ANALISIS INSTRUMEN PENELITIAN

1
Validitas Reliabilitas
2
ANALISIS INSTRUMEN PENELITIAN

1. UJI VALIDITAS
2. UJI RELIABILITAS

3
(No Transcript)
4
VALIDITAS

• Validitas (validity, kesahihan) berkaitan dengan
  permasalahan apakah instrumen yang dimaksud
  untuk mengukur sesuatu itu memang dapat mengukur
  secara tepat sesuatu yang akan diukur tersebut.
  Secara singkat dapat dikatakan bahwa validitas
  alat penelitian mempersoalkan apakah alat itu
  dapat mengukur apa yang akan diukur.

5
Validity Types

• Criterion (Concurrent or Predictive) Validity
  focuses on how well the instrument compares with
  external variables considered to be direct
  measures of the characteristic or behavior being
  examined. Intelligence test scores used to
  predict future performance are an example of
  criterion validity. (expert judgement)
• Content Validity is concerned with how well the
  content of the instrument samples the kinds of
  things about which conclusions are to be drawn.
  For example, a test of school readiness can be
  examined to see whether the content relates to
  knowledge and skills expected of first grade
  children following the kindergarten program
  (based on literature) (expert judgement)
• Construct Vailidity involves the extent to which
  certain explanatory concepts or qualities account
  for performance. For example, a personality test
  can be studied to see how well theoretical
  implications of the typologies account for the
  actual results obtained. (EFA or CFA)
• Face Validity is often used to indicate whether
  the instrument, on the face of it, appears to
  measure what it claims to measure. Here the key
  question is whether the persons making use of the
  instrument accept it as a valid measure in the
  everyday sense of the word? (expert judgement)

6
Validity in term of research finding

• Internal validity is related to what actually
  happens in a study.  In terms of an experiment it
  refers to whether the independent variable really
  has had an effect on the dependent variable or
  whether the dependent variable was caused by some
  other confounding variable. 
• External validity refers to whether the findings
  of a study really can be generalised beyond the
  present study.  External validity can be broken
  down into two types.
• Population validity - which refers to the extent
  to which the findings can be generalised to other
  populations of people.
• Ecological validity - which refers to the extent
  to which the findings can be generalised beyond
  the present situation.

7
EXPLORATORY DAN CONFIRMATORY ANALYSIS

• EXPLORATORY Menguji indikator untuk mendapatkan
  proxi variabel yang benar. Secara teori indikator
  variabel belum jelas atau belum ada. Uji dengan
  SPSS. Item yang akan dimasukkan dalam analisis
  akhir adalah item yang memiliki factor loading
  lebih dari 0,40 (Chia, 1995).
• CONFIRMATORY Menguji kecocokan indikator
  variabel. Secara teori indikator variabel sudah
  jelas dan tegas. Uji dengan Lisrel / Amos.

8
EXPLORATORY ANALYSIS
?
9
CONFIRMATORY ANALYSIS
10
d
?
11
RELIABILITAS

• Reliabilitas (reliability, keterpercayaan)
  menunjuk pada pengertian apakah sebuah instrumen
  dapat mengukur sesuatu yang diukur secara
  konsisten dari waktu ke waktu. Jadi, kata kunci
  untuk syarat kualifikasi suatu instrumen pengukur
  adalah konsistensi, keajegan, atau tidak
  berubah-ubah.

12
RELIABILITAS

• 1. Stabilitas (Test retest)
• Satu tes diujikan dua kali, kemudian
  dikorelasikan hasil tes I dan ke II.
• Konsistensi internal
• Split half (odd even beginning ending)
• Alpha Cronbach
• 3. Equivalensi
• Dua atau lebih tes diujikan sekali, kemudian
  dikorelasikan antar semua paket tes

13
Interpretation Guideline
George, D., Mallery, P. (2003). SPSS for
Windows step by step A simple guide and
reference. 11.0 update (4th ed.). Boston Allyn
Bacon.
14
Table Usage

• Koefisien korelasi hasil penghitungan yang
  diperoleh kemudian dikonsultasikan pada
  tabel-tabel nilai-nilai kritis r.
• rh rt reliable (a 5 /1)
• rh lt rt reliable (a 5 /1)

15
TEKNIK BELAH DUA (SPLIT HALF METHOD)