s and s a de ined eedback's pu pose
as imp o emen , s a mos ly ocused on design elemen s like iming and modali ies
(53%) whe eas s uden s p io i ised he quali y o he commen s.
To add ess hese challenges sys ema ically, Ca less and Boud (2018) in oduced
Feedback Li e acy, de ining i as a li elong capabili y connec ed o p o essional skills.
Recen empi ical alida ion by Woi e al. (2025) showed ha eedback li e acy
comp ises wo key dimensions: eedback a i udes (disposi ions owa d eedback)
and eedback p ac ices (cogni i e and beha iou al engagemen wi h eedback),
suppo ing he need o s uc u ed ac i i ies ha de elop bo h componen s.
A ecu en challenge in highe educa ion, pa icula ly in la ge enginee ing cou ses, is
p o iding de ailed, imely, indi idual eedback. Pee eedback o e s a scalable
solu ion: esea ch shows i can be as e ec i e as ins uc o eedback o imp o ing
s uden unde s anding (P ice e al., 2013). A me a-analysis by Huisman e al. (2019)
examining 24 s udies ound pee eedback signi ican ly imp o es academic w i ing
compa ed o no eedback (g = 0.91) and pe o ms simila ly o eache eedback.
Howe e , e ec i e pee eedback equi es aining, as s uden s o en lack clea
unde s anding o e alua ion c i e ia, leading o supe icial o uncons uc i e eedback
(Philippakos, 2017). The p ocess o unde aking pee assessmen also p omo es he
e-e alua ion o he s uden s' own wo k (in e nal eedback) as well as unde aking
c i ical hinking, and p oblem sol ing (Low e al., 2022). S ill, hese concep s a e no
always included in cu iculum o emain pa o implici lea ning ou comes. This gap
equi es a mo e a ge ed p ocess whe e c i ical hinking, e iewing, and p o iding
eedback a e di ec ly p ac iced h ough s uc u ed e lec ion on eedback quali y and
e ec i eness.
Me a- e iews ( ha is, he p ocess whe e s uden s e lec on and e iew eedback
p o ided by pee s – Figu e 1) p o ide s uc u ed p ac ice by ha ing s uden s assess
he quali y and use ulness o pee eedback, de eloping he analy ical and e lec i e
capabili ies essen ial o Enginee ing p o essional p ac ice.
In his esea ch pape , s aged e iews, ‘me a- e iews’ we e inco po a ed in o ou
di e en cou ses wi hin Ci il and En i onmen al Enginee ing a UNSW Sydney,
Aus alia. Ci il and En i onmen al Enginee ing was selec ed due o he esea ch
eam’s disciplina y expe ise and equi emen o g adua es o de elop p o essional
compe encies. While he co e s ages o he p ocess we e consis en (pee e iew–
e iew o eedback – e lec ion and co ec ion), a ia ions included di e en coho s,
discipline a eas, assignmen ypes, eedback ypes and ma king s uc u es.
Figu e 1. Schema ic o he me a- e iew app oach. G een a ow indica es pee - e iew s age,
blue a ow indica es me a- e iew s age, and pink a ows show po en ial lea ning bene i s
gained by s uden s.
2 METHODOLOGY
2.1 Task Se -up and Adminis a ion
The s aged pee e iew p ocess comp ised h ee sequen ial s eps (Figu e 2). This
p ocess was acili a ed using he Moodle Wo kshop ool (Moodle, 2024), a widely used
pla o m o online pee e iew ac i i ies. The p ocess comp ised:
1. Submission: S uden s submi hei assessmen o Moodle Wo kshop (Module 1).
2. Pee - e iew s age: S uden s would hen be pai ed wi h one o wo s uden s o
p o ide eedback. Rub ics we e p o ided o s uden s showing clea guidelines
abou how o e alua e each sec ion o he submission.
3. Me a- e iew: A e s uden s would ecei e hei eedback, hey would p o ide
eedback on hei eedback (i.e. me a eedback) in Module 2A and B. Ano he ub ic
was p o ided o he s uden s o cla i y he expec ed c i e ia o he me a eedback.
