Mona Chalabi: 3 ways to spot a bad statistic
Mona Čalabi (Mona Chalabi): Tri načina da prepoznate loše statističke podatke
Mona Chalabi tries to take the numb out of numbers. She's left with lots of "ers." Full bio
Double-click the English transcript below to play the video.
about statistics today.
pomalo obazrivo, to je u redu;
a little bit wary, that's OK,
kind of crazy conspiracy theorist,
especially now, you should be skeptical.
treba da budete skeptični.
da možete da prepoznate
which numbers are reliable
izvesna pomagala da biste to umeli.
some tools to be able to do that.
I'm talking about here.
o kojim brojevima ovde govorim.
ovu kremu protiv bora.“
this anti-aging cream."
uvek prevrće očima na takve brojeve.
roll our eyes at numbers like that.
dovode u pitanje podatke poput:
are questioning statistics like,
rate is five percent."
je to što ne proističe iz privatne firme,
it doesn't come from a private company,
ne veruje ekonomskim podacima
distrust the economic data
taj broj je još veći; oko sedam od deset.
it's even higher;
in our society right now,
mnogo linija razdvajanja,
prema tim vladinim brojevima.
with these government numbers.
da su ovi podaci od ključnog značaja,
these statistics are crucial,
kao celina imalo smisla,
of society as a whole
emotional anecdotes
and they don't really reflect
u svakodnevnom životu ljudi.
in people's everyday lives.
is winning the argument right now.
trenutno pobeđuje u raspravi.
of alternative facts,
nekom vrstom zajedničke osnove,
this kind of common ground,
moves in the US right now
nekih vladinih statističkih podataka.
statistics altogether.
o merenju rasne nejednakosti.
about measuring racial inequality.
ne treba koristiti
money should not be used
o rasnoj segregaciji.
pravednu politiku
trenutni nivo nepravednosti?
current levels of unfairness?
već o svemu; razmislite o tome.
o zdravstvenoj zaštiti
on health or poverty?
o zdravlju ili siromaštvu?
about immigration
ako se ne možemo makar složiti
and leaving the country?
ulazi u zemlju i izlazi iz nje?
tako su dobili svoje ime.
that's where they got their name.
bolje mere stanovništva
measure the population
beyond either blindly accepting
njihovo slepo prihvatanje,
da bismo mogli da uočimo loše podatke.
to be able to spot bad statistics.
in a statistical department
je proterano iz svojih domova
had been forced from their homes
ali, takođe, neverovatno težak.
but it was also incredibly difficult.
of our numbers --
bi trebalo da idemo,
of the country we should go to,
koja pitanja treba da postavljamo.
really disillusioned with our work,
u vezi sa našim radom,
a really good job,
da zaista dobro obavljamo posao,
who could really tell us were the Iraqis,
da nam ispriča stvari bili su Iračani,
our analysis, let alone question it.
da naiđu na našu analizu,
numbers more accurate
da brojevi budu tačniji
be able to question them.
može da ih preispituje.
koja se bavi podacima.
and sharing them with the public.
i podelim ih sa javnošću.
you don't have to be a geek or a nerd.
štreber ili bubalica.
koriste ih ljudi
they're used by people
dok se pretvaraju da su skromni.
while pretending they're humble.
da primetite loše statističke podatke.
some bad statistics.
možete li da uočite nepouzdanost?
is: Can you see uncertainty?
odnos ljudi prema brojevima,
people's relationship with numbers,
sa političkim anketama
with political polls
da izveštava o činjenicama,
is actually to report the facts
mogu da naškode demokratiji
can actually damage democracy
da se ne trude da glasaju za nekog
don't bother to vote for that guy,
o preciznosti ovog nastojanja.
about the accuracy of this endeavor.
in the UK, Italy, Israel
Italiji, Izraelu
US presidential election,
predsedničkih izbora u SAD,
za predviđanje ishoda izbora
za predviđanje prijema u bolnice.
to predict hospital admissions.
stvarne podatke iz akademske studije
from an academic study to draw this.
polling has become so inaccurate.
zašto je anketiranje postalo tako netačno.
fini reprezentativni uzorak stanovništva
to get a really nice representative sample
na telefon anketarima,
their phones to pollsters,
ljudi mogu da slažu.
people might lie.
know that to look at the media.
pogledavši medije.
of a Hillary Clinton win
pobede Hilari Klinton
da izrazimo temperaturu.
to describe the temperature.
230 miliona glasača u ovoj zemlji
of 230 million voters in this country
will overstate certainty, and it works --
će prenaglasiti sigurnost, i to deluje -
our brains to criticism.
naš mozak za kriticizam.
you might feel skeptical.
možda ćete biti skeptični.
of objective science,
da ovo bolje prenesem ljudima,
to better communicate this to people,
u našim brojevima.
in our numbers.
real data sets,
stvarne skupove podataka
hand-drawn visualizations,
nacrtane rukom,
koliko su podaci neprecizni;
how imprecise the data is;
da je to uradilo ljudsko biće,
of finding out the probability
distribution of flu season.
sezone gripa.
vizualizacija podataka,
data visualization,
the exact probabilities,
da dobiju vakcine protiv gripa
people to get their flu jabs
these imprecisions,
nužno sa sobom određeni broj,
walk away with a specific number,
da nepravde i nejednakosti
leave a huge mark on our lives.
