Two Approaches For Correcting Correlation Attenuation Caused By Measurement Error : Implications For Research Practice  

Oleh:

Xitao, Fan

Jenis: Article from Journal - ilmiah internasional
Dalam koleksi: Educational and Psychological Measurement vol. 63 no. 6 (2003), page 915-930.
Topik: CORRELATION [STATISTIC]; correlation; reliability; attenuation; confirmatory factor analysis; alpha
Fulltext: 915.pdf (305.0KB)

Ketersediaan

Perpustakaan Pusat (Semanggi) Nomor Panggil: EE30.10 Non-tandon: 1 (dapat dipinjam: 0) Tandon: tidak ada

Lihat Detail Induk

Isi artikel

This study focused on the issue of measuremetn reliability and its attenuation on correlation between two composites and two seemingly different aproaches for correcting the attenuation. As expected, correlation coefficients uncorrected for measurement error are systematically biased downward. For the data conditions examined, the two correction approaches provided not only near identital can unbiased estimates but also near identical confidence intervals for the sampling distribution of the corrected correlation coefficients. The highly comparable results form the two approaches suggest that these two approaches work equally well for these data. It is pointed out that the confirmatory factor analysis (CFA) modelling approach may be less applicable because of more difficult data conditions at the item level in research practice. The findings point to the importance of reporting measurement reliability information in substantive research. The findings further suggest that, in research practice, correction for attenuation should be considered when information about score reliability is available.

Opini Anda

Klik untuk menuliskan opini Anda tentang koleksi ini!

Kembali