Term
|
Definition
"Almost"
-almost a true experiment
-to some degree, in some manner, virtual, in some significant sense or degree |
|
|
Term
|
Definition
-An intervention (1 level of the IV)
-an appropriate comparison or control group (another level of IV)
-High degree of control
--> randomization
--> control of all conditions of the experiment |
|
|
Term
| What distinguishes quasi experiment from a true experiment? |
|
Definition
|
|
Term
| What are the threats to internal validity? (List all of them) |
|
Definition
-History
-Maturation
-Testing
- Instrumentation
- Regression
- Subject Attrition
- Selection
- Additive Effects with Selection |
|
|
Term
| History (threat to internal validity) |
|
Definition
| any past experience you've had |
|
|
Term
| Maturation (threat to internal validity) |
|
Definition
Just because time goes by, things change across time
-Ex: woman in case study w/ $ problems; son acting out etc. --> all this could change over time |
|
|
Term
| Testing (threat to internal validity) |
|
Definition
Taking a test has effect on all subsequent testing. Testing will affect everything else that happens.
If give test now (Ex: Beck Depression Test)-->give test later = affect results of later testing |
|
|
Term
| Instrumentation (threat to internal validity) |
|
Definition
changes in your instruments
-Ex: tests, machines, computers
could be new additions of test (something different any tie);
could be something changed with same instrument (Ex: computer really new 1st test, then time passed, not as new and therefore screen not as bright at the time of the second test). |
|
|
Term
| Regression (threat to internal validity) |
|
Definition
go back to mean
extreme scores regress tower their mean
Ex: fantasy football. Regression to mean does not mean the regression to the mean of the NFL, but rather regression to individual mean (how that person typically does on average)
Ex: An "A" student fails a test. Chances are their score will jump quite a bit the next test. Regression to YOUR individual mean. |
|
|
Term
| Subject Attrition (threat to internal validity) |
|
Definition
| selective and mechanical subject loss |
|
|
Term
| Selection (threat to internal validity) |
|
Definition
| placement of participants in groups is not random |
|
|
Term
| Additive Effects with Selection (threat to internal validity) |
|
Definition
- selection-history
- selection-maturation
- selection-instrumentation
- selection- regression in the mean |
|
|
Term
| What does the true experiment fail to solve for? |
|
Definition
1. Contamination
2. Experimenter Expectancy
3. Novelty Effects |
|
|
Term
| In what ways is contamination an issue in true experiments? (explain each) |
|
Definition
1) Resentment: people get upset if they don't receive the best/ real treatment
2) Rivalry: people want to prove that they are in the better group
3) Diffusion of treatments: |
|
|
Term
|
Definition
-what is it the the experimenter wants to happen
-not solved for by true experiment |
|
|
Term
| In what ways does a true experiment not solve for novelty effects? (explain each) |
|
Definition
1. Enthusiasm: excited about study and want to do well
2. Disruption: "not going to try very hard" --> throw wrench in experiment
3. Hawthorne Effect:
a) no matter how conditions are altered, productivity is increased
b) know being observed and change behavior as a result
c) Two potential thought processes behind this
i. They care!
ii. They're judging me!--> boss watching, won't let me get to me |
|
|
Term
| What difficulties are there in conducting a true experiment in Natural Settings? |
|
Definition
1. Obtaining permission from people in authority
2. Obtaining participants
a) Self-selected samples
3. Randomization
a) Perception that randomization is "unfair"
b) Ethics of the no-treatment control |
|
|
Term
| Alternatives to No-Treatment Control |
|
Definition
1) Alternate order of treatments
2) Use of established treatments as a control
3) Wait-list control
4) Treatment as usual control |
|
|
Term
Quasi-Experiment
1) definition
2) characteristics |
|
Definition
1) Procedures that resemble characteristics of true experiments, for example, that some type of intervention or treatment is used and a comparison is provided, but are lacking in the degree of control that is found in true experiments.
