View
- Term
- Definition
- Both Sides
Study
- All (115)
Shortcut Show
Next
Prev
Flip
PSYC 204: EXAM 3
central issue of quasi experimental designs |
research validity |
How are participants selected for quasi experimental designs? |
from preexisting groups. |
do quasi experimental designs use random assignment? |
no |
levels of control for quasi experimental designs? |
low levels over who what when where how |
confounding variables do what to internal validity? |
decrease. but do not always render invalid |
nonequivalent control group designs are what? |
research designs that have both experimental and control groups but participants are not randomly assigned to this group. |
what is the most common quasi experimental design? |
nonequivalent control group. |
control procedures that further improve the interpretability of nonequivalent control designs? |
matching
moderator variables built into the study |
examples of nonequivalent control groups |
delayed control group design
mixed factorial design |
delayed control group design |
nonequivalent control group design in which the testing of one group is deferred. two groups are tested sequentially with an appreciable time interval between them. |
mixed factorial design |
have one between subjects variable and one within subjects variable. ie study of trait (between) and study of state anxiety (within)
between subjects variable=preexisting |
can mixed factorials be experimental and non experimental? |
yes |
Designs without control groups |
1. interrupted time-series designs
-multiple time series design
2. repeated treatment designs |
interrupted time-series designs |
these designs allow the same group to be compared over time by considering the trend of the data before and after treatment.
collect large number of pretest data, introduce intervention, wait to see what happens. **always contend with historical threats that are confounded with intervention |
multiple time series design. |
variation of interrupted time series design. not a design without a control group. control and an experimental group are implemented to rule out history as a rival hypothesis. |
repreated treatment designs |
this research design allows the same group to be compared by measuring participants' responses before and after treatments.
pre treat post pre treat post
-trying to prove that the treatment is what caused increase in behavior.
|
observational research definition |
researcher observes, records, and quantifies ongoing behavior. |
can observational designs be experimental and non experimental? |
yes |
sampling of observations |
must be randomized or nonsystematic |
what requirements must the observer meet for a good test? |
-reliable (scorer and test-retest)
-valid
-standardized
-objective |
when there is more than 1 observer, observer characteristics (experimenter effects) become what? |
possible extraneous variable and need to be controlled for |
how to control for experimenter effects |
1. train raters to standardize their observations
2. build into design as a moderator.
3. select good observers |
naturalistic/complete observation |
research conducted in such a way that the participants' behavior is disturbed as little as possible by the observation process. ie) observing shoppers at the mall. |
observer-participant |
observations are made such that there is no interaction, but the participants are aware of the observer's prescence. |
participant-observer |
researchers participate in natrally occuring groups and record their behaviors.
ie) cult
|
complete participant |
observations made within the observer's own group. ie) reports of aggie life by student for the battalion. the observer is completely immersed in the activities being observed because apart of the group |
least intrusive level of observation |
naturalistic observation |
most intrusive level of observation |
complete participant. |
potential problems of observing |
intrusiveness
reactivity
issues of privacy. |
offline |
video tape and later watch. |
online |
sitting in the same room with research participants. ie) assessment centers. |
offline is better bc |
you can go back and review and can use multiple observers |
can you make causal inferences from observational designs? |
it depends. if experimental designs yes
if not then no |
do you have to use correlation for correlational design (research design/method)? |
no |
correlational design characteristics |
no manipulation
low control
no causal inferences |
types of correlational designs |
1. predictive
2. concurrent
3. postdictive
|
which type of correlational design is superior to the rest? |
predictive |
which correlational design should you use as the last choice if you can? |
postdictive |
predictive correlational design |
IV data is collected before DV data with an appreciable time interval between the two. |
concurrent correlational design |
IV and DV collected at the same time without any appreciable time interval between the two |
postdictive correlational design |
DV has been occuring in the past before the IV was collected. |
some issues with postdictive designs? |
selective sampling. this can be a treat because people you want to analyze may be unavailable.
there are instances when the DV may influence standings on the IV.
archival data: no control over how the data was collected. |
archival research |
refers to research conducted using data that the researcher had no part in collecting.
archival data are those that exists in public records archives. the researcher simply examines or selects the data for analysis. |
can causal inferences be made about correlational data? |
no. |
limitations of archival data |
most archival data is collected for un scientific reasons by non researchers. may not be complete may be subject to bias (police reports)
archival data is carried out after the fact so ruling out other hypothesis for observed correlations may be difficult.
-reliance on post hoc explanations elevates susceptibility to alternative explanations. |
open ended/constructed responses |
permits the respondants to answer in their own words |
close ended |
limits the respondants to alternatives detwrmined in advance by designer. |
format of item |
-construct response (fill in/write in)
true/false
multiple choice
linkert scale.
|
adjacent question effect |
matters which order is presented. a/b first etc. |
willingness to answer |
some people will not answer items theyre unsure about. others may guess. can control for this by instructing participant answer all questions. |
willingness to answer |
some people will not answer items theyre unsure about. others may guess. can control for this by instructing participant answer all questions. |
position preference |
when in doubt pick c.
control for this by randomization of alternatives.
