Post on 18-Jan-2016
BASIC BETWEEN-SUBJECTS DESIGN
CHAPTER OBJECTIVES:
1. Learn how subject are assigned to conditions of a between-subjects experiment and what random assignment accomplishes.
2. Learn the elements of simple two group designs and how to control for confounding
3. Learn how to conduct experiments with more than two groups
4. Understand why one design is selected over another.
EXPERIMENTAL DESIGN
General structure of the experiment Number of treatment conditions Same or different subjects in different treatment
conditions Determined by
The nature of the hypothesis Information the researcher is seeking Practical problems in running the experiment
Questions to consider How many independent variables are there? How many treatment conditions are needed? Will we use same or different subjects?
BETWEEN-SUBJECTS DESIGN
Different subjects take part in each condition of the experiment
Comparison between the behaviors of different groups of subjects
BETWEEN-SUBJECTS DESIGN
Two Groups Two Independent Groups
Experimental group-Control group 2 Experimental group
Two Matched Groups Experimental group-Control group 2 Experimental group
Multiple Groups
SELECTING AND RECRUITING SUBJECTS
The more the sample resembles the whole population, the more likely is it that the behaviors of the sample mirrors that of the population.
How many subjects are needed for between-subjects design? at least 15 to 20 subject in each treatment
condition More comfortably, 30 per treatment
BETWEEN-SUBJECTS DESIGNS (1 IV, 2 GROUPS)
How were the subject assigned? Two Independent Groups Matched Groups
What are the levels of IV? Experimental group – Control group Design Two Experimental group Design
TWO INDEPENDENT GROUPS
Grouped using random assignment
What are the possible treatment conditions (level of the IV) for 2 independent groups?
Experimental group-Control group; Experimental condition-Control condition; Treatment-No treatment; however, “no treatment” does not really mean “no treatment” all the time.
Two Experimental group (No Control group) – gathers more precise information; different values of the IV.
TWO INDEPENDENT GROUPS
When can 2 independent groups be used?
If there is only one independent variable If randomization can be assumed successful
TWO MATCHED GROUPS
Does/must/can not use random assignment 2 groups of subjects assigned by the
researcher by equating them on a particular characteristic that might affect the DV
Matching can be done before or after the experiment
TWO MATCHED GROUPS
How to match? Precision Matching – matched pairs have
identical scores Range Matching – matched pair fall in a specified
rangeNote: Some subjects have to be discarded Rank-order Matching – subject scores are ranked;
adjacent scores are paired.
TWO MATCHED GROUPS
When can 2 matched groups be used?
Presence of a strong extraneous variable Need for more efficient measures of the IV’s
effect When there is a very small number of subjects
BETWEEN-SUBJECTS DESIGNS (1 IV, MULTIPLE GROUPS)
BETWEEN-SUBJECTS DESIGNS (1 IV, MULTIPLE GROUPS)
Sometimes, it takes more than two treatment conditions to make a good test of a hypothesis
When amount or degree of the IV is important Different values of the same variable can produce
different effects May give complete understanding of how the
variable works Each group is run though a different condition One treatment condition may be the control Usual multiple independent groups Matching is possible
ASSIGNING SUBJECTS
Use of block randomization
CHOOSING TREATMENTS
How does one choose how many treatments to use? Theoretical basis Equal or proportional intervals
What will I gain by adding these extra conditions to the experiment?
As a general rule, select the simplest design that will make an adequate test of your hypothesis.
PRACTICAL LIMITS
More subject are needed Takes more time More complicated statistical procedures
Use of Pilot study (mini experiment) pretest selected levels of an IV before conducting the actual experiment.