Indi idualized exe cises o con inuous assessmen in
enginee ing
7 h Teaching & Educa ion Con e ence, London
S. Pineda 1, N. Alguacil 2, J. P´e ez-Ruiz 1, S. Ma ´ın 1, A. Ruiz 1
1Uni e si y o M´alaga (Spain) 2Uni e si y o Cas illa-La Mancha (Spain)
May 21, 2019
1 / 20
Funding
This wo k has been suppo ed by ”I Plan P opio In eg al de Docencia” a
he Uni e si y o M´alaga
2 / 20
Con inuous assessmen a Uni e si y
4Ea ly indica ions o he pe o mance o s uden s
4Inc eased sense o inclusi eness
4Highe lea ning s anda ds o all s uden s
6Inc ease o p o esso wo king load
6G ea e s udy p essu e
6Risk o plagia ism
3 / 20
Online eaching pla o ms
I educes he p o esso wo king load by p o iding eedback o
s uden s au oma ically
I alle ia es he s udy p essu e since s uden s can decide whe e and
when o use hese pla o ms
The e exis ways o educe he plagia ism by p oposing exe cises
whose nume ical solu ions a e di e en o each s uden
Moodle
4F ee and open-sou ce lea ning managemen sys em w i en in PHP
4Al eady in eg a ed in se e al public uni e si ies
6Complex ope a ions no allowed
6IT suppo equi ed o implemen changes
4 / 20
De elopmen o web applica ion
Goal: au oma ic assessmen o indi idualized exe cises
Requi emen s:
Capable o sol ing complex enginee ing p oblems (in eg als, de i a i es,
sys em o equa ions, complex numbe s, e c.)
Flexible enough o implemen changes acco ding o s uden s and
p o esso s needs
P og amming language wi h a smoo h lea ning cu e so ha IT
suppo is no equi ed
Decision:
Py hon: a high-le el, gene al-pu pose p og amming language
Django: a Py hon-based ee and open-sou ce web amewo k
5 / 20

P e ious s eps
1Di ide he cou se in a se ies o blocks o didac ic uni s wi h di e en
lea ning objec i es
2P opose a se ies o exe cises o each block o de ed om lowe o
highe di icul y
3P opose se e al ques ions o each exe cise and he poin s ob ained
by he s uden i answe ed co ec ly
4W i e down each exe cise o mula ion wi h some da a depending on
pa ame e s α, β, γ, . . . (di e en o each s uden )
5W i e down he Py hon code ha p o ides he solu ion o each
exe cise o gi en pa ame e s α, β, γ, . . .
6 / 20
Example o exe cise and code
7 / 20
Use o exe cises in class
Sol ed by p o esso : some o he exe cises a e sol ed in class by he
p o esso using di e en alues o pa ame e s α, β, γ, . . . . S uden s
ha e o sol e he same p oblem using he same p ocedu e bu wi h
di e en da a.
P oposed in class: o he exe cises a e p oposed o be sol ed in class
by each s uden wi h he help o he p o esso s and he es o
s uden s. I a s uden has sol ed co ec ly he p oposed exe cise, she
can o e help o o he s uden s who ha e no inished ye .
P oposed as homewo k: o he exe cises a e p oposed as homewo k o
be sol ed by each s uden a home wi hou any help. The s uden can
check whe he he solu ion is co ec o no using he web applica ion.
8 / 20
Fea u es o web applica ion (s uden )
S uden can check whe he a solu ion o a p oblem is co ec a any
ime om any place
I a ques ion is answe ed co ec ly, he s uden ge s he co esponding
poin s
Each s uden can check he o al ob ained poin s and he posi ion in
he class anking
Each s uden mus answe co ec ly a pe cen age o he ques ions o
block 1 (decided by each p o esso ) o be able o check he answe s
o he ques ions o block 2.
The e is a maximum numbe o ies o answe each ques ion
(decided by each p o esso ). I his limi is exceeded, he s uden can
s ill check whe he he answe is co ec bu she ge s no poin s o i .
9 / 20
Impac : use o web applica ion
Cou se # Exe cises # S uden s # Co ec answe s
EC 175 1505 329495
PS1 62 127 10300
PS2 25 54 2649
Table: Use o web applica ion
16 / 20

Impac : g ades
’14+’15+’16 ’17
Use o web applica ion 63
Non a endance a e o en olled s uden s 12% 22%
Pass a e o s uden s a ending he exam 24% 64%
Pass a e o en olled s uden s 21% 50%
Table: Impac o web applica ion on g ades
17 / 20
Impac : s uden su ey
(a) The webapp has helped me o wo k
daily on he cou se
(b) The applica ion has helped me o
unde s and concep s o he subjec
(c) The applica ion is a good
p epa a ion o he exam
(d) I would ecommend he applica ion
o o he s uden s o his subjec
Figu e: Su ey esul s (5: o ally ag ee, 1: o ally disag ee)
18 / 20
Conclussions
We p esen a eaching inno a ion p ojec based on he de elopmen
o a web applica ion o con inuously assess he lea ning a enginee ing
uni e si y cou ses
The web applica ion has been es ed in 2 uni e si ies and 4 cou ses.
Mo e han 1600 s uden s ha e used he applica ion and almos
340000 ques ions ha e been al eady answe ed co ec ly
The use o indi idualized exe cises mo i a es s uden s o s udy on a
egula basis, inc eases hei unde s anding o he subjec and inc ease
he a e o s uden who pass he cou se
Su p isingly, he class anking d as ically inc eased s uden
pa icipa ion
19 / 20
Thanks o he a en ion!
Ques ions?
[email p o ec ed]
20 / 20