Lecture 3: Clinical Research Methods

What is a path diagram?

Schematic depicting relationships among variables

What is a correlational hypothesis?

Variables are related in a non-causal fashion

What is a causal hypothesis?

A variable causes/influences another variable

What is a mediational hypothesis?

One variable at least partially accounts for relationship between two other variables

What is a moderational hypothesis?

One variable influences the relationship between two other variables

What is the internal validity of a study?

Extent to which causal interpretations are justified and alternative causal explanations are ruled out

What is the external validity of a study?

Extent to which findings generalize beyond study

What is a case study?

Provides a detailed description of links between variables for small number of people

What is a correlational study?

Examines and quantifies associations between variables for larger number of people

What are the two primary characteristics of a correlation?

Magnitude and direction

What is a scatterplot?

A graph showing the relationship between two variables, where dots on the plot correspond to individual subjects

What is an experimental study?

Examines effect of experimentally manipulated "independent variable" (IV) on "dependent variable" (DV) or "criterion variable" across multiple persons; involves random assignments to levels of IV

What does a meta-analysis study do?

Quantitatively integrates findings from multiple studies asking similar research question

What is a cross-sectional research design?

Evaluates sample at single point in time

What is a longitudinal research design?

Follows same persons/cases over multiple points in time

What is a convenience sample?

A sample of persons who are readily available to participate in a study (typically volunteers)

What is a random sample?

Everyone in the population is equally likely to be in the sample

What does it mean for a measure to be reliable?

Consistent or precise

What does it mean for a measure to be valid?

Accurate

What is internal consistency?

Consistency of responses across items on a measure

What is test-retest reliability?

Consistency of responses over time on a measure

What is inter-rater reliability?

Consistency of judgments across raters on a measure

What is alpha?

A measure of internal consistency reliability

What is kappa?

a measure of inter-rater reliability

What is Content Validity?

Extent to which item content on a measure reflects all aspects of concept of interest

What is Concurrent Validity?

Association of a measure with another measure of a related concept at same point in time is strong

What is Predictive Validity?

Association of a measure with another measure of a related concept at future point in time is moderate to strong

What is Convergent Validity?

Association of measures intended to measure same or similar concepts is moderate to strong

What is Discriminant Validity?

Association of measures intended to assess different concepts is weaker than association of measures intended to assess same concepts

What is Incremental Validity?

Extent to which measure tells us more (or predicts more) than what already knew (or could predict)

What is statistical significance?

Reliability of an effect, typically indicated by p < .05

What is practical significance?

Magnitude of an effect, typically indicated by Cohen's d or correlation coefficient (r)

What is Cohen's d?

A measure of practical significance

Does a p-value tell us about the statistical significance or practical significance of an effect?

statistical significance

Does the magnitude of an effect tell us about the statistical significance or practical significance of an effect?

practical significance

Does a case study, an experimental study, or a correlational study typically show the highest internal validity?

an experimental study

True or False: The strength of a correlation of -.56 is the same as the strength of a correlation of .56.

True

True or False: A perfect negative correlation has a value of 1.0.

False

True or False: Correlations vary between -100 and +100.

False

True or False: Case studies can be useful for generating hypotheses but generally are low in both internal and external validity.

True

True or False: As internal validity decreases, external validity typically increases.

True

True or False: A moderational hypothesis evaluates WHY or HOW one variable impacts another.

False

True or False: A meta-analysis generally has high external validity.

True

True or False: Cross-sectional designs are particularly helpful for increasing our understanding of the development, course, and treatment of psychopathology.

False

True or False: It is almost impossible to obtain a fully representative sample.

True

True or False: Inter-rater reliability tells us about the consistency of responses across items on a measure

False

True or False: As the sample size for a study increases, the statistically significance of an effect generally increases.

True

True or False: Sexual assault is common on college campuses, especially among undergraduate women and undergraduates with non-binary gender identities.

True

True or False: The rates of non-consensual sexual contact by physical force or inability to consent decreased significantly among undergraduate women between 2015 and 2019.

False

True or False: A path diagram depicting moderation of the relationship between X and Y by Z shows a single-headed arrow pointing from X to Y and a single-headed arrow pointing from Z to Y.

False

True or False: A path diagram depicting mediation of the relationship between X and Y by Z shows two single-headed arrows pointing from X to Y and from X to Z, as well as a single-headed arrow pointing from Z to Y.

True

True or False: A path diagram that shows four variables that are all connected by double-headed arrows indicates that the four variables all correlate with one another

True

True or False: The number of dots on a scatterplot indicates the number of conditions in the study.

False

True or False: When considering what a correlation between two variables means, we need to consider both the directionality of the effect as well as the possibility of fourth variables.

False

True or False: Endorsement of rape-supportive attitudes does not predict future sexual aggression.

False

True or False: The extent to which the items on a measure reflect all aspects of the concept being measured is called concurrent validity.

False

True or False: From an ethical perspective, running an anonymous study means that the researcher knows how specific individuals responded but agrees not to share that information with anyone under any circumstances.

False

True or False: From an ethical perspective, an informed consent document must address confidentiality limits and the potential risks and benefits of participating in a study.

True