Campus Review Vol. 29 Issue 2 | February 2019 | Seite 8

NEWS
campusreview.com.au
In Business and Science, a male teacher
from an English-speaking background had
the highest probability of getting the highest
possible grade.
In fact, such a Science instructor was
more than twice as likely to get a higher
score on evaluations from local male
students than a female teacher from a
non‑English-speaking background.
Engineering and Medicine faculties
showed slightly less discrepancy – the
only significant bias found was against the
non-English-speaking female cohort when
evaluated by local students.
Female teachers in the Arts and Social
Sciences faculties fared even better. The
researchers found no statistically significant bias against them. They said this suggests
that faculties with a larger proportion of
female teachers will see less bias.
Still, Arts and Social Sciences isn’t a
bias-free hub, with non-English-speaking
teachers getting lower scores when
evaluated by local students.
The biases observed remained unchanged
between undergraduate and postgraduate
degrees.
Lead author and UNSW statistician
Associate Professor Yanan Fan said this
suggests graduates may carry these biases
with them into the workforce.
“Reducing bias will have great benefits
for society, as university students represent
a large proportion of future leaders in
government and industry,” she said.
Study co-author Professor Emma
Johnston said encouraging more women at
the professorial level, in leadership positions
and in membership of key committees will
help shrink biases.
“We need to continue to support women
at all levels of academia in STEM across
Australia in order to smash stereotypes that
create the partiality that exists within our
community.” ■
than 1.4 million electronic health records.
It was then able to draw on its ‘experience’
to diagnose a broad range of childhood
diseases, with accuracy rates of more than
90 per cent in some cases.
The system performed better than junior
doctors, but not quite as well as more senior
experienced physicians.
The scientists who created the AI model
believe it could speed up the triaging of
patients in hospital emergency departments
and improve diagnosis. But sceptical British experts insist that
intelligent machines could never take
the place of human physicians.
The human doctor who asks targeted
questions and then uses his or her
knowledge and experience to make a
diagnosis “can be considered a classifier of
sorts”, said the researchers, writing in the
journal Nature Medicine.
The AI program works in a similar way,
sifting through vast amounts of clinical
information “to mimic the clinical reasoning
of human physicians”.
To create the system, the scientists
obtained electronic health records from
1,362,559 outpatient visits to the Guangzhou
Women and Children’s Medical Center, a
major government hospital in China.
The records covered physician-patient
encounters from January 2016 and July
2017 involving children and teenagers up to
18 years of age.
In total, 101.6 million data points were
used to ‘train’ the program.
Specific words and phrases, as well as
numerical data such as patient temperature,
were analysed by the system and compared
with what it had learned from the electronic
health records. ■
Bias effects
Faculties with most gender, culture
bias revealed in student surveys.
G
ender and culture both affect the
scores teachers receive on student
experience surveys – but the bias is
more notable in certain faculties.
That’s according to a review of almost
525,000 student experience surveys across
five faculties between 2010 and 2016,
published in PLOS ONE.
Gender and cultural generally have a
negative impact on the scores of women
and teachers from non-English-speaking
backgrounds across almost all faculties,
the researchers found.
Doctor AI
Computer medico diagnoses
diseases in children.
A
n artificial intelligence system
designed to diagnose childhood
diseases can recognise symptoms
more accurately than some human doctors,
a study has shown.
The deep learning program, tested in
China, assimilated information from more
6