Chapter 2: Internal & External Validity

What are characteristics of experiments?

Extreme control;highly unrestricted focus; usually restricted sample

What are problems with this experiment?

Extreme control; constricted; experimental demand

Only experiments can plausibly specifiy causal relations

True

What is this? (heat➔aggression)

Unidirectional causation

What is this? (guilt➔anxiety)(anxiety➔guilt)

Bidirectional causation

What is this?(Marriage rate decline followed by increase in suicides; both effect of economic depression)

Noncausal covariation

What is sun block an example of between sun exposure and sunburn?

moderator

What is a moderator?

Tells us when effects will happen; influences the strength of the relationship

What does mediators tell us?

Tells how or why something happens

What are the threats to internal validity? (MRS SMITH)

History; Maturation, Selection, Statistical Regression, Maturation, Instrumentaton; Testing, History

What is the goal of experimentation?

Eliminate threats to internal validity

What does the success of two group experiment designs depends on?

depends primarily on the assumption that experimental and control groups are equivalent on all factors expect exposure to the independent variable

Experimental Methods

Techniques of inferring causal relations, in which participants’ actions are affected by the controlled manipulation of variables determined by the researcher.

Quasi-Experimental Methods:

Variants of experimental methods that do not involve random assignment, but, as in experiments, participants are exposed to some form of variable manipulation introduced by the researcher or social agents.

Nonexperimental Methods:

In this form of research, no variables are manipulated, but the relations of naturally occurring variables are measured.

Unidirectional causation

A relation in which changes in A produce subsequent changes in B, but changes in B do not influence A.

Bidirectional causation:

 A relation in which changes in A produce changes in B and, in addition, changes in B produce changes in A.

Noncausal covariation:

relation in which changes in A are accompanied by changes in B, because both A and B are determined by changes in another variable,

Moderator:

A third variable that can either augment or dampen the cause-effect relation between variables.

Mediator:

A third variable that serves as an intermediary to help explicate the chain of processes in a predictive relation.

Independent Variable

 A feature or characteristic manipulated independently of its natural sources of covariation to produce different conditions in an experimental study.

Dependent Variable:

A measured outcome or consequence, not manipulated by the researcher, and expected to be influenced by ( or dependent upon) manipulation of the independent variable in an experiment.

Internal Validity

Indicates the extent to which inferences of causation can be made about the obtained relation between an independent variable and a dependent variable.

External Validity:

The extent of generalizability or certainty that results can be applied to other respondent groups, different settings, and different ways of operationalizing the conceptual variables.

Type 1 Error:

the  probability that the observed effect would have been obtained by chance

Statistical significance:

A measure of the probability of obtaining an observed effect by chance, a threat to internal validity.

Type 2 Error:

The probability of failing to reject the null hypothesis when it is actually false ( i.e., there really is a relation between the predictor and outcome variables, but our study failed to detect it at above chance level).

Statistical power

The probability of obtaining a statistically significant effect, if indeed that effect truly exists; dependent on the strength of the relation between variables and the number of observations ( participants)

Confound:

A type of extraneous variable that affects the dependent variable in a way that cannot be separated from the effect of the independent variable, thereby posing a threat to internal validity

History

Differences in outcomes are attributed to events ( unrelated to the experimental treatment) that occurred during the passage of time between the independent and dependent variables.

Maturation:

Changes in outcomes are caused by changes in the internal or physical characteristics of the participants ( e.g., growing older, becoming more tired, less interested). These are termed maturation effects, even though some representatives of this class ( e.g., growing tired) are not typically thought of as being related to physical maturation.

Testing

Scores obtained in a second administration of the dependent variable measure are caused by participants having been exposed to the same measure previously.

Instrumentation

Differences in scores are caused by changes in the properties of the measurement instrument, rather than by changes in the participants being measured. Examples include the instrument being damaged or the measurement device failin

Statistical Regression:

the tendency of extreme scores to regress to the mean of the distribution from which they were obtained. A threat to internal validity that arises as a function of measurement error coupled with the selection of participants based on extreme scores

Selection

If outcomes for two or more groups of participants are being compared, differences are caused by selection procedures employed when participants were assigned to the groups.

Mortality or attrition:

Selection procedures, treatment differences, or issues that yield different proportions of participants dropping out of the study may cause the observed dif- ferences between the groups in the final measurement.

Selection-based interactions

If participants were differentially ( non- randomly) selected to serve in comparison groups, these specially selected groups may experience differences in history, maturation, testing, etc., which may produce differences in the final measurement

Random assignment

An allocation approach that requires that all persons available for a particular research study may participate in either the experimental or the control group, and that only chance determines the group to which any individual is assigned.

Fallible observer:

researcher’s expectation can threaten the internal validity 

Conceptual replication:

 An attempt to reproduce the results of a previous study by using different operational definitions to represent the same constructs

Random selection:

A form of creating a pool of potential research participants that requires that chance determine the selection of participants for a study, thereby helping to assure the generalizability of results from the sample to the entire population of relevant persons.