Project _2 Controlling Threats to Validity by hcj


									Controlling Threats to Validity
                        Chapter 9
                   Friday, July 11
Threats to validity

 Validity can be threatened in many ways
   Presence of confounding variables
   Unrepresentative samples
   Inappropriate statistical tests or violations of
    statistics assumptions
   Subject and experimenter effects

 All of these threats can (and should) be
Types of Control Procedures

 General control procedures
 Control over subject/experimenter effects
 Control through the selection and
  assignment of participants
 Control through specific experimental
General Control Procedures

Preparation of the setting
  Eliminate competing variables
  Increase control over the independent variable
  But remember: a natural setting increases
   external validity
General Control Procedures

Preparation of the setting
• Response Measurement
  –Use reliable and valid measures
General Control Procedures

Preparation of the setting
Response Measurement
• Replication
  –Demonstrates that the finding is consistent
    and robust
Subject & Experimenter Effects

Blind procedures
  Single-blind: The experimenter does not know
   what condition the participant is in
  Double-blind: Neither the experimenter nor the
   participant knows what condition the participant
   is in
     May involve a placebo condition
     A good placebo for psychological research?
Subject & Experimenter Effects

Blind procedures
 Use of objective measures
 Use of multiple observers
Participant Selection

Populations and samples
  General population: all potential participants
  Target population: those participants you are
   interested in
  Accessible population: portion of target
   population that is available to the researcher
  Sample: drawn from the accessible population
    Want a sample that is representative of the
Sampling Procedures

Random sampling
  Every participant has an equal chance of
   being sampled
Sampling Procedures

Random sampling
• Stratified random sampling
  –Random sampling within strata (subgroups)
Sampling Procedures

Random sampling
Stratified random sampling
• Ad hoc samples
  –Random sample from accessible population
  –Must generalize cautiously
  –Should describe sample to help define limits
    of generalization
Participant Assignment

Free random assignment
  Random assignment to groups
Matched random assignment
  Random assignment of participants in
   matched sets to groups
Experimental Research

The highest constraint form of research
  Maximum control over extraneous variables
Relies on manipulation of the independent
  It is important to make sure that any systematic
   differences are due only to the independent
Experimental Design

Main focus of Chapters 10 through 13
Experimental design, coupled with the
 control procedures covered in this
 chapter, will maximize validity
Key elements of experiments
  Two or more groups/conditions (control)
  Random assignment of participants to groups

To top