13.07.2015 Views

Contents

Contents

Contents

SHOW MORE
SHOW LESS

You also want an ePaper? Increase the reach of your titles

YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.

CHAPTER 6: Independent Groups Designs 219combination of more than one independent variable in one experiment requiresthe use of a complex design, which we will describe in Chapter 8. For now, recognizethat drawing causal inferences based on the natural groups design canbe a treacherous enterprise. Although such designs are sometimes referred to as“experiments,” there are important differences between an experiment involvingan individual differences variable and an experiment involving a manipulatedvariable.SUMMARYResearchers conduct experiments to test hypotheses derived from theories,but experiments can also be used to test the effectiveness of treatments orprograms in applied settings. The experimental method is ideally suitedto identifying cause-and-effect relationships when the control techniquesof manipulation, holding conditions constant, and balancing are properlyimplemented.In Chapter 6 we focused on applying these control techniques in experimentsin which different groups of subjects are given different treatments representingthe levels of the independent variable (see Figure 6.6). In the randomgroups design, the groups are formed using randomization procedures suchthat the groups are comparable at the start of the experiment. If the groupsperform differently following the manipulation, and all other conditions wereheld constant, it is presumed that the independent variable is responsible forthe difference. Random assignment is the most common method of formingcomparable groups. By distributing subjects’ characteristics equally across theconditions of the experiment, random assignment is an attempt to ensure thatthe differences among subjects are balanced, or averaged, across groups in theexperiment. The most common technique for carrying out random assignmentis block randomization.There are several threats to the internal validity of experiments that involvetesting independent groups. Testing intact groups even when the groups arerandomly assigned to conditions should be avoided because the use of intactgroups is highly likely to result in a confounding. Extraneous variables, such asFIGURE 6.6In this chapter we introduced three independent groups designs.Independent Groups DesignsHow were the groups formed?RandomassignmentMatching“By nature” (Individualdifferences variables)Randomgroups designMatchedgroups designNaturalgroups design

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!