Laporkan Masalah

PENYETARAAN SEKOR TES MASUK PERGURUAN TINGGI

FITRI ANDRIANI,

2005 | Tesis |

The aim of this research was 1)to analize the items of two university entrance test; 2) to arrange, to try out, and to analize the items of anchor test; 3) to do the equating and determines the conversion score of TPA PMDK UNAIR 2004 I and II by Classical and Racsh approach; 4) to estimate the Standard Error of Equating of each method. The subject of this research were the third grade of Senior High School. There were eight Senior High School, included five public school and three private school. Number of subject for trying out of anchor test was 204 and 882 subject for equating process. Try out of TPA PMDK Unair 2004 used respons of participant registered as college student at PMDK Unair 2004. Number of respons for TPA PMDK Unair I was 4646, and for TPA PMDK II was 1792. Analize was done by ITEMAN and BIGSTEPS software, supported by SPSS and Excel. Item analysis with Classical approach, resulted on 32 anchor test items, 66 TPA PMDK Unair I items, and 67 TPA PMDK Unair II items that can be used because had a biserial correlation 0.30 and up. For Rasch model approach, in TPA PMDK Unair I, there were 57 items, while TPA PMDK Unair II, there were 56 items that fit with the models, that had out fit between 2 and -2 and / or had negative ptbis. The equating formula with Calssical approach was Y* = 0.8990 (X - 25.6157 ) + 25.8938 with Standard Error of Equating was 0.282. While equating with Rasch model approach was bvB = bvA - 0.1615 with Standard Error of Equating was 0.030. It was shown that score of TPA PMDK Unair I will increased if equating with TPA PMDK Unair II. Based on Standard Error of Equating , it was shown that equating with Rasch model approach was more accurate than equating with Classical approach. Keywords: Equating, Classical Test Theory, Modern Tes Theory, Standard Error of Equating.

Kata Kunci :


    Tidak tersedia file untuk ditampilkan ke publik.