2)
a) Has an IV and DV
b) Includes comparison group
c) Lacks full control
i. Lacks Randomization: this is hallmark of quasi-experiment
ii. Lacks tight control of experimental conditions
3) Use a true experiment if at all possible in order to make clear cause-effect |
|
|
Term
| Why should we use a Quasi-Experiment? |
|
Definition
1) Broader questions with broader populations
2) External validity--does the intervention "work" in the real world
3) Some hypotheses can't be tested with a true experiment
4) Translational Research: take experiment in lab and translate into real world |
|
|
Term
| List the various types of Quasi-Experiments |
|
Definition
1) One Group Pre-Test Post-Test
2) Nonequivalent Control Group Design
3) Simple Interrupted Time Series
4) Time Series with Nonequivalent Control Group |
|
|
Term
| One Group Pre-Test Post-Test |
|
Definition
O1-Pretest
X1-Intervention to Treatment
O2-Posttest
O1 X O2
-No control group at all
-BAD CHOICE for an experiment
--> what Dr. does. --Ex: take temp., gives meds... see back in few days and take temp again. compare pre and post tests to see if change (did temp. go down?)
- most commonly done; basis for all others |
|
|
Term
| Nonequivalent Control Group Design |
|
Definition
Quasi-experimental procedure in which a comparison is made between control and treatment groups that have been established on some basis other than through random assignment of participants to groups.
- Compare intervention to a "like" control group but without randomization
O1 X O2 (has intervention)
O1 O2 (no intervention) want to see difference in the two O2's compared
Ex: New peer social support program for sev. mentally ill.
- Intervention group: clubhouse A -nonintervention group: clubhouse B
-pretest for social interaction skills, then intervene with clubhouse A but not clubhouse B |
|
|
Term
| What is the Nonequivalent control group design vulnerable to? |
|
Definition
1) Selection-history
2) Selection-instrumentation
3) Selection-regression to the mean
4) Selection-maturation
5) regression
6) Expectancy, Contamination, Novelty effects |
|
|
Term
| Interrupted Time-Series Design |
|
Definition
Simple Interrupted Time-Series Design
O1 O2 O3 O4 O5 X O6 O7 O8 O9 O10
Time Series with Nonequivalent Control group
O1 O2 O3 O4 O5 X O6 O7 O8 O9 O10
--------------------------------------
O1 O2 O3 O4 O5 O6 O7 O8 O9 O10
1) Multiple observations before intervention
2) Multiple observations after intervention
3) Comparison of baseline before and after intervention
-Ex: Effect of Honor Contract: when implement honor contract, cases of cheating went from 130-->~ 60. Huge drop indicates honor contract= effective
-Good but has NOT shown cause and effect--> ideal= repeat study multiple times (in this ex: = study at numerous different colleges and compare results)
-must be abrupt change w/ intervention
-this design= used regularly in business
Ex: not selling enough product--> new add campaign--> see if have change in sales DV= units sold (could be yearly, monthly, daily) |
|
|
Term
| Threats to Internal Validity associated with Simple Interrupted Time Series Design |
|
Definition
1) History, especially Cyclical change
-seasonal variations
2) Instrumentation
- New measures are often used with new programs/ interventions
3) Maturation, testing, regression to the mean all well controlled by obtaining a baseline |
|
|
Term
| Time Series with Nonequivalent Control Group |
|
Definition
Quasi-experimental procedure that improves on the validity of a simple interrupted time-series design by including a nonequivalent control group; both treatment and comparison groups are observed for a period of time both before and after the treatment
1) Multiple pre and post tests
2) Intervention and Control group
3) Rules out history and instrumentation effects |
|
|
Term
|
Definition
-Can I generalize my results?
1) As usual, related to representativeness of the sample
a. Do the results generalize to the larger population
b. Are the effects of the intervention the same or different for different groups?
2) Replication is the best test for external validity |
|
|
Term
| What is the best test for external validity? |
|
Definition
|
|
Term
| Is a Quasi-Experiment (or True Experiment) Valid? |
|
Definition
1) Researcher as detective
a. consider all threats to internal and external validity
2) Control and test for threats as much as possible
3) Note existing threats in limitations |
|
|
Term
|
Definition
Research that seeks to determine whether a change proposed by an institution, government agency, or other unit of society is needed and likely to have an effect as planned, and, when implemented, to have the desired effect at a reasonable cost.
