TEKNIK ESTIMASI PARAMETER DAN KOMPONEN ERROR PADA DATA PANEL TIDAK LENGKAP DENGAN LSDV, FGLS DAN ANOVA

• Main Authors: , YAYUK SETYANING ASTU, , Prof.Dr.rer.nat, Dedi Rosadi, S.Si., M.Sc.
• Format: Theses and Dissertations NonPeerReviewed
• Published: [Yogyakarta] : Universitas Gadjah Mada 2014
• Subjects: ETD
• Online Access: https://repository.ugm.ac.id/133395/; http://etd.ugm.ac.id/index.php?mod=penelitian_detail&sub=PenelitianDetail&act=view&typ=html&buku_id=74049

This thesis discusses the unbalanced panel data fixed effect and random effect with one-way and two way. In the fixed effects one-way and two-way used method of Least Square Dummy Variable (lSDV) by assuming the intercept varies between individuals and time, whereas the slope is constant. Dummy variable used to determine differences the intercept on the classification individual or time units. For the random effects, differences in characteristics individual and time accommodated on the error component model with unknown error covariance matrix. The method used for one-way is Feasible Generalized Least Square (FGLS). Modifications ANOVA by Wallace and Hussain used for variance component estimation on the error covariance matrix. For two-way ANOVA method was used. Estimated components of variance on the error covariance matrix using the OLS residuals as components of quadratic forms and equating to expectations. In this case study the factors that affect cash dividend represented by ROI, Cash Ratio, Current Ratio, Debt Total Assets, Earning Per Share, Debt Equity Ratio and Dividend Payout Ratio in the period 2009-2013 with 36 of consumer goods manufacturing company in IDX. Key Words : Unbalanced panel data, Fixed effect, Random effect, LSDV, FGLS, ANOVA, Error Components