Psych/Soc: Research Methods Flashcards
Extraneous/confounding variables
Threat to internal or external validity?
Want a _____ control group to rule out extraneous/confounding variables
Extraneous variables not accounted for in the study; another variable offers an alternative explanation for results; lack of useful control; a third variable that affects both the independent variable and the dependent variable, thus potentially obscuring the true relationship between them
Want a homogeneous control group to rule out extraneous/confounding variables
Threat to internal validity
Double blind
To counter placebo effect, neither person administering treatment nor the participants truly know if they are assigned to the treatment or control groups
Sampling bias
Threat to internal or external validity?
Selection criteria is not random
ex. population is not normally distributed
Form of selection bias
If it is not equally likely for all members of a population to be sampled ex. only undergraduate students used but results applied to whole population
Threat to internal validity
Selection bias
General category of systemic flaws in a design
Includes sampling bias, meta-analysis, attrition
Meta analysis
These studies are thought to yield the strongest available evidence on a given topic but…
Form of selection bias so conclusions drawn can be treacherous
Purposely selecting which studies to evaluate, a big picture analysis of many studies to look for trends in data
Attrition
Threat to internal or external validity?
Form of selection bias
participants dropping out of study
Participant fatigue; participants drop out of study
If reason they drop out is non random then this might introduce extraneous variable
Threat to internal validity
Between subjects design vs within subjects design
Between subjects design: comparisons made b/w subjects from one group to another
Within subjects design: compare same group at different time points
Type I error and Type II error
Type II error = false negative
Type I error = false positive (worse)
Null hypothesis
Assume no causal relationship b/w the variables and any effect that they measure, if there is one is due to chance
Want to reject null hypothesis and have a significant difference b/w the two groups
Want a low p value
P-value
What increases the power?
A number 0 to 1 that represents the probability that a difference observed in experiment is due to chance
if p<0.5 then reject null hypothesis
Usually 30 or more participants needed, a larger sample size is preferred because it increases the power
Operationalized Variables/ operational definition
Process of defining how a particular variable will be measured.
Determine independent and dependent variables, specify each
A formal definition used for research that can be measured by experimenters
Ex. Operational definition for sympathetic arousal: HR, BP
for intelligence: IQ, EQ scores
External Validity
A flaw or limitation that might make it difficult to apply our conclusion to the real world
Ex. only students participate in study
Internal Validity
Construct Validity
Internal Validity -> The extent to which we can say the change in outcome variable is due to the intervention. Causality
Leaves doubts about the conclusion because of some inherent flaw in the design
Construct validity -> specifically related to how well-designed an instrument is, how well a study is able to measure what researchers intended to measure
*Difference between Demand Characteristics
(Threat to what kind of validity?), Hawthrone Effect, and Internal Validity?
Social facilitation
Social interference
Tendency of participants to consciously or subconsciously act in ways that match how they are expected to behave which can threaten internal validity
Online:
Hawthorne effect comes from observation. If I stand behind your shoulder as you do a task your performance will change.
Demand characteristics comes from thinking you know what the psychologist wants for results, and you act accordingly.
The difference is that the Hawthorne effect doesn’t result in behavior that you think the researcher would approve of. Rather, it results in altered behavior due to being watched. This is quite similar to social interference and social facilitation.
The Hawthorne effect is a type of measurement bias in which participants change their behavior in response to their awareness of being observed.
In an experiment or research project, demand characteristics are cues that may influence or bias participants’ behavior, for example, by suggesting the outcome or response that the experimenter expects or desires.
Impression management is characterized by deliberate, conscious behaviors intended to control how others perceive oneself, especially by guiding them to attribute desirable traits to oneself.
Social facilitation is the tendency to perform better on simple tasks when one’s efforts can be evaluated by others.
Social interference is the tendency to perform worse on complex tasks when one’s efforts can be evaluated by others.
Impression management
Threat to internal or external validity?
Participants adapt their responses based on social norms or perceived researcher expectations; self-fulfilling prophecy; methodology is not double-blind, Hawthorne Effect
Threat to internal validity
Reliability
Lack of reliability
Threat to internal or external validity?
Reliability -> consistency of measures
Measurement tools do not measure what they purport to, lack consistency (lack of stable consistent results)
Threat to internal validity
Is “experiment doesn’t reflect real world” threat to internal or external validity?
Lab set up that doesn’t translate to the real world, lack of ________
External
Lab set up that doesn’t translate to the real world, lack of generalizability
Can selection criteria be a threat to internal or external validity?
