1. UKM Learning and Research Repository
2. Master Thesis / Tesis Master
3. Faculty of Education / Fakulti Pendidikan

Title:	Kebolehpercayaan dan kesahan konstruk Instrumen Kecerdasan Intelek secara dalam talian
Authors:	Tuan Norasiah Tuan Ibrahim (P47630)
Supervisor:	Siti Rahayah Ariffin, Prof. Dr. Datin
Keywords:	Intelligence levels
Issue Date:	10-Oct-2011
Description:	Kajian ini bertujuan untuk meneliti kebolehpercayaan dan kesahan konstruk Instrumen Kecerdasan Intelek secara dalam talian. Instrumen ini telah dibina oleh sekumpulan penyelidik Universiti Kebangsaan Malaysia. Instrumen ini terdiri daripada 13 konstruk kecerdasan yang mengandungi 370 item iaitu menyusun blok, persamaan, tempoh menyusun digit, konsep gambar, perbendaharaan kata, urutan nombor dan abjad, penaakulan matriks, kefahaman, carian simbol, pemerhatian gambar, maklumat, aritmetik dan penaakulan perkataan. Responden kajian terdiri daripada 497 pelajar prasiswazah di salah sebuah Institusi Pengajian Tinggi Awam di Malaysia. Data dianalisis menggunakan program Winsteps yang mengaplikasi Model Rasch berasaskan IRT (Item Response Theory). Analisis kebolehpercayaan dilakukan dengan melihat nilai kebolehpercayaan item dan individu serta indeks pemisahan item dan individu. Analisis kesahan konstruk dilakukan dengan melihat indeks korelasi Point Measure, statistik kesesuaian dan analisis pemetaan kesukaran item – kebolehan individu. Hasil kajian menunjukkan nilai kebolehpercayaan item bagi 13 konstruk ialah di antara 0.69 hingga 0.99 manakala nilai kebolehpercayaan individu pula adalah antara 0.00 hingga 0.92. Indeks pemisahan item bagi kesemua konstruk menghasilkan 2 hingga 12 kumpulan strata manakala indeks pemisahan individu pula menghasilkan 0 hingga 3 kumpulan strata. Analisis kesahan konstruk mendapati semua item dalam setiap konstruk mempunyai nilai ukuran korelasi (PTMEA Corr) positif kecuali item q072. Kajian ini menunjukkan 231 item adalah sesuai dengan Model Rasch (julat MNSQ 0.50-1.50), 70 item overfit (<0.50) dan 69 item underfit (>1.50). Kajian ini mencadangkan item-item overfit dalam instrumen ini patut digugurkan manakala item-item underfit perlu diperbaiki atau diubahsuai. Kesimpulannya analisis kebolehpercayaan dan kesahan konstruk menggunakan Model Rasch penting dalam pembinaan item psikometrik bagi memastikan ketekalan sesuatu instrumen. Analisis yang dijalankan juga dapat mengenal pasti keupayaan item-item mendiskriminasi pelbagai tahap kebolehan individu dan melihat sejauh mana item-item tersebut dapat mengukur konstruk yang hendak diukur dalam ujian kecerdasan intelek ini.,The purpose of this study is to examine the reliability and construct validity of Intelligence Quotient Instrument through online services. The instrument was developed by a group of UKM researchers. This instrument consists of 370 items from 13 constructs which are block design, similarities, digit span, picture concept, vocabulary, letter-number sequential, matrix reasoning, comprehension, symbol search, picture completion, information, arithmetic and word reasoning. The respondents involved in this study were 497 undergraduate students from one of the public universities in Malaysia. The Item Response Theory (IRT) is used with the application of Rasch Model of measurement on the analysis procedure. The instrument reliability was examined by analysing item and person reliability as well as the item and person separation index. The construct validity was examined by analysing the point measure correlation, fit statistics and item – person map of abilities and difficulties. The findings showed that the item reliability of 13 constructs were 0.69 to 0.99 whereas the person reliability were 0.00 to 0.92. Item separation index of all constructs produced 2 to 12 groups whereas person separation index produced 0 to 3 groups. Construct validity analysis showed that each item in the constructs was having positive correlations except one item (q072). This study showed that 231 items were fit to Rasch Model (MNSQ range 0.50-1.50), 70 items were overfit (<0.50) and 69 items were underfit (>1.50). The study suggests to eliminate overfit items and for reviewing or modify the underfit items. As a conclusion, the reliability analysis and constructs validity using Rasch model should be emphasized in the development of psychometric items to ensure the consistency of the instrument. The analysis would identify the ability of the items in discriminating levels of person ability and to see whether the hierarchy of item meaning corresponds with the construct we want to measure in this intelligence quotient test.,Master/Sarjana
Pages:	171
Call Number:	BF431.5.M4T833 2011 tesis
Publisher:	UKM, Bangi
Appears in Collections:	Faculty of Education / Fakulti Pendidikan

Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.