This app oach was implemen ed ac oss ou cou ses (Table 1) wi h consis en ub ic
s uc u es (shown in APPENDIX A).
Figu e 2. Adminis a ion o he s aged pee e iew p ocess using he Moodle Wo kshop
ac i i y.
Table 1. Summa y o cou ses inco po a ing s aged pee eedback ac i i ies
Code
Type
Size
Submission
ype
Fi s le el
Second
le el
Thi d le el /
Re lec ion
UGRD-1
Fou h
yea co e
caps one
cou se.
34
3-page si e
isi epo
de ailing
cu en
in as uc u e
sys ems, each
s uden
ocused on one
o 4 sys ems
2 pee s (one
same opic, one
di e en opic)
indi idually
p o ide w i en
commen s and
summa i e g ade
(5%) acco ding o
3x3 ub ic
Feedback
quali y o
pee s
e alua ed wi h
w i en
commen and
summa i e
g ades (5%)
acco ding o a
3x3 ub ic
Re lec ion
encou aged in
inal g oup
epo and
in e iew.
Op ional su ey
UGRD-2
Fou h
yea co e
caps one
cou se.
Repea
54
PGRD
Elec i e
cou se o
PG and
Fou h
yea
elec i e
cou se o
UG
40
P elimina y
P ojec Design
B ie ela ed o
a anspo
planning and
managemen
opic
2 pee s
indi idually
p o ide w i en
commen s and
summa i e g ade
(5%) acco ding o
3x3 ub ic
Feedback
quali y o
pee s
e alua ed wi h
w i en
commen and
summa i e
g ades (5%)
acco ding o a
3x3 ub ic
Re lec ion
encou aged in
inal g oup
epo ,
Resubmission
o he
p elimina y
epo and
in e iew.
Op ional su ey
UGRD+
PGRD
Fou h
yea
elec i e
cou se,
also
aken by
pos g ad
s uden s
127
D a
p esen a ion on
he use o
sus ainabili y
assessmen
me hods, each
s uden
ocused on one
o 9 hemes
2 pee s (one
same opic, one
di e en opic)
indi idually
p o ide w i en
commen s and
indica i e g ades
( o ma i e)
acco ding o 3x3
ub ic
Feedback
quali y o
pee s
e alua ed wi h
w i en
commen and
summa i e
g ades (5%)
acco ding o a
4x3 ub ic
No o malized
e lec ion
ac i i y.
Op ional su ey
2.2 Da a collec ion me hods
Da a collec ion included s uden pe cep ions ia su eys, cou se con en om he
pee e iew ac i i ies, and eache e lec ions o he p ocess. Pe mission o su ey
s uden s as well as o use anonymised cou se da a (w i en commen s and g ades)
was acqui ed (Human E hics iRECS6901). Pos cou se su eys we e used o ga he
s uden esponses ega ding he e ec i eness o he ac i i ies in imp o ing e iewing
skills and imp o ing s uden unde s anding o assessmen equi emen s. Su eys
we e dis ibu ed o s uden s ia email ollowing cou se and/o ac i i y comple ion,
howe e esponse a es we e low (11 ou o 250 s uden s).
2.3 Mixed-me hod analysis
A mixed-me hods app oach was adop ed o encompass bo h he quan i a i e and
quali a i e in o ma ion ga he ed. Quan i a i e analysis examined s uden -assigned
ma ks o i s ( eedback) and second-le el (me a- eedback) e iews. The ollowing
hypo heses we e es ed: a) In cou ses whe e s uden s p o ided eedback on same
and di e en opics (UGRD-1, UGRD-2, UGRD+PGRD), do he ma ks gi en di e ?
b) Do s uden s who ecei ed highe ma ks subsequen ly gi e highe me a eedback
a ings? Fo quali a i e analysis, he w i en eedback was analysed using ini ial
quali a i e da a ca ego iza ion and explo a ion o opics using a la ge language model
(LLM)-based app oach as ou lined by Hayes (2025). This LLM amewo k p o ided a
sys ema ic and in e ac i e ounda ion o gene a ing labels and iden i ying illus a i e
quo es (p omp examples p o ided in APPENDIX B). Using he p ocessed
in o ma ion, esea ch ques ions examined included: how impo an is s uden
expe ise in p o iding use ul eedback, how do s uden s espond o eedback one,
and e ec i eness based on pos ask e lec ions.
