RELIABILITAS DAN VALIDITAS KONSTRUK TES INTELIGENSI KOLEKTIP INDONESIA TINGKAT TINGGI (TIKI-T)

Apriliantri, - (2017) RELIABILITAS DAN VALIDITAS KONSTRUK TES INTELIGENSI KOLEKTIP INDONESIA TINGKAT TINGGI (TIKI-T). S1 thesis, Universitas Pendidikan Indonesia.

Abstract

Penelitian ini bertujuan untuk mengetahui reliabilitas dan validitas konstruk Tes Inteligensi
Kolektip Indonesia tingkat Tinggi (TIKI-T). Reliabilitas/keandalan alat Tes Inteligensi Kolektip
Indonesia tingkat Tinggi (TIKI-T) diuji dengan analisis data split half reliability, formula Kuder-
Richardson-20 dan Alpha Cronbach. Validitas TIKI-T menggunakan validitas konstruk yaitu
analisis factor dan validitas by known group. Validitas konstruk diuji dengan analisis data
confirmatory factor analysis model tiga faktor (space and non verbal reasoning, general scholastic
aptitude dan accuracy and speed), exploratory factor analysis setiap subtes dari alat tes TIKI-T
yang bertujuan untuk melihat validitas konstruk TIKI-T. Selain itu peneliti melihat validitas
konvergen TIKI-T dengan Intelligenz Structur Test (IST) menggunakan metode exploratory factor
analysis. Validitas by known-groups dilakukan untuk membandingkan perbedaan jenis kelamin
pada setiap subtes TIKI-T diuji dengan analisis data T-Tes. Kualitas item pada Tes Inteligensi
Kolektip Indonesia tingkat Tinggi (TIKI-T) diuji dengan Item Respon Theory (IRT) 2 Parameter
Logistik (2PL), yang berupa a parameter diskriminasi (discrimination parameter) dan b parameter
kesulitan (difficulty). Sampel penelitian berjumlah 316 subjek kategori akhir SMA, SMK dan
mahasiswa perguruan tinggi. Hasil yang diperoleh menunjukkan bahwa reliabilitas TIKI-T
memiliki estimasi yang baik. Interkorelasi pada setiap subtes TIKI-T terdapat 3 faktor/ komponen
yaitu space and nonverbal reasoning, general scholastic aptitude dan speed and accuracy yang
menunjukkan fit dengan model (teori yang digunakan sebagai dasar penyusunan alat ukur).
Interkorelasi TIKI-T dan IST menghasilkan 5 faktor/ komponen yang terbentuk, yaitu verbal and
reasoning, spatial and reasoning, number, speed and accuracy dan spatial and memory. Kualitas
item pada TIKI-T memiliki keberagaman tingkat kesulitan dan diskriminasi. Tidak adanya
perbedaan antara jenis kelamin perempuan dan laki-laki pada setiap subtes TIKI-T, hal ini berarti
bahwa tes inteligensi ini tidak bias jenis kelamin.----------This study aims to determine the reliability and construct validity of Tes Inteligensi Kolektip
Indonesia tingkat Tinggi (TIKI-T). Reliability of Tes Inteligensi Kolektip Indonesia tingkat Tinggi
(TIKI-T) was tested with split half reliability analysis, Kuder-Richardson formula-20, and Alpha
Cronbach. TIKI-T validity is measured by construct validity this is factor analysis and by known
group validity. Construct validity was tested by confirmatory factor analysis, three-factor model
(space and nonverbal reasoning, general scholastic aptitude, and also accuracy and speed), and
exploratory factor analysis of each subtest of TIKI-T, which aims to test the construct validity of
TIKI-T. In addition, the researchers looked at the convergent validity of TIKI-T with Intelligenz
Structur Test (IST) by exploratory factor analysis method. Validity by known-groups was
performed to compare sex differences in each TIKI-T subtest tested by T-Test data analysis. Items
quality on Tes Inteligensi Kolektip Indonesia tingkat Tinggi (TIKI-T) were tested by the Item
Response Theory (IRT) 2 Parameter Logistic (2PL), which is in the form of a discrimination
parameter and b difficulty parameter. The sample of the research used 316 final grade of senior
high school, vocational high school, and college students. The results show that the reliability of
TIKI-T has a good estimate. Intercorrelation on each subtest TIKI-T shows 3 factors/ components
(space and nonverbal reasoning, general scholastic aptitude, speed, and accuracy) is fit with the
model (basic theory of instruments). Intercorrelation of TIKI-T and IST resulted in 5
factors/components formed, namely verbal and reasoning, spatial and reasoning, number, speed
and accuracy and spatial and memory. The quality of items on TIKI-T has a diversity in levels of
difficulty and discrimination. The absence of sex differences in each subtest of TIKI-T shows that
this intelligence test is not biased by gender.

	Text S_PSI_1301140_Title.pdf Download (815kB)
	Text S_PSI_1301140_Abstract.pdf Download (184kB)
	Text S_PSI_1301140_Table_of_Content.pdf Download (551kB)
	Text S_PSI_1301140_Chapter1.pdf Download (317kB)
	Text S_PSI_1301140_Chapter2.pdf Restricted to Staf Perpustakaan Download (706kB)
	Text S_PSI_1301140_Chapter3.pdf Download (452kB)
	Text S_PSI_1301140_Chapter4.pdf Restricted to Staf Perpustakaan Download (2MB)
	Text S_PSI_1301140_Chapter5.pdf Download (190kB)
	Text S_PSI_1301140_Bibliography.pdf Download (414kB)
	Text S_PSI_1301140_Appendix.pdf Restricted to Staf Perpustakaan Download (14MB)

Item Type:	Thesis (S1)
Additional Information:	No.Panggil:S PSI APR r-2017; Pembimbing: I. Medianta Tarigan, II. Helli Ihsan; NIM: 1301140
Uncontrolled Keywords:	Tes Inteligensi Kolektip Indonesia tingkat Tinggi (TIKI-T), exsploratory factor analysis, confirmatory factor analysis, item respon theory, split half reliability, formula kuder- richardson-20.
Subjects:	B Philosophy. Psychology. Religion > BF Psychology L Education > L Education (General)
Divisions:	Fakultas Ilmu Pendidikan > Program Studi Psikologi
Depositing User:	Mrs. Santi Santika
Date Deposited:	28 Aug 2018 08:04
Last Modified:	28 Aug 2018 08:04
URI:	http://repository.upi.edu/id/eprint/31380

Actions (login required)

View Item