Topic 4 Internal and External Validity

What are the five research strategies?

Descriptive, correlational, non-experimental, quasi-experimental, and experimental

What are the two categories of research strategies?

Not experimental and experimental

Which research strategies are not experimental?

Descriptive, correlational, and non-experimental

Which research strategies are experimental?

Quasi-experimental and experimental

What is a research strategy?

Refers to general approach and goals of a research study

What is a research design?

Refers to how you will implement the strategy

What is a research procedure?

Refers to details on how the study will be done

What is the descriptive research strategy?

Describe current state of a variable(s), does not examine relationship between variables and there is no comparing between groups

What are potential descriptive research questions?

How many hours do you spend studying per week? How many neurons are in the hippocampus?

What is the correlational research strategy?

Measure 2+ variables for every individual/subject in the group, identify correlations between variables but no manipulation of any variables

What are potential topics in correlational research?

Correlation between household income and GPA, correlation between mouse locomotor activity and food intake

What is the non-experimental research strategy?

Examine relationship between variable, looking for difference between two groups, lacking a control group and no manipulation

What are potential topics for non-experimental research?

Difference in resting heart rate between cyclist and swimmer, difference in locomotor activity between single or group-housed mouse

What is the experimental research strategy?

Explain relationship between 2 variables to establish cause-and-effect relationships, compare 2+ groups, there is manipulation and a control group

What are potential topics for experimental research?

Effect of stress on word recall test, effect of endocannabinoid on neuronal firing of hippocampal cell

What is the quasi-experimental research strategy?

Has a control group, but groups cannot be randomly assigned because of preexisting participant/subject variable, does not provide definitive answer about cause-effect

What are potential topics in quasi-experimental research?

Effect of BDNF polymorphism on stress response, reading comprehension before/after literacy program at school A versus school B

What are the two kinds of validity of an experiment?

Internal and external

What is internal validity?

Research has internal validity if it produces unambiguous explanation for the relationship between two variables

What is a threat to internal validity?

Any factor that raise doubts or allow alternative explanations

What is external validity?

Extent to which your experimental result holds true outside your specific study

What is a threat to external validity?

Any characteristic that limits the generalizability of the results

What trends are seen in validity in experimental research strategies?

High internal validity and limited external validity

What trends are seen in validity in not experimental research strategies?

High external validity and limited internal validity

What are the three main categories of threats to external validity?

Generalizing to all participants/subjects, generalizing across study features, generalizing across different measures

What are external validity threats relating to participants or subjects?

Selection bias, convenience sample, volunteer bias, participant characteristics in the demographic, cross-species generalization

What are external validity threats relating to study features?

Novelty effects, multiple treatment interference, experimenter characteristics

What are external validity threats relating to different measures?

Sensitization, timing of measurement

What are the main threats to internal validity?

Confounding and obscuring variables

What are confounding variables?

Change with the independent variable, thus can affect the dependent variable

What are examples of confounding variables?

Assignment bias, environmental variables, time-related variables, observer bias

What are obscuring variables?

Make changes in dependent variable hard to observe and can lead to measurement error

What are examples of threats to internal validity?

Ineffective manipulation, measurement error, and excessive variation in data

What is ineffective manipulation?

Inadequate/ineffective manipulation of independent variable, thus no detectable change in dependent variable

What is measurement error in relation to threats to internal validity?

Too much variation in the measurement due to poor instrumentation, inconsistent training/scoring

What is excessive variation in data in relation to threats to internal validity?

Individual differences contribute the vast majority of variation

What are the three main criteria for generalization?

Controlled comparison rule, sampling rule, operational definitions rule

What is the controlled comparison rule?

When comparing data between experimental and control groups, the only difference must be the independent variable

What is the sampling rule?

Adequate sample size to control for random sampling error, unbiased sample so it is representative of population

What is the operational definitions rule?

Determines what principles can be generalized, single experiment only allows generalization within specific terms of the operational definition

How do you broaden generalizations to entire constructs?

Replicate experiments using different operational definitions of the same construct

What are the two ways to replicate an experiment?

Exact and conceptual replications

What is an exact replication?

Repeat experiment, get same results, and increase confidence in findings

What is a conceptual replication?

Same construct with different experiment/methodology, get consistent results, conclusions better generalized to entire constructs

What happens if you fail to replicate an experiment?

Original results may be invalid or replication may not be exact