18
As we work with school
districts on the use of student
growth data for teacher
evaluation purposes, some
interesting revelations are
occurring.
High
school
teachers are finding out that
students know a whole lot
more than the teachers thought
they
knew.
Elementary
teachers are discovering that
not all students go “brain dead”
over the summer months and
many maintain and even grow
the knowledge from the
previous grade.
I have written before about the use of data to
make decisions. I have learned a lot about the use
of data from individuals with MBA type degrees. I
remember talking to an alternative certified
superintendent (did not travel through the traditional
education duties of teacher, building level
administrator to superintendent) whowas askingme
questions concerningwhat factors I used to analyze
the hiring of new teachers. I had to admit that we
used a certain cadre of questions, but in the endwe
made the hiring decision on non-scientific type
responses such as the candidates’ love for children.
This administrator related to me an analysis he
was doing with teachers in his district. The district
had a majority of students from Spanish-speaking
families. He had analyzed student growth scores
disaggregated by the native language of the
teacher. He divided teachers into three groups: 1)
NativeEnglish speakers noSpanish skills; 2) Native
English speakers with Spanish as a second
language; and 3) Native Spanish speakers with
English as a second language. His analysis
determined that students in classrooms with native
Spanish speakers with English as second language
outperformed all others. Second were native
English speakers with Spanish secondary and last
were English only speakers. He told me that they
were going to use these results as a screening tool
in the interviewprocess for new teachers.
While I do not think using student growth for
teacher evaluation purposes will be an evolutionary
changing experience for public education, I do think
using data to think about what we are doing will be
important for improving education.
In another example of using data to make
decisions was an analysis I did concerning high
school student performance results on the ACT vs.
the same students’ results on theWorkKeys portion
of thePSAE.
High school administrators and teachers were
concerned that students did not try as hard on the
WorkKeys as they did on theACT because theACT
counted for college admission and the WorkKeys
had no high stakes result. I analyzed the scores
using a decile (each of 10 equal groups intowhich a
population can be divided according to the
distribution of values of a particular variable)
analysis and discovered that there was no
significant difference in student scores. In other
words, if a student scored in the top 10 percent on
the ACT they also scored in the top 10 percent of
theWorkKeys.
A high school district superintendent shared
another example to me. In this use of data the
district analyzed the math courses that students
took in middle school against the grades and level
of math the same student earned in high school. It
was determined that students who successfully
completed Algebra I in the 8
th
grade had a 93
percent chance of getting a 24 or better on the ACT
test as a high school junior. Conversely, a middle
school student who only progressed through 8
th
grade remedial math hadonly a 2 percent chanceof
scoring a 24 or better on the ACT. Once these
statistics were communicated to parents at the
elementary level, the feeding elementary districts
hadmuch greater success getting students to enroll
inmore rigorousmath classes.
Student growth: Usingdata tomakedecisions
Dr. RichardVoltz
AssociateDirector/
Professional
Development
