Educator Evaluation Systems & Effectiveness Labels

Educator Evaluation Systems & Educator Evaluation Systems & Effectiveness LabelsEffectiveness Labels

Venessa Keesler, Ph.D. & Carla HoweOffice of Psychometrics,

Accountability, Research & EvaluationMichigan Department of Education

Overview of Current Plan and Issues • November release date for the November release date for the aggregateaggregate

effectiveness labels by school (number of teachers effectiveness labels by school (number of teachers reported as highly effective, effective, minimally reported as highly effective, effective, minimally effective, ineffective)effective, ineffective)

• Key important messages:Key important messages:– This is the FIRST YEAR; 800+ different systems This is the FIRST YEAR; 800+ different systems

(we have data to show this)(we have data to show this)– Districts did MASSIVE amounts of work to Districts did MASSIVE amounts of work to

accomplish thisaccomplish this– We do not believe that huge numbers of MI We do not believe that huge numbers of MI

teachers are ineffectiveteachers are ineffective

Current Circumstances

Our current legislation has allowed for local systems of Our current legislation has allowed for local systems of evaluations, which has given districts flexibility to design systems evaluations, which has given districts flexibility to design systems that work best for them.that work best for them.• Over 800 systems across the stateOver 800 systems across the state• Varying degrees of implementation across the state Varying degrees of implementation across the state Public reporting of effectiveness labels is required by SFSFPublic reporting of effectiveness labels is required by SFSF• Scheduled for release in November via Scheduled for release in November via mischooldata.orgmischooldata.org• Teachers labels reported in aggregate by school (number of Teachers labels reported in aggregate by school (number of

teachers in each of the four categories)teachers in each of the four categories)• Principals/Administrators reported at the district level.Principals/Administrators reported at the district level.

Important Context for the Important Context for the 2011-12 Results2011-12 Results

• First year of implementation of NEW systems based First year of implementation of NEW systems based on student growth measureson student growth measures

• State provided student growth measures are only State provided student growth measures are only available in grades 4-8 for reading and mathematicsavailable in grades 4-8 for reading and mathematics

• Varying components across systems (i.e. between Varying components across systems (i.e. between districts)districts)

• Varying percentages of growth across systems (i.e. Varying percentages of growth across systems (i.e. between districts)between districts)

• Some districts on prior contract (i.e. No new system, Some districts on prior contract (i.e. No new system, but reporting labels was required)but reporting labels was required)

K-12 Educator Evaluation Survey

• 792 districts completed the survey about their Evaluation systems from April to August

• Required to be completed by SFSF • Results provide valuable insight into local systems • The types of frameworks used • The % of student growth as a component (law states

“significant”, but it isn’t defined until 2013-14)• Types of growth measures included• Types of decisions informed by the results of

evaluations

Other Frameworks reported include: Charlotte Danielson Framework AND a local component, Teacher Advancement Program, My Learning Plan, 5 Dimensions of Teaching and Learning, Local District or ISD framework, McREL, STAGES, Kim Marshall Rubrics

50% of reporting districts

54 districts with a prior contract did not have to incorporate growth or a new system in 2011-12#

of

dis

tric

ts

PRELIMINARY/DRAFT FINDINGS

Appropriate given the FIRST year of local evaluation systems

# o

f dis

tric

ts

Other ways growth data are measures include: Combination of data from multiple assessments, pre/post test data, combination of local, state, national measures, benchmark testing, several sources as agreed upon in the professional growth plan

# o

f dis

tric

ts

Others types of assessment data reported that factor into educator evaluations include: AIMSweb, DRA, Ed Performance Series, Fontes & Pinnell, STAR Reading and Math, CBM for Math, DELTA Math

# o

f dis

tric

ts

Others types of assessment data reported that factor into educator evaluations include : AIMSweb, DRA, Ed Performance Series, Fontes & Pinnell, STAR Reading and Math, TerraNova, ITBS, DELTA Math

# o

f dis

tric

ts

Others types of assessment data reported that factor into educator evaluations include : AIMsweb, Ed Performance Series, STAR Reading and Math, Study Island

# o

f dis

tric

ts

Others types of assessment data reported that factor into educator evaluations include : common assessments, district benchmark assessments, Scantron Performance Series