3 RESULTS
3.1 Quali a i e analysis o eedback commen s
Recu en eedback hemes in he w i en commen s om all cou ses we e iden i ied
using he LLM based app oach, he ollowing key hemes we e epo ed.
Cla i y and Speci ici y: Ac oss all cou ses, s uden s emphasised he necessi y o
eedback ha was clea , speci ic, and unambiguous. In UGRD-1 and UGRD-2,
epo s explici ly no ed eedback ha lacked speci ici y o cla i y led o con usion and
limi ed i s u ili y. Simila ly, in PGRD, s uden s epo ed ha ague connec ions
be ween esea ch ques ions and me hodologies weakened eedback applicabili y.
P ac icali y and Ac ionabili y: S uden s epea edly alued eedback o e ing
p ac ical s eps owa d imp o emen . Acco ding o he UGRD+PGRD and UGRD
ac ionable sugges ions signi ican ly enhanced s uden comp ehension o assessmen
equi emen s and helped in e ining hei submissions. S uden s speci ically p aised
ins ances whe e pee eedback ou lined clea pa hways o ac ions.
Balance and Cons uc i eness: Many s uden s app ecia ed eedback ha
inco po a ed ecommenda ions and cons uc i e c i icism. In UGRD-2 he e we e
ins ances whe e o e ly nega i e o posi i e eedback was no aligned wi h he
cons uc i e ecommenda ions, diminishing i s use ulness.
Technical De ails: Technical speci ici y, including quan i a i e da a and p ecise
me hodological de ails, was equen ly highligh ed as c i ical o he pe cei ed
c edibili y and use ulness o pee eedback. Fo ins ance, PGRD no ed s uden
in e es in de ailed echnical explana ions ela ed o queuing heo y models and da a
collec ion p ocesses. While o UGRD-1 and UGRD-2, s uden s we e o en
sugges ed o inco po a e igu es, maps, and diag ams o enhance hei epo s.
Focusing on he me a- e iew eedback s uden s gene ally epo ed imp o ed cla i y
ega ding assessmen c i e ia and g ea e con idence in unde s anding assignmen
expec a ions. Cou se speci ic de ails showed:
- UGRD-1 and UGRD-2: S uden s pa icula ly alued he dep h o eedback
ega ding comp ehensi e si e analyses and in as uc u e assessmen s,
con ex ualising he onsi e obse a ions and linking hem o assignmen equi emen s.
- UGRD+PGRD: S uden s emphasised he impo ance o clea ly jus i ied
me hodological choices and explici a icula ion o da a bounda ies and inpu s,
highligh ing he alue o anspa ency in sus ainabili y assessmen me hodologies.
- PGRD: The applicabili y and ele ance o heo e ical concep s such as queuing
heo y eme ged as a c i ical ac o . S uden s exp essed app ecia ion o eedback
clea ly linked o p ac ical implica ions and ealis ic scena ios, unde sco ing he
impo ance o in eg a ing heo e ical knowledge wi h eal-wo ld applica ions.
3.2 T ends in G ades
Analysis o pee e iew and me a eedback ma ks es ed wo hypo heses ega ding
he opic ma ching e ec s and co ela ion be ween ecei ed ma ks and subsequen
gi en ma ks on he me a eedback p ocess. Desc ip i e s a is ics o pee e iew
eedback and me a eedback ma ks o he ou cou ses a e p esen ed in Table 2.
Independen sample - es s examined whe he s uden s ma ked di e en ly when
e iewing same e sus di e en opics ac oss h ee cou ses (UGRD-1, UGRD-2,
UGRD+PGRD). No signi ican di e ences we e ound in any cou se o ei he pee
e iew ma ks (p- alues: 0.910, 0.400, 0.431) o me a eedback ma ks (p- alues:
0.732, 0.720, 0.810). This sugges s ha opic ma ching did no in luence s uden
ma king beha iou .
