PSYC 204: EXAM 2
52 Cards in this Set
| Front | Back |
|---|---|
|
control
|
any means used to rule out possible threats to a piece of research. applies to all research designs and not just experimental designs
|
|
contol or experimental control
|
ability to restrain/guide sources of variability. (extraneous threats and threats)
with experimental control, we are concerned with out ability to rule out possible threats to out study and rule out other explanations of research results.
|
|
extraneous variable
|
variable other than independent variable that is not the focus of the study but may confound the results if not controlled.
want to minimize extraneous variable
|
|
confounding variable
|
an extraneous variable whose levels covary with that of the independent variable
|
|
common strategies to achieve/enhance control
|
random assignment to groups
manipulation checks
subject as own control (within subjects designs)
instrumentation of response
matching
building nuisance variables into the experiment
statistical conclusion
|
|
random assignment to groups
|
rules out plausible alternative interpretations due to chance
controls for BOTH known and unknown effects
|
|
manipulation checks
|
access the efficacy of manipulation
associated with the IV
did the manipulation work as intended?
Why is it important?
techniques of assessment and when (at end of study ask research participant)
|
|
what does one do if manipulation did not work?
|
-make manipulation stronger
-if ^ not possible, drop participants that failed.
|
|
subject as own control used for what type of design?
|
Within subjects design and single subject design
|
|
subject as own control
|
one of the most powerful control techniques is to have each participant experience every condition of the study
since each indicidual is unique and varies in several ways that could effect research outsome, participant serves as own control and rules out these variations
|
|
potential problems of subject as own control?
|
1. practice effects
2. irreversibility of treatment effects
3. dependability of treatments
|
|
order effects
|
changes in participant's performance resulting from the position in which a condition appears in the study.
ie) movie preference, warm up or practice effects in learning studies.
|
|
sequence effects/carry over effects
|
changes in participants performance resulting from interactions among conditions themselves.
ie) taste tests, eating toothpaste and then drinking orange juice
|
|
counterbalancing
|
a method of control in which the conditions are presented in one order the first time and then another.
|
|
reverse counterbalancing
|
conditions are presented in order the first time and then in reverse order
|
|
block randomization
|
can be considered to be a specific example of incomplete counterbalancing. control procedure in which the order of conditions is randomized, which each condition being presented once before any condition is repeated (BCAD ADBC) this controls for both oder and sequencing effects.
|
|
instrumentation of response
|
using a measurement or instrumentation process that is objective, standarized, sensitive, and also generates reliable and valid scores.
-want to take subjective states- dreams, lying etc and use objective instrumentation to generate measurement/scores.
-trying to maximize amt of control…
|
|
matching
|
procedure whereby participants are matched on some variables or characteristics of interest.
|
|
when is matching typically used?
|
when not possible to use random assignment. basically a procedure that attemps to obtain equivalent groups in the absence of random assignment.
sometimes matching is implemented before participants randomly assigned to groups.
|
|
matching controls for what type of effects?
|
ONLY KNOWN EFFECTS
|
|
What do you suspect for mathcing?
|
important variable or characteristic on which participants differ that can be measured and participants matched on.
also suspect relationship between matched variable and D.V.
|
|
moderator variable
|
a nuisance variable (another IV) that moderates or influences the relationship between IV and DV
|
|
in order to control for extraneous variables you can:
|
build them into the study as a moderator variable by measuring them and analyzing their effects. this is basicallt having another IV.
|
|
a moderator variable is a third variable (in the context of one IV and one DV)
|
that differentially affects the relationship between IV and DV.
|
|
Moderator variable controls for
|
only KNOWN EFFECTS
|
|
moderators?
|
...
|
|
mediators
|
Iv-->mediator->DV
relationship between the IV and DV frows through
|
|
statistical control
|
first determine if the design permits analysis by accepted statistical methods.
increase stat power of a design (increase sample size)
use the appropriate stat technique to enhance control (ancova, correlations, regression)
increase the number of trials or items-enhances reliability of…
|
|
greater degrees of control result in
|
higher levels of internal validity
|
|
Experimenter Effects
|
when a researcher unintentionally affects the results
ie) if research focuses on judgements of warmth, researchers mood could influence performance.
if study concerns smoking, then whether the researcher appears to be a smoker may be important variable.
|
|
experimenter expectancy
|
type of experimenter effect.
issue is if researcher behaves in ways that bias the research results in the expected direction.
|
|
mechanisms that produce experimenter expectancy
|
nonverbal communications influence.
expectancy effects also observed in animals.
|
|
more than one researcher
|
reduces experimenter bias.
|
|
standardization of researcher's behavior in participant-researcher interactions
|
this can be done by standardizing the instructions and procedures and in the maitenance of a constant environmental setting.
the use of "end of session" checklists that are completed by both the researcher and participant to document perception of session
|
|
double blind procedures
|
reduces experimenter bias. keep researcher and participant in the dark about expected outcomes or hypothesis being tested.
|
|
single blind procedure
|
those in which only participant is not fully informed of either the nature of study or the conditions under which she/he is participating. may take forms of deception
|
|
deception
|
conditions under which it can be used
debriefing
|
|
single blinds reduce
|
participant expectations but NOT experimenter bias or expectancies.
|
|
automation
|
automate the procedures treatments and all other aspects of experiments as much as possible.
advantage: uniformity
disadvantage: inflexibility
does this decrease experimenter bias?
|
|
experiment is a study or design in which _____.
|
we manipulate variables.
random assignment is also implied.
high control over who, what, when, where, and why too.
|
|
ideally all research designs should use:
|
random sampling.
|
|
experimental research designs call for ____ _____ to groups.
|
random assignment.
*this is what distinguishes research design from other design
|
|
Random sampling and random assignment give the strongest case for for _____ _______ and _________.
|
causal inferences and generalizability.
|
|
random sampling
|
the process of choosing a "representative" sample from an entire population such that every member of the population has an equal and independent chance of being selected into the sample.
**probabilistic sampling
|
|
Random assignment:
|
(randomization)
control technique that equates groups of participants by ensuring every member of the sample has an equal chance of being assigned to any group.
|
|
Random assignment controls for:
|
both known and unknown effects and threats.
|
|
principle of concern is whether differences between groups AFTER the introduction of the IV are...
|
due solely to chance fluctuations or to the effect of IV and chance fluctuations.
|
|
experimental research designs
|
1. control experiment with control group and experimental group
2. control experiment with no control group
3. control experiment with control condition within subjects???
|
|
Solomon four group design
|
generally accepted as the best design but requires a large amt of participants.
|
|
why should you pretest?
|
equivalence of groups
provides a baseline
effects of testing or practice effects
|
|
what does one do if there are baseline or pretest differences?
|
try and account for the differences:
-difference scores: posttest-pretest (generally a poor method)
-ANCOVA
-Partial/semi correlations
-regression (hierarchical)
|
|
research designs to avoid
|
1. one group posttest only design
2. one group pretest-posttest design
3. posttest only design with nonequivalent control groups. (not randomly selected from the same population as the experimental groups)
|
Study Guide: Final Exam