Education Week - March 25, 2015 - 10

New Studies Challenge
Notion That Teachers
Plateau Early in Careers
CONTINUED FROM PAGE 1
from all sorts of people, Bill Gates on
down," said John P. Papay, an assistant
professor of education and
economics at Brown University, in
Providence, R.I. He is the co-author
of one of two new studies on the topic.
"But teacher quality is not something
that's fixed. It does develop, and if
you're making a decision about a
teacher's career, you should be looking
at that dynamic."
Better With Age
Investigating the connection
between a teacher's experience
and his or her teaching quality
has long proved methodologically
challenging, largely because of the
difficulty in comparing cohorts
of students taught by teachers of
varied experience levels with different
training and backgrounds.
Studies based on such cross-sectional
comparisons have tended to
find few performance differences
between early- and later-career
teachers.
Beginning in the early 2000s,
scholars began to track the same
teachers over time, linking them to
their students' test scores. But there
are pitfalls to that type of statistical
modeling, too. For one, it requires
researchers to make assumptions
about a typical teacher's growth
trajectory over time in order to disentangle
the effects of each year of
experience from other possible influences,
such as a change in class
size or curriculum that might have
occurred.
In their new study, Mr. Papay and
his co-author, Matthew A. Kraft, also
of Brown University, show that some
assumptions in prior research have
had a tendency to depress the effect
of teachers' experience on student
achievement.
For their study, forthcoming in
the Journal of Public Economics,
the researchers looked at a set of
some 200,000 student test scores
linked to about 3,500 different
teachers from an unnamed urban
district. They analyzed those data
using three different methods, each
of which relies on different baseline
assumptions about how to capture
growth in teacher effectiveness as
teachers gain experience.
Under all three of the models
studied, the researchers found
teachers' ability to improve student
achievement persisted well beyond
the three- to five-year mark.
While the teachers did make the
most progress during their first
few years in the classroom, teachers
improved their ability to boost
student test scores on average by
40 percent between their 10th
and their 30th year on the job, the
study shows.
The improvements were seen in
both reading and math teachers, but
were stronger in mathematics.
Beyond Test Scores
What's more, teachers with more
years of experience are better
equipped to boost more than just
test scores, according to a second
new study, released as a working
paper by the Washington-based
BLOGS www.edweek.org/go/blogs
Science Standards Call
For Teaching 'Mysteries'
| CURRICULUM MATTERS | In a packed
session at the National Science Teachers
Association's recent conference, a professor
who helped lead the development of the Next
Generation Science Standards described the
new standards as "a shift from learning about
something to figuring out something."
Brian J. Reiser, a professor of learning
sciences at Northwestern University, offered
this example: "Ngss does not ask you to
explain photosynthesis, ngss asks you to
explain how a tree gets all its stuff."
Traditionally, science classes have been
taught a few different ways, he said. One way
is through application: The teacher presents
the idea, then students do the lab experiment
to see it in action.
The ngss storyline is different. Students
are given a big question that they can relate
to-a "mystery" of sorts. Through their
investigation of that question, they hit on other
phenomena along the way that they also need
to investigate and explain.
Reiser showed a lesson in which students
were told that there was a large decrease in
the number of Galapagos finches between 1976
and 1977. Students were tasked with figuring
out why so many finches died and why some
survived. They were given access to data on
the Web and had to figure out which questions
to ask and what information was relevant.
Eventually, students determined that there
was a drought at that time and that the seeds
the birds ate were nearly depleted. Birds with
longer beaks survived because they were able
to open the leftover, tough-shelled seeds.
From there, students probe a similar
phenomenon-say, why peppered moths
were more prominent during the Industrial
Revolution. "Then you ask students to tell the
story without the finch or moth," Reiser said.
Eventually, they come up with a model.
Reiser said that's when you deliver the
kicker: "Scientists have built a story like this,
too, and it's called natural selection." -LIANA HEITIN
Using Data to Match
Students to Career Paths
| COLLEGE BOUND | More K-12 and college
systems are turning to technology and
analytics to better engage and track students.
