241109 rm-p.l.-experimental design

Post on 27-Jun-2015

1.315 views 1 download

Tags:

Transcript of 241109 rm-p.l.-experimental design

EXPERIMENTAL DESIGN

DR. PREMILA LEE,Professor,

Dept of Surgical Nursing & CNE & Research,

College of Nursing,Christian Medical College,

Vellore

INTRODUCTION

• Purpose of scientific research is to determine cause & effect relationship

• To determine the efficacy of an intervention

• Researcher – active

• Provides for evidence based practice

Is there an intervention?

No

Non experimental research

yes

Is there random assignment to treatment groups?

Yes

Experimental

research

Are there measures to compensate for lack of random assignment

No

No Yes

Quasi experimentalPre experimental

TYPES

• True experimental

• Quasi experimental

CHARACTERISTICS

• Manipulation

• Control

• Randomization

MANIPULATION

• Researcher gives intervention or treatment

• Manipulates the independent variables

• Observe the effect

CONTROL

• Used as a basis of comparison

• Possibilities for the counterfactual

- Alternative intervention

- Placebo effect

- Standard care

- Wait list control group

RANDOMIZATION

• Random assignment

• Every subject has an equal chance

• Cluster randomization

• Advantages - No systematic bias - Allows comparison - Increases confidence

EXPERIMENTAL DESIGN

• Pre test - Post test control group design

• Solomon four group design

• Factorial design

• Randomized block

• Cross over design

PRETEST POSTTEST DESIGN

R

E

C

Pre test X Post test

Pre test Post test

SOLOMON FOUR GROUP DESIGN

R

E2

Pre test X Post test

X Post test

C1 Pre testPost test

E1

C2 Post test

FACTORIAL DESIGN

Type of stimulation

Daily exposure

Auditory

A1

Tactile

A2

15 minutes

B1

A1 B1 A2 B2

30 minutes

B2

A1 B2 A2 B2

45 minutes

B3

A1 B3 A2 B3

RANDOMIZED BLOCK DESIGN

• Similar to factorial design in structure

• Two independent variable (factor) - one is not manipulated – known as blocking variable

• Can include more than one blocking variable

CROSS OVER DESIGN

• Repeated measures design

• Within subject design

• Subjects as their own control

• Equivalence among subjects

STRENGTHS

• Powerful to test hypothesis

• Yields highest quality of evidence

• Provides greater confidence

LIMITATIONS

• Impractical or impossible at all times

• Artificiality

• Difficulty in clinical setting

• Hawthorne effect

• Ethical concern

• Costly

QUASI EXPERIMENTAL

• Non equivalent control group before and after design

• Time series design

NON EQUIVALENT CONTROL GROUP BEFORE AND AFTER DESIGN

O1 X O2

O1 O2

E

C

LIMITATIONS

• Lack of randomization

• Control group is not equivalent

TIME SERIES DESIGN

O1 O3O2 O4 X O5 O6 O7 O8

Time series non equivalent control group design

O1 O3O2 O4 X O5 O6 O7 O8

O1 O3O2 O4 O5 O6 O7 O8

PRE EXPERIMENTAL

• One group pre test post test design

• One group posttest- only design

• Non equivalent control group after only design

One group pretest post test design

O1 X O2

ONE GROUP POSTTEST ONLY DESIGN

X O2

POSTTEST ONLY DESIGN WITH COMPARISON GROUP

R

E

Non equivalent comparison group

X Post test

Post test

QUASI EXPERIMENTAL - STRENGTHS

• Practical & feasible

• Adaptable to real world

• Some studies it is the only way

QUASI EXPERIMENTAL - STRENGTHS

• Cause & effect influences less convincing

• Not objective

• Experiment & comparison groups are not equal

CONCLUSION

• To meet the growing demand for evidence based practice