Stats - Research Design Flashcards
What design has at least one manipulated IV and subjects are randomly assigned?
True Experimental Design
What design has at least one manipulated IV, and subjects are not randomly assigned
Quasi-Experimental Design
What design has no manipulation of an independent variable?
Non-experimental, observational or passive design
How do you eliminate the possible carryover effects in “within-subject design”?
Counterbalancing
(Latin square = how you can do counterbalancing)
Difference between idiographic and nomothetic
Idiographic: single-subject approach
Nomothetic: group approached
What is the significant problem associated with a single-subject design called that results in highly correlated data from measuring the same person repeatedly?
autocorrelation
What single-subject design has a baseline and then a treatment phase?
AB Design
What’s a threat to an AB Design & what does it mean?
The threat of history which is when something co-occurs at the time of research that impacts the outcomes of the research
Explain the process of an ABAB Design
When a single study research design starts with a baseline phase, then a treatment phase, then a return to baseline phase and then another treatment phase
What does an ABAB design protect against?
A threat of history
What single subject design could have problems with the ethics & a failure of returning to baseline?
ABAB Design
What single subject design helps you try out two different treatments at once?
Simultaneous (alternating) treatment design
What single-subject design is when the treatment is done sequentially, across either subjects, settings or behaviours
Multiple Baseline Design
What design resolves the problems found in AB & ABAB designs?
Multiple Baseline Design
What would the single-subject design be called if a researcher set a criterion of 8 cups per day and then over time, changed it to 6 cups per day to eventually no cups at all
Changing Criterion Design
What type of sampling measurement do you use when a behaviour is not discrete?
Time sampling
What’s momentary time sampling?
Only recording a behaviour that it is present or absent at the precise moment that the interval ends
What’s whole interval sampling?
Observed for the entire interval and only received a checkmark if the target behaviour is exhibited for the full duration of the interval
What type of sampling will tally the number of times a target behaviour occurs or doesn’t occur? It’s also a sampling best for discrete behaviours or ones that don’t happen often.
Event Recording
What’s the research called that evaluates treatment under conditions that only resemble or approximate clinical settings where the problems studied are less severe?
Analogue Research
What’s it called when you do outcome investigations conducted in clinical settings?
Clinical Trials
Cross-sectional research
When you research differences across sections (e.g., different ages)
What does the problem of cohort effects mean in cross-sectional research?
inherent differences between groups that limit the research being done
What type of research has problems with significant expenses and high dropout rates?
Longitudinal Research
What research follows subjects over many years to understand the changes as we age?
Longitudinal research
What’s cross-sequential or cohort sequential research?
Takes several cross-sections and follows them for a briefer timeframe (e.g., 10 years)
What are sampling procedures?
How we select our subjects from the population
What type of sampling provides an equal chance of every member of a population being randomly selected?
Simple Random Sampling
What type of sampling is first divided into strata, and then a random sample of equal size of each strata is selected
Stratified Random Sampling
Proportional Sampling
Individuals are randomly selected in proportion to their representation in the general population
What’s sampling called where you pick every 10th person is selected in a way that’s not bias
Systematic Sampling
What type of sampling involves identifying naturally occurring groups of subjects and randomly selecting certain groups (e.g., classes at a university)
Cluster sampling
What’s a threat in research design?
Something interfering with drawing clear conclusions in research
What causes threats to internal validity?
Factors other than the IV that may have caused the change to the DV
What’s a threat of maturation?
Factors that affect a subject’s performance because of the passing of time
What’s the best control for a threat of history or maturation?
Having a control group within your study
What’s a threat of testing or test practicing?
it comes from subjects familiarity with the testing that affects scores on repeated testing
What’s the best control for Testing or Test Practice threats?
Solomon Four Group Design
What’s involved in the Solomon Four Group Design?
Involves dividing subjects into 4 groups where:
1. one group is measured pre and post and gets the intervention in between
2. second group is measured pre and post but does not get the intervention
3. third group is measured post only and gets the intervention
4. fourth group is measured post only and does not get the intervention
What’s an instrumentation threat?
Changes in observers or the calibration of equipment that’s causing change other than the subjects themselves
The threat of statistical regression or a tendency to the mean
is the tendency for extreme scores (scores very much above or below the mean) to become less extreme (closer to the mean) on retesting without any intervention.
What’s a very common threat to internal validity?
Statistical regression
What’s the best way to manage the threat of statistical regression?
A control group
The threat of selection bias
it’s a threat of non-random assignment, when you don’t randomly assign your subjects like using volunteers
Threat of Attrition or Experimental Mortality
Subjects quitting the experiment and there’s a differential loss of subjects from the groups
Threat of Diffusion
When the no-treatment group actually gets some of the treatment
What’s a threat of construct validity
if our treatment was effective, was it the essential components of treatment that caused the results or was it something else (if it’s something else then it was a threat to construct validity)
Threat of attention & contact with clients (threat to construct validity)
did the treatment group do better because they simply got attention from the therapist
Experimental expectancies is also called what?
The Rosenthal Effect
What is experimental expectancies (Rosenthal Effect)? (threat to construct validity)
Unintentional cues or clues given to the subjects about how they should respond by the experimenter
How do we control experimenter expectancy threat?
keeping the experimenter blind to the control vs non control group
Threat of demand characteristics (threat to construct validity)
factors in the procedures that suggest how the subject should behave, like telling subjects the medication has side effects and then subjects report more side effects
How to control the threat of demand characteristics?
Keeping the subjects blind
What’s the John Henry Effect or compensatory rivalry? (threat to construct validity)
it occurs when the control group tries harder than the experimental group in the spirit of competition
What is a threat to external validity?
When something interferes with the generalizability - can we generalize from our research situation to people who aren’t participating
Threat of sample characteristics (external validity)
The volunteers of the research are somehow different than the rest of the population
Threat of stimulus characteristics (external validity)
the setup of the artificial research not meeting the same characteristics as naturalistic settings
What’s an example of a contextual characteristics threat to external validity
Reactivity which is when subjects behave in a certain way just because they are participating in the research and are being observed
the Hawthorne effect is an example of what type of threat to external validity.
Contextual characteristics
Threat of low power to statistical conclusion validity
diminishing the ability to find significant results like small sample sizes or inadequate interventions
Threat of unreliability of measures to statistical conclusion validity
The measures are unreliable
Threat of variability in procedures to statistical conclusion validity
inconsistency of treatment procedures
Threat of subject heterogeneity
It’s when during research where the goal is to look for differences, the greater the subject variability/diversity/heterogeneity the less likelihood of finding significance
What two concepts minimize moat threats to internal validity but then what’s the tradeoff?
using a control group and random assignment
the more controlled or precise the experiment is, the less generalizable it will be
The greater the internal validity, the lower the _____ _______
external validity