Table 2: Desc ip i e s a is ics o Feedback Ma ks and Me a Feedback Ma ks by cou se.
No e: cou ses in which he asks we e o ma i e a e unde lined.
Cou se
Pee Re iew Ma ks
Me a Feedback Ma ks
UGRD-1
Mean: 80.00, S d: 14.15
Mean: 85.69, S d: 18.65
UGRD-2
Mean: 79.62, S d: 11.60
Mean: 79.02, S d: 20.12
PGRD
Mean: 74.39, S d: 23.41
Mean: 79.74, S d: 21.40
UGRD+PGRD
Mean: 74.67, S d: 17.54
Mean: 83.95, S d: 23.79
Co ela ion analysis es ed whe he s uden s who ecei ed highe submission ma ks
subsequen ly ga e highe ma ks on he me a eedback p ocess. Resul s a ied
ac oss cou ses: UGRD+PGRD showed a small bu signi ican posi i e co ela ion ( =
0.174, p = 0.021), while UGRD-1 ( = 0.219, p = 0.073), UGRD-2 ( = 0.027, p =
0.787), and PGRD ( = 0.046, p = 0.689) showed no signi ican ela ionships. The
hi d hypo hesis ecei ed limi ed suppo , being suppo ed in only one cou se.
No ably, he signi ican co ela ion was ound only in he o ma i e assessmen
cou se, sugges ing ha g ading s akes may in luence s uden eedback beha iou .
4 DISCUSSION
4.1 How impo an is opic amilia i y in he quali y o eedback p o ided?
Fo subjec s UGRD-1, UGRD-2 and UGRD+PGRD s uden s we e assigned in o
di e en opics, and hei wo k was e iewed by pee s who comple ed he
assignmen on he same opic, o a di e en opic. In UGRD-1 and UGRD-2, s uden s
w o e si e epo s on one o ou in as uc u e sys ems: wa e , was e, ene gy o
anspo . In UGRD+PGRD, s uden s used 9 di e en sus ainabili y assessmen
me hods o ack consump ion o lows o esou ces in hei households (e.g. ca bon
oo p in , lows o single use plas ic). The e was no s a is ically signi ican di e ence
in ma ks assigned a ei he e iew s age be ween hose in Same s Di e en g oups.
The quali a i e analysis o pee eedback commen s e ealed di e ences be ween
same and di e en opic e iewe s, p ima ily in he dep h e sus b ead h o eedback.
Same- opic e iewe s p o ide mo e specialised and echnically de ailed eedback
and we e mo e c i ical a poin ing ou misconcep ions o missing echnical concep s.
In con as , di e en - opic e iewe s o e ed mo e gene al, s uc u e- ocused, and
audience-o ien ed eedback wi h sugges ions ocusing on he gene al s uc u e,
eadabili y, b oade con ex , and communica ion e ec i eness.
Analysis o he me a eedback commen s showed ha same- opic e iewe s had a
highe endency o dispu e/deba e hei ecei ed eedback compa ed o di e en -
opic e iewe s. On he o he hand, eedback ecei ed om di e en - opic e iewe s
was p aised o i s o ganisa ion and cla i y. Some s uden s also exp essed
app ecia ion o insigh s om ou side hei opic a ea. O e all, s uden s om bo h
g oups alued clea , speci ic eedback wi h ac ionable s eps.
These indings sugges ha opic amilia i y can in luence he na u e o pee
eedback, e en i i didn' signi ican ly a ec he ma ks gi en. Domain expe ise le el
can de e mine he sophis ica ion o pee eedback, wi h high-knowledge s uden s
p o iding me acogni i e eedback and low-knowledge s uden s p o iding su ace-
le el ask eedback (Alqassab e al., 2018). Ou indings align wi h he wo k on
domain expe ise and demons a es he educa ional alue o pee e iew ype
ac i i ies in p o iding co ec i e eedback, e en when lacking ce ain expe ise. F om
an educa ional p ac ice pe spec i e, he e is alue in eedback om a ied coho s
u he suppo ing he u ili y o pee e iew ype ac i i ies.