At the recent SXSWedu conference,
administrators shared how expanded access
to data helped them improve career planning
10 | EDUCATION WEEK | March 25, 2015 | www.edweek.org
for high school students, deepen learning
experiences on college campuses, and retain
students who might otherwise fall off track.
Concerns over privacy, along with limited
time and budgets, keep many systems from
fully realizing their visions for leveraging data
and innovation.
On one panel, Doyle Vogler, an assistant
superintendent of schools in Lubbock, Texas,
explained how high school students in his
district create online personal graduation
plans that they update yearly with the input
of counselors and parents. With the program,
students track their own progress and link
to universities with degrees in their area of
interest and job-market prospects for those
fields.
It allows students to become "consumers
of their own education," said Vogler of the
approach being used in his 30,000-student
district where 70 percent of students are
low-income. By looking at past performance,
students are given projections of their likely
success in future courses-although not
"tracked," he added. The analysis helps
students create career pathways, which
might not include college but can help
them see potential matches in fields from
manufacturing to information technology.
-CARALEE J. ADAMS
.2
.15
LONG-TERM GAINS
.1
.05
North Carolina middle school math
teachers' ability to boost their students'
scores improved for more than a decade
in the 2000s-not just in their first 3 to 5
years. (The tails for each point represent
confidence intervals.)
SOURCE: National Center for Analysis of Longitudinal Data
in Education Research
0
0
10
Teachers' years of experience
20
30
National Center for Analysis of
Longitudinal Data in Education
Research.
Researchers Helen F. Ladd and
Lucy C. Sorenson, both of Duke University,
in Durham, N.C., analyzed
records from about 1.2 million middle
school students in North Carolina
from 2007 to 2011, including
absences, reported disciplinary offenses,
and test scores. The data also
contain responses from 6th through
8th graders about time spent on
homework and their reading habits.
Using a value-added method
similar to that of the Brown University
scholars, Ms. Ladd and Ms.
Sorenson similarly found that, on
average, the students' teachers
continued to improve their effectiveness
in boosting academic outcomes
for at least 12 years.
Regarding nontest outcomes, the
data show that as teachers gained
experience, they were linked to lower
rates of student absenteeism. The
researchers postulate that more experienced
teachers got better at motivating
students and in classroom
management, resulting in better attendance
and fewer infractions.
The study also found suggestive
evidence of benefits to time spent on
reading and homework completion.
But because of statistical "noise" surrounding
those findings, they are not
as precise.
The nontest findings were most
marked for reading teachers rather
than for math teachers, in contrast
to the student-achievement findings,
which were stronger for the math
teachers.
Unions Respond
In all, the new studies paint
teacher quality as a mutable characteristic
that can be developed, rather
than a static one that's formed in the
first few years on the job.
That's a welcome change for the
3 million-member National Education
Association, which has long
maintained that teacher experience
matters and should be considered in
determining pay and promotions.
"These are incredibly important
studies, and I think we'd make a big
mistake if we didn't look at them carefully
and re-examine some assumptions,"
said Segun Eubanks, the director
of teacher quality for the nea. "The
idea of teachers maxing out in five
years was so contradictory to what we
know about other professions."
Mr. Eubanks said that the findings
suggest policymakers redouble
efforts to improve teacher retention
and evaluating teachers on factors
beyond test scores.
"It isn't that you scrap all reforms
and go back to the good old
days, but it's time to look at a third
way-career ladders, shortened
salary schedules, hybrid teaching
roles," he said.
Both sets of researchers stressed
that their findings concern the average
teacher's rate of improvement
over his or her career. They shouldn't
be interpreted to mean that experienced
teachers are always better
than novices.
The studies also dovetail with a
small but growing body of research
suggesting that high-quality coaching
and professional development
can improve teacher effectiveness.
"My policy conclusion from this is
that we have to help teachers grow.
They have the potential," said Ms.
Ladd, a professor of economics. "You
want to get high-quality teachers in
the first place and then you want to
stick with them."
Students' math test score (standard deviation)
http://www.edweek.org/go/blogs http://www.edweek.org
Education Week - March 25, 2015

Table of Contents for the Digital Edition of Education Week - March 25, 2015