# o

f dis

tric

ts

PRELIMINARY/DRAFT FINDINGS

Others types of decisions include: Assignment to committees or roles beyond the classroom, classroom support and assistance, layoff/recall/transfer, mentoring, staff placement, scheduling, setting improvement goals, merit pay

# o

f dis

tric

ts

Other Factors Reported As Part of Evaluations

Absenteeism rate Professional responsibilitiesClassroom management Student achievement dataContent knowledge Adherence to School

Improvement PlansInstructional practices Commitment to School &

DistrictUse of technology Learning environmentPedagogical knowledge and practice

Parental communication/involvement

Professional development participation

Relationships with students

PRELIMINARY/DRAFT FINDINGS

UNDERSTANDING EDUCATOR EVALUATION LABELS IN MI

Overview of Statewide Results

Caveat….

• Labels are not EQUAL across districts• However, we know that people will want this

type of analysis and we want it done appropriately

Statewide ResultsLabel Number Percent

Ineffective 775 0.82%

Minimally Effective 1,998 2.11%

Effective 70,742 74.74%

Highly Effective 21,141 22.33%

IMPORTANT NOTES:•Based on the labels as determined by the local evaluation system; rigor of label designation is not consistent across districts•THERE is differentiation in label reporting now, 22% of teachers are reported as “highly effective” moving away from a satisfactory/unsatisfactory system•We do not believe that 1% of teachers labeled as “ineffective” is unreasonable in the first year

Impact of growth

• Law required districts to implement systems based in “significant part” on student growth

• How do the labels look different when the district used growth in greater percentages?

Growth and eval labelsLess than 10%

11-20% 21-30% 31-40% 41-50% >50% State

Ineffective 0.6% 0.6% 1% 0.5% 2% 1.4% .82%

Min Effective

1% 2% 3% 6% 3.5% 1% 2%

Effective 83% 71% 74% 64% 68% 82% 74%

Highly Effective

15% 26% 23% 30% 27% 15% 22%

26,253 22,639 25,309 5,650 3,143 1,485 87,575

More differentiation in

labels when growth counts at

a higher rate

LESS differentiation without growth

Distribution of Labels By Percent of Evaluation Based on Growth

Key Takeaways

• Distribution of labels (i.e. number of teachers in each category):– Is appropriate in Year 1 of implementation– Reflects differentiation (esp highly effective/effective)

• BUT we also see that systems using higher proportions of growth are able to make those differentiations more accurately

• The statewide evaluation system will move us toward more growth measures at higher rates

Who is more likely to be rated as highly effective or effective?

Teachers more likely to appear in highly effective category (versus other three) and in effective category (versus other two):•Female teachers•Those with more time in the same district•Teachers with a professional certificate (as opposed to all others)•Those with a master’s degree or higher•Teachers in districts with growth over 40% in their system

Who is less likely to be rated as effective or highly effective?

• Older teachers• New teachers (those in their first year of

teaching)• Mathematics, science, social science, special

education and world language teachers (relative elementary teachers)

• Teachers in systems where growth is less than 10% of the evaluation system

Relationship between effectiveness labels and Priority/Focus/Reward

• Important to remember:– A school-level designation does not mean that all

teachers within that school are in a given level of effectiveness

– Example: In a Priority School, there will be effective teachers as well as ineffective teachers

Ineffective Minimally Effective

Effective Highly Effective

Priority 2.5% 6.3% 74% 18%

Focus 0.5% 1.5% 80% 18%

Reward 0.5% 0.85% 74% 24%

Statewide 0.8% 2% 74% 22%

Notes:There are significantly more teachers reported as ineffective and minimally effective in Priority Schools than the statewide number, and in Focus or Reward schools.

Effectiveness Labels in Priority, Focus and Reward

Schools

Key Takeaways from the Results 1) These results are reasonable for the first

year; represent a huge effort on the part of districts

2) There is differentiation in the system; there will be more as growth becomes a higher component; but we still do not believe large numbers of Michigan teachers are “ineffective”

Questions?

• Contact Office of Evaluation, Strategic Research and Accountability (OESRA)

• 517-373-1342