A inal s uden pe cep ion su ey (Thi d le el/Re lec ion on Table 1) was conduc ed,
bu only comple ed by a hand ul o pa icipan s in he PGRD and UGRD+PGRD. I
showed 67% o esponden s sligh ly disag eeing ha i ’s easonable o ha e
eedback om pee s ins ead o an ins uc o . This scep icism owa ds pee eedback,
also iden i ied in o he s udies (Kau man e al., 2010; Kau man & Schunn, 2010),
sugges s ha despi e bene i s, s uden s may s ill doub he legi imacy o pee
eedback compa ed o ins uc o s.
4.2 Cons uc i e o C i ical? Tonal nuances in eedback
Tone o he eedback p o ided eme ged as an impo an elemen du ing he s udy. In
some cou ses (UGRD 2), a numbe o s uden s pe cei ed he eedback as o e ly
ha sh as exempli ied by "You c i icism one needs a li le bi o de uning…”. In o he
cou ses, s uden s epo ed app ecia ing a p o essional and espec ul one when
iden i ying p oblems, wi h a ocus on he wo k i sel a he han pe sonal commen s.
Tone was in oduced as one o he c i e ia agains which eedback was compa ed in
UGRD+PGRD. When s uden s used mo i a ional language in hei eedback, his
encou agemen was no ed and app ecia ed in he me a e iew eedback. Simila ly,
he con as was also ue, as eedback ha only iden i ied p oblems wi hou any
posi i e commen s o encou agemen was iewed nega i ely. Simila indings ha e
been ound in a eas such as pee - e iew p ocess in academic jou nals (Chong & Lin,
2024). Feedback one ha is p o essional, has a espec ul language and ocuses on
he wo k was consis en ly alued, while hos ile, demeaning, o o e ly c i ical one
hinde ed he ecipien 's abili y o engage cons uc i ely wi h he eedback.
Ano he issue encoun e ed was when he e was a disconnec be ween he w i en
commen s and he g ades assigned on he ub ic, wi h s uden s no ing con usion
when posi i e-sounding eedback was accompanied by lowe g ades.
4.3 E ec i eness o eedback in imp o ing assessmen ou comes
The inal s uden pe cep ion su ey included nine ques ions abou he e ec i eness
o he pee e iew ac i i y using Like scales alongside a inal open esponse ield.
Commen s highligh ed a iabili y in pee eedback quali y. Some s uden s exp essed
conce ns abou ecei ing eedback ha was occasionally supe icial o pe unc o y,
indica ing inconsis encies in pee engagemen . One s uden explici ly ema ked,
"Task is only e ec i e when s uden s comple e i p ope ly," emphasizing he need o
consis en pa icipa ion and hough ul pee engagemen .
Despi e hese conce ns, he quan i a i e esul s indica ed obus suppo o he pee
eedback app oach. App oxima ely 94% o s uden s ag eed o s ongly ag eed ha
pee commen s led di ec ly o imp o emen s in hei manusc ip s. Addi ionally,
a ound 67% pe cei ed he eedback hey p o ided o pee s as use ul, e lec ing a
gene ally posi i e ou look owa d he eedback exchange.
In PGRD, he p elimina y design epo was g aded by he ins uc o o examine how
he ac i i y a ec ed assessmen g ades. S a is ical analysis o g ades be o e and
a e pee eedback con i med i s e ec i eness, showing a signi ican imp o emen in
s uden pe o mance (mean g ades inc eased om 7.99 o 8.54, p < 0.001). The end
o e m Cou se E alua ion Repo u he ein o ced hese indings, highligh ing ha
100% o esponden s b oadly ag eed ha he eedback ecei ed du ing he cou se
acili a ed hei lea ning as did eedback p o ided by he ins uc o . These indings
unde sco e he alue o s uc u ed, quali y eedback wi hin he eaching en i onmen .
