I am not normally a podcast guy but a podcast that covers the recent AERA study on the edTPA problems actually got me to listen to the whole thing. This is pretty amazing stuff and you should listen to the researchers talk about the MAJOR issues with the edTPA.
What is the podcast about – “Nearly 20 states use the edTPA performance assessment system to determine if pre-service teachers are ready for the classroom. A new study led by Rutgers University’s Drew Gitomer and UCLA’s Jose Felipe Martinez now raises questions about the reliability and validity of edTPA measurements. Gitomer and Martinez join CPRE Knowledge Hub managing editor Keith Heumiller to discuss their findings, recently published in the American Educational Research Journal, and some important implications for performance assessment policy and future research.”
Some pretty crazy things I took from the podcast –
- Candidates who meet or exceed the cut score can go on to teach – those who don’t are not permitted to seek their license and either have to retake the exam or give up on teaching.
- What we do know is that edTPA does not offer appropriate estimates of reliability precision or decision consistency and overall the information that is offered is highly misleading.
- The area is heavily understudied.
- Reliability of assessments like this is usually much lower – approaches and formulas they (edTPA) use are erroneous. .6 to .8 reliabilities when you use normal formulas – edTPA is using their own formulas in order to show higher reliabilities.
- “It is critical to examine the implications of an assessment on historically marginalized groups of examinees particularly those groups that are underrepresented in the teaching force. Now in our paper we point out that African American candidates are much less likely to pass the assessment and more likely to be misclassified in terms of passing status. And so that creates disparate access to the teaching profession. It’s a critical issue and those results are discounted by edTPA in their reports.”
- “We discovered another issue after we published that is quite troubling – decision makers rely on the technical abilities of the assessment – and edTPA in its technical reports claims that all analysis have been reviewed by national recognized psychometricians and they meet the technical standards for licensure assessments set forth by AERA and NCME standards – but based on confidential statements they have received from multiple sources they believe this statement is highly misleading and that technical reviews by experts has not occurred as described.”
This really begs further study and consideration especially since Texas is thinking of using the edTPA as a certification exam for teachers.