Help
Options
Didn't know it?
click below
click below
Knew it?
click below
click below
Don't Know
Remaining cards (0)
Know
retry
shuffle
restart
0:00
Embed Code - If you would like this activity on your web page, copy the script below and paste it into your web page.
Normal Size Small Size show me how
Normal Size Small Size show me how
PSYC 204- ASSIGN 3
In class assignment 3
| Term | Definition |
|---|---|
| Matched Samples Experiments | Matched samples increase similarity of groups assigned to different treatment/control conditions |
| Weakness of Random Assignment | Relies on the laws of probability |
| Between Subjects Experiments | -Reliant on random assignment to ensure group similarity prior to treatment -Researchers recruit, then randomly assign to treatment/control condition |
| Priming Effect | Casual exposure to information makes it more likely that it will influence processing of further related information |
| Placebo Group | Non-treatment group |
| Maturation Effects | Change in participants over time (natural development) |
| Matched Sampling: Within-Subjects Design | A.K.A. Repeat measure design Measure same people prior to treatment and compare the pre-test data with data from post-treatment measurements -Assume that any change is because of treatment --Hold all potential types of individual variation constant |
| Problem with Within-Subjects Design | There's no way to rule out alternative explanations for any observed differences (placebo or maturation) |
| "One-Off" Experiment | Experiment that is only conducted once and is not intended to be repeated, study unique outcome -Lack of scientific rigor |
| Matched Pairs Designs | -Used if more than one treatment condition or concerns about maturation effect. -Match participant data (age, experience, etc) with similar participant, then randomly assign them to different groups -Eliminates maturation effect and carryover effect |
| "Carry-Over" Effect | The idea that the influence of a past condition on a participant's response to a new condition |
| Factors to "Carry-Over" Effect | -Fatigue -Learning -Previous task context |
| Mixed Design (With and Between Subjects) | -Treatment and control are tested twice to test for maturation effects --If maturation effects are unlikely, no need to measure control twice ---THINK BANDURA |
| Importance of Replication Studies (Slight Variation) (Bandura) | Account for potential confounding variables that might have altered or qualify initial findings |
| Statistical Controls | -Used after data has been collected -When performing data analysis, the research holds one or more variables constant |
| Simpson Effect | By holding variation constant, researchers can identify experimental effect, then test if conclusions are valid |
| Partial Counterbalancing technique (Within-Subjects) | Randomize sequence/order that treatments occur e.g. 1.Treat1 – Rdm – Rdm 2. Treat2 – Rdm – Rdm 3.Treat3 – Rdm – Rdm 4.Rdm – Treat1 – Rdm 5.Rdm – Treat2- Rdm 6.Rdm – Treat3 – Rdm 7.Rdm- Rdm – Treat1 8.Rdm- Rdm – Treat2 9.Rdm- Rdm – Treat3 |
| Counterbalancing technique (Between-Subjects) | Eliminates treatment order effects entirely at the cost of introducing unwanted random variation |
| Drawback: Counterbalancing technique | May require unrealistically large sample sizes depending on the number of treatments |
| Partial Counterbalancing | -Order sequences are randomized -Researchers ensure balance in terms of when each treatment in a sequence occurs |
| Counterbalancing technique (Within-Subjects) | Changes sequence/order that treatments occur |
| Drawback: Random Counterbalancing | Might not cover every possible combination of treatment sequences |
| Complete Counterbalancing technique | Grouping participants in terms of treatment order -More treatments=More subgroups |
| Compensatory Equalization | When a control group becomes aware of the treatment and demands equal treatment |
| Fix Compensatory Equalization | Provide control group with treatment once the study is concluded -Less applicable when time is a factor |
| Compensatory Rivalry | When control group becomes aware that they are the control group so they make an effort to match the treatment group in terms of their "score" on measures of the independent variable --Overcompensating |
| Resentful Demoralization | Control realizes they aren't receiving treatment and become less motivated because they see the treatment group as having an advantage over them --Adds unwanted variation to experiment |
| Single Blind | Only participants don't know -Deception |
| Controlling for Participant Interpretation Effects | Make placebos believable -Deception |
| Deception | Used to conceal or disguise the presence of a treatment -Ethical concerns |
| Ethical Deception | -Minimal/no risk to participants -Altering consent requirements -If it's impossible/impractical to get answers without it -Participants are provided with debrief and opportunity to refuse consent/request withdrawal of their data from the study |
| Partial Disclosure | -Deception by omission -Incomplete disclosure of information about research to participants --Hypothesis omitted --Describe around topic- conformity=group dynamics |
| Deception Classification: Misleading | Mild deception |
| False Feedback | Self-concept/efficacy research -Careful of negative feedback |
| Double-Blind | Research associate and participants don't know |
| Partial Blind | Experiment conductors are ignorant of some portion of the experiment e.g. person 1 supplies, person 2 measures |
| Triple-Blind | People analyzing data are unaware of research objectives |
| Control Considerations: Rapport | How the person administering treatment presents themselves e.g. friendly or cold |
| Logic of Disconfirmation | We cannot prove that a hypothesis is correct based on a single test using sample data, but we can be confident if there's counter-evidence |
| Test Hypothesis | The one we seek support of empirically |
| Null Hypothesis | Analyze experimental data with opposite hypothesis to see which to reject |
| Significance Testing | Rule out coincidence correlation |
| Significance Test Formula | Difference of group means/sample error=T --Group mean difference cannot be used as basis for determining stat significance |
| Statistics: Confidence | How certain you want to be that your observed results are not due to chance |
| Confidence Level | Usually 0.01 (1% chance that sig result is due to error) |
| Illustration | Testing mean differences using a t-test |
| Benefits: Mixed Method Design | Compromise between practical benefits of WS (fewer participants) and method of BS |
| Difference between QED and CE | Quasi researchers don't have direct involvement in the experiment or manipulation of variables -Do not create their sample groups, deliberately/directly manipulate variables, more naturalistic settings |
| QED equivalent to a "treatment" | Interventions |
| Quasi-Experimental Designs | Treatments arising from circumstances, Naturalism -Internal validity threats |
| Internal Validity Challenge in QED | More difficult to rile out external factors that might affect variation in dependent variable -No genuine control group |
| QED Control/Treatment Condition Selection | Treatments are administered by something/someone other than the researcher e.g. Natural disasters |
Created by:
user-1982862
Popular Psychology sets