1) Are human services effective?
-->Ex: given to charity. Do you know that that $ helped? no.
2) Perhaps the most applied of all research
3) Key Questions:
a. What are the needs?
b. What are the processes involved in program implementation?
c. What are the outputs
d. What are the outcomes
e. How efficient is the delivery of services? |
|
|
Term
|
Definition
- Large scale, real-world tests
- often more significant consequences
- often greater investment of time, money, effort
- often driven by or affected by political concerns
--> ex: Does adding extra police reduce crime? (5% of adolescence create 60% of crime)
- However, the outcome of social experiments cannot be "truly" evaluated.--> usually time lag in implementation, too many variables, |
|
|
Term
| Applied Research (Definition and Types) |
|
Definition
Examination of psychological principles and treatments in real world settings
Types:
1) Case studies
2) Single Subject experiments
3) Quasi-experimental designs
4) Clinical trials
5) Program evaluation |
|
|
Term
|
Definition
1) Intensive description and analysis of a single case
a. does not use experimental techniques
b. Low degree of control
c. Based on clinical impressions
d. Can include empirical data
2) Exploratory method
3) contributes to development of hypotheses and theories
a. Cannot test them directly
b. Can sometimes disprove them |
|
|
Term
|
Definition
-working on railroad, explosives go-off, metal rod shoots all the way through cheek and up and out of top of head. Didn't die but parts of pre-frontal lobe= destroyed. (Pre-frontal lobe associated with ability to control emotions, and not do stupid things)
-After the accident, he became much more likely to be profane, to swear, etc. Lost ability to stop self from saying those things etc.
-This is why people with ADHD say things they shouldn't say, because signals in pre-frontal lobe are not firing quick enough. |
|
|
Term
|
Definition
1) Develop new ideas or hypotheses
2) Theory development (challenge and support)
-"Counterinstance"-- a case that defies a generally accepted principle
- Support
3) Intensive detailed information-- not usually available in study of a group
- Idiographic approach (study the individual)
- Nomothetic approach (seeks to determine the typical or average of a group)
4) Trying out new techniques or clinical innovations
5) Very rare events
- Brain injury
- Feral children
- Lesch-Nyhan
6) Can be more compelling and meaningful |
|
|
Term
| Disadvantages of Case Study |
|
Definition
1) Can't make causal inferences
2) Observer bias
3) Experimenter bias
4) Threats to external validity--is it generalizable?
--Sometimes yes, sometimes no
5) The danger of the testimonial
--notoriously awful |
|
|
Term
| Disadvantages of Case Study |
|
Definition
1) Can't make causal inferences
2) Observer bias
3) Experimenter bias
4) Threats to external validity--is it generalizable?
--Sometimes yes, sometimes no
5) The danger of the testimonial
--notoriously awful |
|
|
Term
| When to use the case study |
|
Definition
1) new, rare, or unstudied phenomena or treatment
2) Initial data
3) Complement the nomothetic approach |
|
|
Term
| When NOT to use case study |
|
Definition
1) To establish causal inferences
2) To try to make statements about groups
3) To try to prove a theory or test a hypothesis
4) To demonstrate the effectiveness of an intervention |
|
|
Term
| Single Subject (or Small-n) Experiments |
|
Definition
Manipulating IV and analyzing.
-only use when can do something and see an immediate change (very hard to do effectively)
-Skinnerian analysis of behavior
-Apply experimental designs to a single subject or small group of single subjects
-Manipulate an IV within the subject
- Examine changes in behavior (the DV)
- Possible to establish causal inferences
1) first establish a stable baseline (meaning nothing really changing; typically 5 units at minimum)
a. want 5+ units b/c want to show stability
b. If don't have stable-baseline, you don't know if the intervention works. To ensure not by chance; if not stable, have to wait unit it is stable
c. IV= intervention; DV= behavior |
|
|
Term
| 2 Types of Single Subject Experiments |
|
Definition
1) Reversal Design a.k.a ABAB Design
2) Multiple Baselin |
|
|
