Search
Program Calendar
Browse By Day
Browse By Time
Browse By Person
Browse By Room
Browse By Unit
Browse By Session Type
Search Tips
Annual Meeting Housing and Travel
Personal Schedule
Sign In
The assessment of instructional features represents a central challenge to educational researchers. One reason for this is that it requires a large number of observed school classes to unbiasedly estimate so-called doubly-latent-multilevel-models (DLMM). Therefore we examine whether the Bayesian framework helps to overcome sample size limitations that result from the frequentist approach. We conducted a simulation study comparing the performance of the two statistical approaches when estimating DLMMs. Additional factors of the Monte-Carlo study are: number level-2 clusters, intraclass correlation, sampling error, number indicators. With regard to the specification of the Bayesian models informative priors are derived from existing large scale data sets. From the findings, we draw up recommendations for applying DLMM in case of small level-2 cluster sizes.