Statistical tests and concepts Flashcards
Alpha
The probability of REJECTING the null hypothesis when it is TRUE
AKA Type I error
Beta
The probability of ACCEPTING the null hypothesis when it is TRUE
AKA Type II error
Power
The probability of REJECTING the null hypothesis when it is FALSE, or the likelihood of finding differences b/w conditions when the conditions are in fact truly different
Probability = 1-Beta
Effect size
A way of expressing the difference b/w conditions in terms of a common metric across measures and across studies
Method = based on obtaining the difference b/w the means of interest on a particular measure and dividing this by the common pooled std deviation
Standard deviation
Measure of variation or variability about a mean
Effect size
Refers to the magnitude of the difference b/w 2 or more conditions
Expressed in SD units
Reduce variability by controlling features of the experiment - want to
Dependent on the design of the study
Holding constant
Increases experimental precision
Conditions held constant incl potential sources of influence on subjs bx
held constant when identical or very close to that across subjs and conditions
Eg race, gender, SES, etc.
Controlling sources of variation
Extraneous factors that can contribute to variation
eg time of day, weather, how IV implemented
Let factors vary unsystematically
Disperse factors equally across groups by assigning subjs randomly and by running all subjs in each condition
No systematic bias = well-controlled study
Relationships among variables of interest
Correlation
Risk Factor
Cause
Correlation
Two or more variables associated
Not cause/effect
Risk factor
Character that is an antecedent and increases likelihood of outcome of interest
Cause
One variable infl the other directly
What factors infl the relationship b/w variables; that is the direction or magnitude of the relation?
Moderator –> A variable that info the relationship of two variables of interest
The relationship b/w A and B changes b/c of some other variable eg gender
WHAT - direction or magnitude
How does the phenomenon work; that is, thru what relation or mechanism or thru what process does A lead to B?
Mediator –> The process, mechanism, or means thru which a variable produces a particular outcome
Mechanism elaborates precisely what happens that explains how B results
HOW - mechanism of effect
Deeper understanding
Can we control or alter the outcome of interest?
Intervention –>
Is there something we can do to increase or decrease the likelihood of an undesired outcome (prevention)
Or decrease etc something that has already occurred (tx)
Theory
refers to a conceptualization of the phenomenon of interest
Incl:
Nature, antecedents, causes, correlates, and consequences of a particular aspect of functioning or characteristic
May specify the relations of various constructs to ea other
Different levels of theory
Circumscribed
Conceptual view and model
Broad or narrow
Circumscribed theoretical views
characterize contemporary research in an effort to develop specific models or integrated sets of findings and relations among variables
Explain relations b/w characteristics and a d/o and how lead to other dysfxn
Conceptual underpinnings
Overall orienting view rather than global concepts
Relate approaches insofar as they often present broad orienting views that encompass constructs of interest
eg CBT for depression
Why theory is needed:
- Bring order to areas where findings are diffuse or multiple
- Explain the basis of change and unite diverse outcomes
- Direct our attn to which moderators to study
- Application and extension of knowledge
Overall goal of science
To understand - and this entails connecting empirical relations w statements of mechanisms and processes
How do we put theory into practice?
Operational definitions
Operational definition
Refers to defining a concept on the basis of the specific operations used in the experiment
Limitations of operational definitions
- Over simplification or focus only on part of the concept
- Incl irrelevant features
- Use single measure to define a construct
Multiple operationism
Combined measures that attempt to explore a concept w different operations
May allow an investigator to place more confidence in capturing concept of interest
Compensates for inadequacies of using a single measure
Eg To measure cardiovascular functioning: measure blood pressure, stress test, cholesterol, heart rate
Notion of latent variable used to…
Reflect the idea of a construct represented by several measures
Bc latent variable = defined by several different measures - it is not confounded by measurement error of a single measure
Observed variables
Specific measures
Represent construct of interest
In designing a study, what types of variables should I be looking at?
Environmental/situational
Instrumental
Individual differences
Environmental/situational variables
Varying what is done to, with, or by the subj eg: special intervention type of tx task
Instructional variables
Refer to a specific type of environmental or situational variable
Variations in what participants are told to do or are led to believe thru written/verbal statements about the experiment and their participation
Eg
Comparing different instructions
Subj or indiv differences
Refer to attributes or characteristics of indiv subjects
Not manipulated directly
Varied by selecting subjs w different characteristics, experiences, attributes, or traits
Eg
Age, education, sex, social class, exposure to trauma, scores on personality measures, dx, being twin or only child, etc.
Person w or w/out or w varying degrees of characteristic = compared