External
Too restrictive of inclusion/exclusion criteria for participation (i.e., sample is not representative)
Can situational effects be a threat to internal or external validity?
External
Presence of lab conditions changes outcome (e.g. pre-test and post test, presence of experimenter, clausterphobia in an MRI machine)
Can the lack of statistical power be a threat to internal or external validity?
External
Sample groups have high variability; sample size is too small
Correlational studies
Type of cross-sectional study
Non-experimental design that explores the relationship b/w two quantitative variables. Most commonly used correlation is Pearson Correlation which assigns a number from -1 to 1 to a pair of variables. If the value is neg, the variables are negatively correlated (so if one increases the other decreases)
Pos valued means positively correlated
value of 0 means no LINEAR relationship but there could be another type of relationship
Causality cannot be inferred
Ethnographic studies
Non-experimental design that is qualitative method/analysis in which researchers immerse themselves into lives of people studying
Twin studies
Analysis of heritability through measuring characteristics of twins
Longitudinal studies
Long term analysis
Multiple measures are made over time on same group
Case studies
Case series
Case-control studies
Case studies - Deep analysis of an individual or single case or example, researchers report their experiences with a certain condition or treatment (good for providing warning for new disease)
Case series - same but for more than 1 case (good for providing warning for new disease)
Case-control studies type of observational study, retrospective, gathering some “cases” or people with an outcome of interest and comparing them to “controls” or people who do not have that outcome of interest, with the goal of identifying differences between the two groups, important in epidemeology for investigating the cause of an emerging disease
(Now compare this to cohort studies)
Phenomenological studies
Self-observation of a phenomenon by researcher or small group of participants
Researcher study self!
survey
Series of Q’s to allow participants to self-report behaviors or tendencies
Archival Studies
Analysis of historical records for insight into a phenomenon
Biographical
Exploration of all the events and circumstances of an individual’s life
Observational
Broad category that includes any research in which experimenters do not manipulate the situation or results
Naturalistic observation of circumstances as they are
Ppl observed and outcomes measured with no attempt to control the outcome
Non experimental designs differ from experimental designs in that…
Experimental studies…
Non experimental designs lack a control group
Experimental studies -researchers directly manipulate an independent variable
Hawthorne Effect
the alteration of behavior by the subjects of a study due to their awareness of being observed.
Mixed-methods research
when qualitative and quantitative methods are both used
Moderating variable
- modulates the strength/intensity of the relationship between two variables. For example, a relationship may generally exist between workplace stress and anxiety. For individuals who engage in healthy coping strategies like exercise, the relationship between stress and anxiety might be weaker than it is for individuals who do not engage in healthy coping. In this example, healthy coping is a moderating variable.
Mediating variable
explains an observed relationship between two variables (causal link between them). It is one that provides a mechanistic link for an observed relationship between two variables. It can be important to understanding why the independent variable affected the dependent variable. Ex. Mediator variables In a study on socioeconomic status and reading ability in children, you hypothesize that parental education level is a mediator. This means that socioeconomic status affects reading ability mainly through its influence on parental education levels.
Negative control
- experimental treatment which does not result in the desired outcome of the experiment.
A placebo drug is an example of a negative control, not a positive control
Positive control
- experimental treatment which is performed with a known factor to get the desired effect of the treatment.
prospective analysis
- data are gathered moving forward in time
Randomized control trial (rct)
Single blinded
Double blinded
RCT is where participants are randomized to either a treatment group or control group
Single blinded -> either researchers or participants blind
Cross-sectional study
take a set of people representative of a pop and measure various things about them at a single point in time and look for correlations among those measurements
ex. opinion polls
protective factors
risk factors are independent variables with a higher risk of a negative outcome
Protective factors are associated with a lower risk of a negative outcome
Cohort studies
type of prospective study, subset of longitudinal studies and observational studies, in which a group of subjects is assembled according to some organizing principle, often age, and followed up with over time (since data is gathered going forward, this is known as prospective analysis)
- group of subjects is assembled according to some organizing principle - such as age or location - and followed up over time.
Systematic reviews
researcher combs through literature on given topic and critically assesses the outcomes of various studies
Content validity
Criterion validity
Predictive validity
How well the test covers the full scope of content the researchers intend to measure
indicator of how well our test predicts performance of a relevant outcome
sub-type of criterion validity that involves predicting future performance
Precision
How close together experimental measures are
