PSY 4372 Lecture Notes - Lecture 7: Odds Ratio, Statistical Significance
Research
Science should be: relevant to the public, applicable, transparent, objective.
Formulating a hypothesis; observations, previous research, theory.
Testing the hypothesis: data, analysis, etc.
Acting on best evidence: conclusions.
Ethics in Research: ethical principles in research and publication, informed consent for all
participation, ethics board approval.
Research Design: depends on the state of knowledge, the question to be addressed, and the
frequency of the phenomena under study
• if it is a new phenomena then you may want to start with correlation designs before you start
on experimental designs to create or view an independent effect
• comparative designs are quasi experimental and occur when you can not separate the
independent variable randomly (ie. gender)
Internal Validity: how much the study is actually getting at the goal and not being confounded by
other factors.
• extent to which interpretations from a study can be justified about alternative explanations
ruled out
• controlled
• laboratory
• experimental
• efficacy studies
External Validity: how it applies to the exterior world.
• extent to which findings from a study can be generalized
• naturalistic
• real-world
• effectiveness studies
Challenges in Research
• balancing internal and external validity
• trying to resolve discrepancies
• opens to unexpected results
Randomized Control Trials (RCTs)
• all participants are assessed prior to intervention
• participants are randomly assigned to a treatment or no treatment group
• all participants are assessed after the intervention period
Efficacy Studies: the extent to which a treatment is actually working, this could be through
randomized study of treatments.
• make sure it is looking solely on the treatment and not other factors
Clinical vs. Statistical Significance
A p value of <0.05 indicates statistical significance, but does this constitute clinical significance?
find more resources at oneclass.com
find more resources at oneclass.com
Document Summary
Science should be: relevant to the public, applicable, transparent, objective. Ethics in research: ethical principles in research and publication, informed consent for all participation, ethics board approval. External validity: how it applies to the exterior world: extent to which findings from a study can be generalized, naturalistic, real-world, effectiveness studies. Challenges in research: balancing internal and external validity, trying to resolve discrepancies, opens to unexpected results. Randomized control trials (rcts: all participants are assessed prior to intervention, participants are randomly assigned to a treatment or no treatment group, all participants are assessed after the intervention period. Efficacy studies: the extent to which a treatment is actually working, this could be through randomized study of treatments: make sure it is looking solely on the treatment and not other factors. Reliable change index: is the change from pretest to post test larger than what you would expect it to be from error or chance.