Bayesian Posterior Estimation of Logit Parameters With Small Samples  

Oleh:

Galindo-Garre, Francisca ; Vermunt, Jeroen K. ; Bergsma, Wicher

Jenis: Article from Bulletin/Magazine
Dalam koleksi: Sociological Methods and Research vol. 33 no. 01 (Aug. 2004), page 88.
Topik: Small samples; logith models; Bayesian estimation; prior distributions;

Isi artikel

when the sample size is small compared to the number of cell in a contingency table, maximum likelihood estimates of logit parameters nd their associated stamdard errors may not exist or may be biased. This problem is usually solved by " smoothing" the estimates, assuming a certain prior istribution for the parameters. this article investigates the performance is point and interval estimates obtained by assuming various prior distributions. the authors focus on two logit parameters of a 2x2x2 table: the interaction effect of two predictors on a response variable and the main effect of one of two predictors on a responses variable, under the assumption that the interaction effect is zero. the results indicate the superiority of the posterior mode to the posterior mean.

Opini Anda

Klik untuk menuliskan opini Anda tentang koleksi ini!

Kembali