Americans have shorter life expectancies
i Indijanci imaju kraći životni vek
can be kept in solitary confinement cells
u SAD mogu držati u samicama
of an average parking space.
prosečnog mesta za parkiranje.
is also to remind people
takođe je da se ljudi podsete
statistical concepts,
sadrži 6,23 fekalnih nezgoda.“
contains 6.23 fecal accidents."
swimming pool in the country
iz Centra za kontrolu i prevenciju bolesti
which comes from the CDC,
47 objekata za plivanje.
redistributing poop.
u preraspodeli kake.
how misleading averages can be.
kako prosek može da obmane.
that you guys should be asking yourselves
koje treba da postavite sebi
odnosi i na prosečne vrednosti,
about averages in a way,
zašto ljude toliko frustriraju
why people are so frustrated
o tome ko pobeđuje a ko je na gubitku
of who's winning and who's losing
globalne prosečne vrednosti
are frustrated with global averages
sa njihovim ličnim iskustvom.
with their personal experiences.
data relates to their everyday lives.
sa njihovim svakodnevnim životom.
pod nazivom „Draga Mona“,
called "Dear Mona,"
i iznosili svoja pitanja i probleme,
with questions and concerns
pomoću podataka.
in a separate bed to my wife?"
u odvojenom krevetu od svoje žene?“
zbog svojih tetovaža?“
of natural causes?"
because they make you think
jer vas teraju da razmislite
and communicate these numbers.
i saopštite ove brojeve.
„Koliko piškenja je mnogo?“,
"How much pee is a lot of pee?"
that the visualization makes sense
da vizualizacija ima smisla
in the appendix of an academic study.
u prilogu akademske studije.
ove brojeve o količini mokrenja,
these numbers on urination volume,
and try it for yourself.
mora da se izričito odnosi na vas.
has to relate specifically to you.
je dobilo novčanu kaznu u Francuskoj
were issued fines in France
niti nosim veo preko lica.
or wear the face veil.
gde se vi uklapate
is to get as much context as possible.
što je više konteksta moguće.
sa jednog podataka,
from one data point,
is five percent,
by educational status --
s obzirom na status obrazovanja -
wanted you to go to college --
da idete na fakultet -
s obzirom na pol.
muškaraca viša
it was the other way around.
of the biggest changes
kada sagledate stvari izvan proseka.
once you look beyond the averages.
možete promeniti priču.
you can change the story.
that I want you guys to think about
o kojem želim da razmišljate
na koji se podaci saopštavaju,
data is communicated,
podjednako je bitan.
matters just as much.
i nekako dosadna,
and actually kind of boring,
pomoću kojih možete ovo proveriti.
you can take to check this.
in this country support jihad,
muslimana u ovoj zemlji podržava džihad,
svuda 2015. godine.
the original questionnaire.
koji su izveštavali o tom podatku
who reported on that statistic
lower down on the survey
kako definišu „džihad“,
how they defined "jihad."
da budu religiozniji“.
to be more religious."
kao „nasilan sveti rat protiv nevernika“.
"violent holy war against unbelievers."
it's totally possible
who defined it as violent holy war
definisao kao nasilni sveti rat
uopšte ne preklapaju.
kako je istraživanje sprovedeno.
how the survey was carried out.
opciona anketa,
da je nađe na internetu i popuni je.
on the internet and completed it.
uopšte identifikuju kao muslimani.
if those people even identified as Muslim.
je bilo 600 ispitanika.
respondents in that poll.
približno tri miliona muslimana,
Muslims in this country,
otprilike jednom od svakih 5 000 muslimana
one in every 5,000 Muslims
često bolji od privatnih.
are often better than private statistics.
par stotina ljudi, možda hiljadu,
hundred people, maybe a thousand,
proizvode za negu kože 2005. godine,
skin care products in 2005,
to claim that they work.
da biste tvrdili da deluju.
interest in getting the numbers right,
da dobiju ispravne brojeve,
odgovarajući brojevi.
they're totally impartial,
obavlja svoj posao
their jobs regardless of who's in power.
to a couple hundred people.
na koje se uporno pozivam
I keep on referencing
Ministarstva za rad,
businesses in this country.
preko 140 000 firmi u ovoj zemlji.
koji dolaze iz privatne kompanije,
that comes from a private company,
za sebe i gomilu prijatelja, isprobate,
and a bunch of friends, test it out,
da su brojevi bili pogrešni.
you can say the numbers were wrong.
vladine podatke?
government statistics?
sve što treba da vidite.
on the chart you need to see.
jer ako odustanete,
altogether, because if you do,
o javnoj politici u neznanju,
decisions in the dark,
lične interese kao smernice.
interests to guide us.
ABOUT THE SPEAKER
Mona Chalabi - Data journalistMona Chalabi tries to take the numb out of numbers. She's left with lots of "ers."
Why you should listen
After working for a humanitarian organisation, Mona Chalabi saw how important data was, but also how easily it could be used by people with their own specific agendas. Since then, her work for organizations like Transparency International and The Guardian has had one goal: to make sure as many people as possible can find and question the data they need to make informed decisions about their lives.
Chalabi is currently the Data Editor of the Guardian US, where she writes articles, produces documentaries and turns data into illustrations and animations. In 2016, her data illustrations were commended by the Royal Statistical Society.
Mona Chalabi | Speaker | TED.com