PENELITIAN EKSPERIMENTAL
HEALTH /
MEDICAL
STUDY
OBSERVASIONAL EXPERIMEN
(obs. In nature) (manipulation & observation)
DESCRIPTIVE ANALYTIC EXPERIMENT
(ekspl-desk for (analysis of (analysis of
phenomena) dynamical CAUSATION
Phenomena)
CROSS-SECTIONAL STUDY LONGITUDINAL STUDY
(point-time approach) (period time approach)
CASE CONTROL STUDY COHORT STUDY
(Backward in time) (Cause-effect)
EXPERIMENTAL STUDIES
A. TRUE EXPERIMENTAL;
B. QUASI EXPERIMENTAL;
C. PRA-EXPERIMENTAL.
(1) The One-Shot Case Study;
(2) The Group Pre and Post Design;
(3) The Static Group Comparison Control Group Only Design
(4) Randomized Control Group Pre-Post Test Design;
(5) Randomized Solomon Four Group Design;
(6) Factorial Design;
(7) Post-test only design;
(8) The Same Subject Design;
(9) Design by level;
(10) Time series Design;
(11) The Non Equivalent Control Group Design.
PENELITIAN EKSPERIMENTAL
• Dua keadaan yang identik, sejajar
dalam segala hal, kecuali satu yg
berbeda (perlakuan/manipulasi);
• Perlakuan prosedur yang efeknya
akan diukur dan dibandingkan;
• Mengadakan percobaan untuk
melihat hasil (variabel dependen);
• Menentukan hubungan kausal di
antara variabel-2 yang diteliti.
TRUE EXPERIMENT
Penelitian untuk menyelidiki
kemungkinan hubungan sebab-akibat
dengan memberikan 1/> kel.eks, 1/>
kondisi perlakuan & bandingkan
hasilnya dgn kel.kontrol;
Ciri-ciri :
. Pengaturan variabel perlakuan;
. Menggunakan kel.pembanding;
. Memusatkan pengendalian var.luar;
. Perhatian pada validitas Eks/Int.
LANGKAH-2 TRUE EXPERIMENT
• 1. Survai kepustakaan;
2. Identifikasi masalah;
3. Rumuskan Hipotesis;
4. Definisi pengertian dasar & variabel utama;
5. Menyusun rencana penelitian (susun var.eks,
desain, sampel, perlakuan, instrumen,
kumpul data dan trial run test);
6. Melaksanakan Eksperimental;
7. Managemen data;
8. Interpretasi hasil;
9. Menulis laporan & disiminasi.
(1) RA : Experimental Group
Control Group
Time
(2) - Start of Intervention
- The end of Intervention
(3) X - Intervention
(4) O - Observation / Measurement
O1,2,3 - Observation (1), (2), (3) etc.
(5) Separated line sign of non randomization assigment
Example : RA Exp.Group O1 x O2
Cont.Group O3 O4
Other points related to experiments
• Validity
– An experiment is valid if :
• the observed change in the dependent
variable is, in fact, due to the independent
variable >> internal validity
• if the results of the experiment apply to the
“real world” outside the experimental
setting >> external validity.
(1) The One-Shot Case Study;
(2) The Group Pre and Post Design;
(3) The Static Group Comparison Control Group Only Design
(4) Randomized Control Group Pre-Post Test Design;
(5) Randomized Solomon Four Group Design;
(6) Factorial Design;
(7) Post-test only design;
(8) The Same Subject Design;
(9) Design by level;
(10) Time series Design;
(11) The Non Equivalent Control Group Design.
Study with limitation of characteristic criteria
of true experimental.
THE One-Shot Case Study
INTERVENTION Posttest
X 01
EXP GROUP TIME
X O1
Disadvantages
1. No Control and Internal Validity Control (-)
2. Negative Comparation Group.
Advantages
Developed of idea, Conclusion ??
One Group Pre-Posttest
Design
Pre test INTERVENTION Post test
O1 X O2
Exp group time
Pretest intervention Postest
Exp. group X O1
Kel.Kont O2
RA (-)
Validity challenges :
History
Maturation
Testing
Instrumentation
Advantages :
No Pretest, cost is expensive and
.Interaction pretest and X
STUDY
POPULATION
Selection by
defined criteria
Potential Non participants
participants (do not meet
selection criteria)
Invitation to
participate
Non
Participants
Participants
Randomized
Treatment Control
RA Exp.Group O1 X O2
Cont.Group O3 O4
RA Exp.Group O1 Xa O2
Exp.Group O3 Xb O4
Cont.Group O5 O6
Randomized Controlled trial
(RCT)
Group Pre Interv Post
1.Pretested O1 X O2
2.Pretested O3 O4
3.Unpret X O5
4.Unpret O6
1D = O1, X, H, M O2
2D = O3, H, M O4
3D = X, H, M O5
Treatment Response
R X O1
Compare
R (-) O2
Intervention Response
Cont.Gr. R (-) O1
Exp.Gr.1 R Xa O2
Compare
Exp.Gr.2 R Xb O3
Exp.Gr.3 R Xc O4
(2) NON EQUIVALEN CONTROL GROUP DESIGN
Kel. Eksp. Waktu
O1 X O2
Kel Eqiv. O3 O4
Classical Quasi Experiment
1. Randomisation –-> (Exp group and Control group).
2. R (-) the same characteristic of group (Exp & Control)
3. R (-) >
Group (-) > Time Series ---> Trends ?
4. Time Series (-) Pretest (+) to compare (Pretest – Post
test Design)
5. Pretest (-) limitation of analysis multivariate analysis
6. Validity problem :
• True measurement ?
• Function of equipment ?
• Chalenges of validity internal and external
