Thursday, November 6, 2014

Triangulating Your Teacher Evaluation Data

Most teacher evaluation protocols prescribe a fairly rigorous process for observers to establish an acceptable level of inter-rater agreement among their peers when rating the categories of their District’s teacher practice instrument. Though protocols may vary, a vast majority involve observers reviewing a series of video scenarios and ranking each video according to specific look-fors embedded in their teacher evaluation rubric.
Unfortunately, many school systems and states have witnessed inflated scores for the teacher practice portion of the teacher evaluation protocol even though observers had achieved “certification” with their chosen instrument. Why the inflated scores? Theories abound ranging from threats of teacher grievances to staged “dog and pony” performances during announced observations. Regardless of the theory, the net effect of the millions of dollars invested in teacher evaluation training to improve instruction has been negligible at best.
How can school systems establish an acceptable level of fidelity between the key “look-fors” of their teacher evaluation instrument and actual teaching behaviors? One method is to begin triangulating the teacher evaluation scores using additional sources. For example, districts using any version of the Danielson Framework for Teaching can triangulate their results using informal walkthrough data and teacher lesson plan submission data to determine the degree of alignment among the three data sources. Provided below is a sample triangulation that uses the empirically-validated LoTi (Levels of Teaching Innovation) Framework to triangulate Danielson Evaluation Practice scores, H.E.A.T. Walkthrough rating scores, and lesson plan submission scores.
LoTi Level Ratings
Danielson Evaluation Practice Ratings
H.E.A.T. Walkthroughs (LoTi Level)
Lesson Plan Submissions (LoTi Level)
Danielson Evaluation Practice Scores
Level 0 - Nonuse
Unsatisfactory
2%
0%
0%
Level 1 - Awareness
Basic
8%
6%
3%
Level 2 - Exploration
Basic
28%
26%
5%
Level 3 – Infusion
Proficient
46%
52%
72%
Level 4 – Integration
Distinguished
8%
12%
12%
Level 5 – Expansion
Distinguished
6%
2%
4%
Level 6 - Refinement
Distinguished
2%
2%
4%
 A cursory glance of this table reveals a dramatic gap between walkthrough and lesson plan submission scores and the Danielson Evaluation Practice scores based on the LoTi Framework. This scenario describes the majority of school systems during the initial year of formal implementation of their teacher evaluation instrument. Moving into Year 2 implementation and beyond, the goal might be to increase the alignment among the three indicators by 10% annually and over time begin to document increased fidelity between the instructional (How we plan instruction) and operational (How we actually teach) curricula. Only by following a model of triangulation can the teacher evaluation practice score possess any real clout in terms of quantifying growth as it pertains to improved professional practice in the classroom.