Chapter 4 – Research Methods in Clinical Psych Copyright © 2014 John Wiley & Sons, Inc. All rights reserved. Ways of Generating Research Hypotheses Everyday Experience and Observation Professional Experience and Observation Addressing Applied Problems and Needs Previous Research Theory Copyright © 2014 John Wiley & Sons, Inc. All rights reserved. Relationships Between Variables Correlation – the variables are associated in some way Moderation – one variable influences the direction or size of another Mediation - one variable explains the relationship between two others Copyright © 2014 John Wiley & Sons, Inc. All rights reserved. Ethics in Research (some principles) Institutional approval Informed consent Inducements for participation Deception in research Debriefing Humane care for animals Reporting results Copyright © 2014 John Wiley & Sons, Inc. All rights reserved. Important Concepts to Clinical Research Internal validity – controlling for biases External validity – how representative and applicable the study is Statistical conclusion validity – whether the study was designed in a way to adequately test hypotheses through statistical methods Statistical versus clinical significance – Effect size Copyright © 2014 John Wiley & Sons, Inc. All rights reserved. Clinical Research Designs Case Study – detailed description of a case, which can allow for a great deal of hypothesis generating – Down side: many threats to internal validity making it difficult to generalize Single case designs A-B (two measures of symptoms pre and post treatment – A-B (designs with a number of clients) – A-B-A (designs with one person at different time points) Copyright © 2014 John Wiley & Sons, Inc. All rights reserved. Clinical Research Designs Correlational designs – association of variables (most common) – Note even DSM group v control is a correlational design – ‘Median splits’ on a particular variable – Factor analysis – the underlying structure of a variable – Mediator v Moderator designs – Structural equation modeling Copyright © 2014 John Wiley & Sons, Inc. All rights reserved. Clinical Research Designs Quasi-experimental designs – comparing groups when random assignment is not available or ethical Experimental Designs (the ‘gold standard’ in clinical research) – random assignment and experimental manipulation – Randomized controlled trials (RCT) Copyright © 2014 John Wiley & Sons, Inc. All rights reserved. Clinical Research Designs Meta-analysis – summarizing several similar studies through a statistical analysis – Effect size – a statistical measure of how strong the experimental effect is (i.e., statistical significance does not say how strong the effect is, only that it is unlikely to have randomly occurred) Copyright © 2014 John Wiley & Sons, Inc. All rights reserved. Selecting the Participants Importance of selecting a representative sample (age, gender, SES, ethnicity, etc.) Sampling strategy – how participants are chosen/recruited – Probability sampling – requesting participants from say, every 10th person in a neighborhood – Non-probability sampling – recruiting through a method that will reach as many people as possible Copyright © 2014 John Wiley & Sons, Inc. All rights reserved. Selecting the Participants Setting the sample size – making sure there are enough participants to detect differences in groups – Low power – Sufficient sample size – Statistical significance – Clinical significance Copyright © 2014 John Wiley & Sons, Inc. All rights reserved. Selecting the Measures Self-Report Measures – participant report. Informant-Report Measures – report by someone who knows the participant Rater Evaluations – research assistant rates the participant Performance Measures – how a participant does on a task (e.g., reaction time or behavioral measure) Copyright © 2014 John Wiley & Sons, Inc. All rights reserved. Selecting the Measures Projective Measures – participant responses to ambiguous stimuli, which may reflect the internal state of the participant Observation of Behavior - coding systems used to summarize complicated behavior Psychophysiological Measures – a range of measures reflecting biological markers (e.g., heart rate, blood pressure, neural activity) Archival Data – Data stored for some other purpose used for research (e.g., police records, health care utilization records, and academic records) Copyright © 2014 John Wiley & Sons, Inc. All rights reserved. Psychometric Properties of the Measures Reliability Internal Consistency: How homogeneous the test items are Test-Retest Reliability: How stabile over time scores on a measure are Inter-Rater Reliability: The consistency of scores on a measure across different raters or observers. Validity Content Validity: How fully and accurately the measure represents the construct being assessed Face Validity: How much the measure overtly appears to be measuring the construct of interest. Copyright © 2014 John Wiley & Sons, Inc. All rights reserved. Psychometric Properties of the Measures Validity (cont.) Criterion Validity: The association of a measure with a related criterion Concurrent Validity: The association of a measure with other relevant data measured at the same point in time. – Predictive Validity: The association of a measure with other relevant data measured at some future point in time. Copyright © 2014 John Wiley & Sons, Inc. All rights reserved. Psychometric Properties of the Measures Validity (cont.) Convergent Validity: The association between a measure and either other measures of the same construct or conceptually related constructs Discriminant Validity: The association between measures that, conceptually, should not be related Incremental Validity: The extent to which a measure adds to the prediction of a criterion beyond what can be predicted with other measurement data Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.