PSYC 200W Lecture Notes - Lecture 12: External Validity, Social Desirability Bias, Internal Validity
28 Jun 2018
School
Department
Course
Professor
Properties of an Experiment
Random assignment to various conditions in a way that ensures initial equivalence
Manipulate 1 or more independent variable (IV) to assess its effects
Control extraneous variables (confounds) for their effects on responses
-> Goal is to determine causality
Manipulate IV:
Independent Variable
oIV must have 2 or more levels - levels are the different values that an IV takes
oLevels are also referred to as conditions
o3 types of IV
Environmental manipulation - environmental modifications of aspects of the
research setting
Instructional manipulation - vary the IV by differing the instructions or
information that the participants receive
Invasive manipulation - creating physical changes in the participants body
through physical stimulation, surgery, or drugs
Control groups receive zero level of IV. They are important when
oResearcher wants to know baseline level of behavior in absence of IV
Pilot test is used to check the levels of IV they plan to use on a couple of participants before the
experiment to make sure they are detected differently by the participants
Manipulation check includes questions to determine if the IV manipulation was successful by
asking the participants questions to check if they detect the levels differently
Subject/participant variables are nonmanipulated variables and cannot be considered as IV
because they are not experimentally manipulated by the researcher such as age, gender
Dependent variable
oResponse being measured
oIncludes observations, self-report measures, or physiological reactions
Condition Assignment:
Between group design/Randomized groups
oGroups must be roughly equivalent at the beginning of the study and any changes in
their behavior at the end of the study should be an effect of the IV
oThus, experimenter should take steps to ensure initial equivalence of groups before
introducing the IV
Simple random assignment - placing participants in such a way that each
participant has an equal probability of being placed in any condition. This ensures that
participants in different groups do not differ.
Matched random assignment - obtain scores on a measure that is relevant to
the experiment and then divide the participants into clusters, after which they will
divide them equally into all the conditions so there is one participant of each cluster in
each condition
Repeated Measures/Within subjects
oSame participants serve in all conditions of an experiment
oCalled repeated measures because each participant is measured more than once and
thus, requires no random assignment
oAdvantages:
find more resources at oneclass.com
find more resources at oneclass.com
Document Summary
Random assignment to various conditions in a way that ensures initial equivalence. Manipulate 1 or more independent variable (iv) to assess its effects. Control extraneous variables (confounds) for their effects on responses. Iv must have 2 or more levels - levels are the different values that an iv takes. Environmental manipulation - environmental modifications of aspects of the research setting. Instructional manipulation - vary the iv by differing the instructions or information that the participants receive. Invasive manipulation - creating physical changes in the participants body through physical stimulation, surgery, or drugs. Control groups receive zero level of iv. Researcher wants to know baseline level of behavior in absence of iv. Pilot test is used to check the levels of iv they plan to use on a couple of participants before the experiment to make sure they are detected differently by the participants.
