Study Guides (299,240)
CA (140,909)
UBC (4,804)
PSYC (710)
Quiz

PSYC 217 Study Guide - Quiz Guide: Scatter Plot, Observational Error, Sample Size Determination

14 pages85 viewsWinter 2014

Department
Psychology
Course Code
PSYC 217
Professor
Catherine Rawn
Study Guide
Quiz

This preview shows pages 1-3. to view the full 14 pages of the document.
Single Case, Quasi-Experimental, Dvlpmtal Research
Recall: Internal Validity
- Does the IV cause change in DV? (Or is something else at work)
- A continuum
No internal validity High internal validity
No causal claim possible Strong causal claim
3 Components of Causal Relationships:
- Temporal precedence cause comes first
- Covariation of cause & effect
- Eliminate plausible alternative explanations
Key Features of True Experiments That Help Achieve Internal Validity
- Manipulated IV(s) at least 2 levels to enable comparison
- Measured DV(s)
- Experimental control over extraneous variables
No confounds
Random assignment to conditions of IV
RESEARCH DESIGNS FOR SPECIAL CIRCUMSTANCES
Single Case
Quasi-Experiments
Developmental
ABA Reversal design
ABAB Reversal design
Multiple baseline design
One-group posttest only
One-group pretest-posttest
Nonequivalent control group
Nonequivalent control group pretest-posttest
Multiple repeated measures:
Interrupted time series
Control series
Longitudinal
Cross-sectional
Sequential
EXAMPLES OF SPECIAL CIRCUMSTANCES (often in applied contexts)
Single Case
Quasi-Experiments
Developmental
Therapist testing whether a
theap edues this patiet’s
symptoms
You want to train your dog to stop
peeing on the floor
A researcher explores decision
making abilities of someone w/a
specific brain injury
Community agency evaluating how
well an intervention program works
Teacher evaluating whether a
classroom activity enhances
learning
Go’t ealuatig effets of a e
law
Researcher comparing
attention span of older vs.
younger ppl
University tracking critical
thinking skills as students
progress through their degrees
find more resources at oneclass.com
find more resources at oneclass.com
You're Reading a Preview

Unlock to view full version

Only half of the first page are available for preview. Some parts have been intentionally blurred.

SINGLE CASE EXPERIMENTAL DESIGNS
- Single P clinical, counseling, educational, etc
- Developed from need to determine if a manipulation had effect on a single P
- Baseline control period (A): behaviour is measured over time
- Treatment period (B): manipulation introduced, continued observation behaviour
- Change in behaviour from baseline to treatment periods = manipulation is effective
- Lack of generalizability
Reversal Designs: manipulation can be undone/reversed
- ABA/Withdrawal design: A baseline period B treatment period A period after treatment removed
o Long treatment periods
- ABAB design: treatment introduced a second time (single reversal is not powerful enough)
Multiple Baseline Designs: behaviour changes only after manipulation is introduced
- Observed under multiple circumstances to rule out possibility that other events were responsible
- Multiple baseline across subjects: behaviour of several subjects measured over time (manipulation at diff times)
- Multiple baseline across behaviours: diff behaviours of single subject measured over time
QUASI-EXPERIMENTAL DESIGNS
- Quasi-exp designs: need to study effect of an IV in settings in which the control of true exps cannot be achieved
- Lack control conditions & random assignment lower internal validity than a true experiment
One-Group Posttest Only Design:
- Lacks control/comparison group, lack internal validity
- Does not allow us to draw causal inferences about effect of IV on DV b/c many potential alternative explanations
- Do not know if DV would have been equal, lower, higher w/o program
One-Group Pretest-Posttest Design:
- Measure Ps before manipulation (pretest) and again afterwards (posttest)
- Many potential alternative explanations
Participant Group Pretest Treatment Posttest
PSYC218 Tutoring Students who Avg Sample Final Tutoring Avg Final Exam
Program offered sign up Exam Score: 50% Program Score: 70%
*The program might have had an effect
*Program evaluation: research on programs that are proposed to achieve some +ve effect on a group of indiv
BEWARE: What alternative explanations are there? What could you do to improve this study? Why would that
improve it?
find more resources at oneclass.com
find more resources at oneclass.com
You're Reading a Preview

Unlock to view full version

Only half of the first page are available for preview. Some parts have been intentionally blurred.

Non-equivalent Control Group Design:
- Ps in the two conditions (exp & control group) are not equivalent b/c not randomly assigned confounding
- Selection differences: Ps who form the 2 groups in the exp are chosen from existing natural groups
Non-equivalent Control Group Pretest-Posttest Design:
- Most useful internally valid
- Not randomly assigned, but have pretest scores to see if the groups were the same on the pretest
- Look at changes in scores from pretest to posttest
Interrupted Time Series Design: Impact of Ontario Drunk Driving Law
- Archival data (no manipulation)
- Before and after a particular time
- Look over time
- Interruption = when law was enforced
- Rules out History threat to internal validity
Non-equiv pretest posttest Count for every year
Control series take average
- Multiple pretests & postests (instead of just one)
- Eaie effets of atual aipulatios i soiet
- Eg. passing of laws examine traffic fatality rates over period of time, before & after law was passed
Control Series Design:
- Improve interrupted time series design have a control (to compare other scores with)
DEVELOPMENTAL RESEARCH DESIGNS
- Longitudinal method: same group of ppl observed at diff points in time as they grow older
Best way to study how scores on a variable at one age are related to another variable at a later age
Greater advantages (but expensive & difficult)
Eg. study how home environ of children at age 5 is related to school achievement at age 13
- Cross-sectional method: persons at diff ages are studied at only one point in time (more common, less exp)
Confounds age & cohort effects
Cohort: group of ppl born at about same time, exposed to same events
- Sequential method:
1st phase = cross-sectional method (eg. study groups of 55 & 65yr olds)
2nd phase = use longitudinal method each indiv tested again
find more resources at oneclass.com
find more resources at oneclass.com
You're Reading a Preview

Unlock to view full version


Loved by over 2.2 million students

Over 90% improved by at least one letter grade.