Analyze The Data Of Teaching Capability Assessment Based On Generalizability Theory And Many-facet Rasch Measurement

Nowadays, the teaching capability is highly valued at home and abroadin the process of Authenticating Teacher’s Qualification. The ratingmethod of Teaching Capability Assessment is mainly that several ratersmark the participant on the performances. The rating method is subjectiveto lead to the large error, which influences the justification of TeachingCapability Assessment.While the previous studies rarely make quantitative analysis on therating method of Teaching Capability Assessment, this study analyzes theteaching capability assessment result on the basis of GT and MFRM by GENOVAand FACETS which is to explore the influence of the scoring factors andtheir mechanism and provide the basis to improve the test design and therating training.The study shows that:1. The Generalizability Theory and Many-facet Rasch Measurement whichhave their own emphasis can make a good complementary to each other. Ifthe actual situation meets all conditions, they can be used in combination.Or can choose one according to the analysis emphasis.2. The capability assessment reliability can also be improved in thefuture, which is just suitable for the relative decision not for theabsolute decision.3. The main factors affecting teaching capability assessment are whetherthe raters are strict or not and the difficulty of the tasks to differentparticipants.4．By increasing the number of tasks, the accuracy of the teachingcapability assessment can be improved, while the rate of increase isdeclined. When the number of raters is five, the results will meet thetesting requirements well.5. It is proposed to strengthen the training and guidance of the raters,especially guide the pertinence of marking during the interview process.For long-term considerations, we can make a specialized raters team andtry to establish a “raters pool”.