4.4 Ins uc o e lec ions on eedback and ecommenda ions
Ins uc o s om each cou se p o ided hei e lec ions on he me a eedback p ocess
and iden i ied he ollowing commen s and ecommenda ions.
Ac i i y Adminis a ion in Moodle Wo kshop: The s aged pee e iew ac i i y was
adminis e ed using a se ies o laye ed Moodle wo kshop ac i i ies. This wo k low
equi ed s uden s o copy he eedback hey p o ided in Module 1 and submi i in
Module 2. This double handling c ea ed some challenges: some s uden s didn’
comple e he ac i i y, couldn’ ind whe e he eedback was o mixed up which o he
Module 2 o submi o (luckily he las was a e). This c ea ed some ex a wo k o
ins uc o s. Al e na i es o his me hod could be unning he session in class a he
han asynch onously o e a week, whe e oubleshoo ing could un in eal- ime.
S eamlining he p ocess could also be enabled by specialised so wa e which
appea s o o e an au oma ic wo k low bu equi es in es men and in eg a ion.
Explici Ins uc ion in Feedback P o ision: While he ask was explained in each
class, and he impo ance p e aced in e ms o p o essional enginee ing skills,
s uden engagemen emained an issue in some classes. Fu he a ge ed aining in
p o iding and ecei ing eedback, along wi h clea p o essional expec a ions, could
inc ease s uden engagemen in he ac i i y.
Rub ic Enhancemen : Du ing he s udy, we iden i ied gaps in he exis ing ub ic,
pa icula ly ega ding eedback one, and subsequen ly we included one c i e ia in
he UGRD+PGRD ub ic in esponse. As ub ics a e impo an in guiding s uden
expec a ions and s uc u ing eedback, hey may equi e ongoing e inemen o
enhance consis ency and cla i y in u u e pee e alua ions.
Emphasis on Re lec i e P ac ice: To close he eedback loop we sugges
in eg a ing s uc u ed e lec i e ac i i ies wi hin all cou ses, o solidi y s uden
lea ning om he e iew p ocesses. These may ake he o m o s uden s w i ing up
and documen ing changes made o hei assessmen s (much like he academic
publishing pee e iew p ocess), jou nal en ies iden i ying a eas o imp o emen o
ac i ely discussing in class co ec i e ac ions.
Ins uc o s we e keen o con inue wi h he ac i i y in hei own cou ses and explo e
po en ial applica ions in younge coho s in o de o sca old and build eedback skills
h oughou p og ams.
5 CONCLUSIONS
The indings o he me a- e iew p ocess analysis iden i ied he u ili y o he ac i i y
no only in inc easing s uden con idence and amilia i y wi h he submission
expec a ions, bu also in p o iding and ecei ing cons uc i e and use ul eedback.
The ac i i y was used in ou di e en cou ses anging in disciplines and coho s, and
in di e en assessmen s om o ma i e, summa i e, epo s and p ac ice
p esen a ions. Ac oss all he cou ses, i was ound ha a balance be ween posi i e
ein o cemen and cons uc i e c i icism was posi i ely ecei ed. The mos posi i ely
ecei ed eedback main ained a espec ul, encou aging, and suppo i e one while
s ill p o iding hones and cons uc i e c i icism. While di e en ypes o eedback
we e p o ided by hose comple ing he same o di e en opic (dep h s b ead h),
bo h s ill p o ided cons uc i e eedback, and di e ences in assigned g ades we e
no signi ican . Su ey ou comes sugges his pee eedback p ocess suppo s
s uden lea ning and manusc ip imp o emen . The s aged e iew app oach o e s a
scalable p ocess o helps s uden s in e nalize assessmen c i e ia, de elop nuanced
unde s anding o e ec i e eedback, and p omo es e lec ion – all o which a e
impo an o p o essional enginee ing p ac ice. Ta ge ed e inemen s, including
aining in eedback guidelines and expec a ions, comp ehensi e ub ics, and
s eamlined ac i i y adminis a ion, could u he imp o e he engagemen ,
consis ency and scalabili y o he p ocess.
