STAY SAFE OUT THERE ON THE ROADWAYS.

>> ALL RIGHT.

THANK YOU.

>>> TURNING TO EDUCATION, SOPHISTICATED ALGORITHM CALLED DROP OUT EARLY WARNING SYSTEMS DO'S USED BY WISCONSIN ONLY A FEW LIKE IT.

SYSTEM TRIES TO PREDICT WHETHER A STUDENT EXITING MIDDLE SCHOOL IS LIKELY TO GRADUATE FROM HIGH SCHOOL.

ACCORDING TO THE REPORT FROM MARK UP.

PUBLICATION THAT LOOKS AT USE OF BIG TECH.

NUMBER OF DATA POINTS, TEST SCORES, DISCIPLINARY RECORDS AND FAMILY INCOME.

UNIQUE AND CONTROVERSIAL DATA POINT IS RACE.

DPI SAYS DO'S USES WHAT HAS ALREADY HAPPENED IN THE PAST WHAT WILL HAPPEN IN THE FUTURE.

ALL IN ORDER TO CHANGE COURSE AND HELP INSURE STUDENTS IDENTIFIED AS AT-RISK DOING IN FACT GRADUATE ON TIME BY DIRECTING RESOURCES, SUPPORT AND INTERVENTIONS TO STUDENTS THAT MAY NEED THEM.

WE TURN TO REPORT'S AUTHOR TODD FEATHERS, THANKS FOR BEING HERE.

>> THANK YOU SO MUCH FOR HAVING ME.

>> YOU SEEM TO HAVE UNCOVERED PREDICTION FOR HIGH SCHOOL GRADUATION MOST OF US HAVE NEVER HEARD OF.

HOW DOES IT WORK?

>> IT IS A SYSTEM BUILT IN 2012 TO PREDICT WHETHER STUDENTS WILL GRADUATE ON TIME.

GENERATES PREDICTION EVERY STUDENT 6-9TH, MILD, MODERATE OR LOW RISK OF DROPPING OUT.

SUPPOSED TO HELP STOOLS AND DIRECT RESOURCES, SUPPORT FOR STUDENTS THAT NEED IT MOST.

WHAT WE FOUND SYSTEM IS WRONG NEARLY THREE QUARTERS OF THE TIME A STUDENT WON'T GRADUATE.

LACK OF HISPANIC, THAN WHITE STUDENTS.

ON TOP OF THAT, ACADEMIC RESEARCH BASED OUT OF UNIVERSITY OF CALIFORNIA BERKELEY CONDUCTED STUDY LARGEST OF ITS KIND ON DO'S.

NOT ACHIEVED PRIMARY GOAL.

>> SO THIS QUESTION IF IT IS WRONG ACCORDING TO YOUR REPORTING, THREE QUARTERS OF THE TIME, WHY IS THAT?

>> SO BY SOME STANDARDS, DO'S IS VERY ACCURATE.

IT IS CORRECT 97% OF THE TIME A STUDENT WILL GRADUATE.

WHAT DPI DID IS CALL -- CALIBRATE IT.

SO THEY ARE SAYING, WE ARE WILLING TO LABEL A LOT OF STUDENTS LIKELY NON GRADUATES IN ORDER TO CAPTURE ALL OF THE ONES.

>> COULD IT BE ERROR AS A RESULT OF RESOURCE AND INPUTS GIVEN TO THE STUDENTS THAT MIGHT BE PREDICTED NOT TO GRADUATE?

>> THAT IS ONE OF THE THINGS UNIVERSITY OF CALIFORNIA BERKELEY RESEARCHES TESTED FOR.

THEY WERE GOING TO STUDENTS LABELED HIGH RISK, YOU WOULD EXPECT TEN YEARS OF DATA FOR IMPROVED GRADUATION RATES.

DATA SHOWED NO IMPACT ON GRADUATION RATES AND IDENTIFIED.

>> TURNING TO ANOTHER DATA POINT LOADED INTO THIS ALGORITHM, THAT IS RACE.

SOMETHING YOU CALL IN YOUR ARTICLE QUOTE A RACIALLY INEQUITABLE ALGORITHM BUILT BY THE STATE OF WISCONSIN.

WHAT DID YOU LEARN WHY RACE IS USED AS A PREDICTOR?

>> IMPORTANT TO REMEMBER THIS ALGORITHM DESIGNED BACK IN 2012.

WE HAVE COME A LONG WAY HOW WE UNDERSTAND ALGORITHMS AND BIASES IN THE WAY, THINGS LIKE RACE CAN FACTOR INTO THESE PREDICTIONS.

THIS ALGORITHM HASN'T CHANGED MUCH.

BACK AT THAT TIME, EVEN STILL CURRENTLY, RESEARCHERS IN THIS FIELD OF EARLY WARNING SYSTEMS, USING DATA POINTS LIKE RACE IMPROVES ACCURACY.

BETTER AT PREDICTING WHO WILL AND WON'T GRADUATE HIGH SCHOOL.

YOU CAN SEE WHY THAT WOULD BE THE CASE.

WISCONSIN STARK RACIAL DISPARITIES, WHITE STUDENTS GRADUATING AT A MUCH HIGHER RATE THAN BLACK STUDENTS.

GROWING BODY OF RESEARCH AND ACADEMIC ARGUMENT SAY A SMALL ACCURACY IS NOT ENCODING HISTORICAL BIASES.

>> I WANT TO GET TO SOMETHING THE DPI TOLD US, YOU QUOTED A SPOKES PERSON SAID EDUCATION SYSTEM IS SYSTEMATICALLY RACIST.

SHE TOLD US, WHEN IT COMES TO LABELLING STUDENTS, I AGREE WE SHOULD BE CONCERNED ABOUT THE IMPACT THE WORDS HAVE ON OUR LEARNERS.

YET INTENTION OF THE LABEL IS TO PROVIDE ACCESS TO THE INTERVENTIONS THESE STUDENTS NEED.

I WISH WE DIDN'T NEED A LABEL.

IT IS AN ATTEMPT TO PROVIDE RESOURCES TO THE LEARNERS THAT NEED IT MOST AND SCHOOLS NEED TO IDENTIFY FOR THEM.

HOW DID THE STUDENTS LIKE LABELED HIGH-RISK AND RACE WAS PUT INTO THE ALGORITHM?

>> YOU KNOW, FIRST THING NONE OF THE STUDENTS SPOKE TO KNEW THE ALGORITHM EXISTED OR SCHOOL WAS LOOKING AT THIS INFORMATION AND MAKING DECISIONS HOW THEY TREAT THEM.

SHOCKED A LOT OF THE STUDENTS.

SHOCKED ME WHEN I FIRST HEARD.

BLACK AND HISPANIC STUDENTS OFTEN FELT LIKE SECONDARY SCHOOL SYSTEMS WITHIN THEIR OWN SCHOOLS.

THEY WERE TREATED DIFFERENTLY THAN WHITE STUDENTS.

THEY FEARED HAVING THE BEHIND THE SCENES LABELS WOULD PERPETUATE KIND OF THE CONSCIOUS BIASES.

YOU CAN IMAGINE FOR INCOMING 9TH GRADERS, FIRST THING YOU SEE ABOUT A STUDENT I AM MODERATELY LOW RISK.

>> ALL RIGHT.

TODD FEATHERS, REALLY COMPLICATED AND INTERESTING REPORTING.

THANKS VERY MUCH.

>> THANK YOU SO MUCH AGAIN.

>> FOR MORE ON THIS AND OTHER ISSUES FACING WISCONSIN VISIT