|
acquience or yea and nay saying |
tendency to agree or disagree with questionaire statements regardless of content. controls are: repeat item and reverse. and bi directional responses |
response sets |
tendancy to respond to a questionaire or test content with a particular goal in mind. |
social desirability |
is the most common response set. |
self deception |
type of social desirability. occurs when an individual unconsciously views him/herself in an inaccurately favorable light. lack of self awareness. |
impression management |
refers to a situation in which an individual consciously presents him/herself falsely to create a favorable impression. |
social desirability over reports _____ and underreports ______.
|
socially desirable and socially undesirable characteristics. |
which big 5 factors are most susceptible to impression management/faking effects? |
conscientiousness |
common method variance and collection of data from single source |
potential to inflate and confound observed relationships especially where there is a theoretically justifiable reason to expect this.
ie) relationship between task performance, job satisfaction, organizational commitment, and using self report measures of all variables from employees. |
major survey techniques |
face to face interviews
telephone interviews
mail
magazine
internet-based surveys and sample recruiting sources.
-study response
E-research global
zoomerant
survery monkey
|
sampling |
key to meaningfulness of any survey is soundness of the sampling procedure used. |
types of sampling procedures |
1. uncontrolled
2. haphazard sampling
3. probabilistic sampling |
uncontrolled sampling |
researcher has no control in the selection of respondents
ie) magazines, radio, tv call ins.
usually small sample (2%)
usually biased to more vocal individuals.
|
haphazard sampling |
sampling procedure where researcher has some control over selection into study but is still basically a hit or miss method for selecting participants.
ie) telling reporter to go interview 2 girls 1 man on street
|
probabilistic sampling |
typically use if you can't get the whole population. sampling procedure in which researcher makes an effort to assure that each person in population has an equal chance of being selected. |
simple random sample |
sampling procedure from an entire population. everyone has an equal chance of being chosen. |
stratified random |
sample is chosen to proportionally represent certain segments in larger population. |
cluster |
sample is selected by using clusters or groupings from the population.
ie) ebery sudent in 10th class rather than every 10th student (simple random) |
margin of error |
if sample is 100/100= 0 margin of error. N=1/m2
27% like snickers
25% mars bars. margin of error is 3% so since betwee nthat is abt the same.
|
can we make causal inferences on surveys? |
no |
cross sectional designs what what type of designs? |
BETWEEN subject designs |
cross sectional designs |
different people are tested at different points in time.
advantage: economical
|
longitudinal designs are what type? |
typically within subject designs but can be between too. |
advantage of longitudinal designs? |
strength in allowing us to assess the change in variables or constructs over time. |
longitudinal designs are typically _____ than cross sectional designs |
stronger bc temporal sequencing of IV and DV is more clearly established. |
ime lag designs |
these designs permit us to investigate changes across or differences between coherts. |
time lags hold _____ constant |
age |
time lags use what type of designs over time? |
cross sectional |
do time lags totally eliminate confounding? |
no |
specific threats to internal validity faced by longitudinal and cross sectional designs? |
1. selective survival
2. selective dropout
3. practice effects or retest effects
4. history cohort or generation effects |
selective survival affects what types of designs? |
longitudinal and cross sectional.
more critical with older adults. difficult to make inferences because population isnt the same if people die. |
selective dropout applies to what designs only? |
longitudinal.
people may die, move away, lose interest in study etc. so people who continue the study may be inherently different |
practice effects or retest effects applies to what designs? |
repeated measures longitudinal designs bc same individual is tested and retested on same psych behavior over a long period of time.
-may become used to test
may score higher on a test if practice |
history, cohort, or gneration effects is a threat to what design? |
cross sectional designs. |
cohort |
some group that has some characteristics in common; usually thought of in terms of different age groups. |
cohort effect |
the variable by which the cohort is grouped confounds the IV
-older people may have problems programming a phone because they didnt grow up with them. |
primary |
data comes from research participants |
secondary |
data comes from the results of other studies. |
what type of research is meta analysis? |
secondary |
if have stronger research design..... |
have stronger effects |
calculating the effect size (d) |
d= Me-Mc/Sw |
a positive d means? |
indicates that the ecperimental group responded better than the control group on the dependent variable. |
negative d means? |
control group responded better than the experimental group. |
0 d value means? |
no diff between control and experimental group. |
cumulating effect sizes across studies? |
mean sample weighted effect sizes (d) can be calculated using formula 2
d= sumdiNi/Nt |
how do meta analysis obtain a more stable effect size? |
controls for sampling error by assigning more weight to studies with larger samples and therefor obtains more stable effect size estimates |
stat significance testing? |
seen as holding social sciences back. focus on magnitude of effect instead. |
meta analysis use a common _____ |
metric to aggregate effect size across studies. |
meta analysis can summarize ____ volumes of literature |
large |
how can mmeta analysis be used to resolve conflicts between 2 or more bodies of literature? |
compare effect sizes across them. |
meta analysis and relationships |
can investigate relationships that werent investigated in first study |
meta analysis can identify ____ trends |
subtle. too subtle to identify with narrative reviews |
meta analysis are more ____ and relatively more _____ |
standarized and objective |
disadvantages of meta analysis |
garbage in and garbage out
apples and oranges comparisons
# of primary studies available
selection of primary studies for inclusion. (convenience sample)
file drawer problem
judgement calls |
ethics |
truth in reporting
-accurate reporting of research
-plagiarism
treatment of research participants (informed consent)
internal v external controls and checks
|
ethical |
a piece of behavioral research is ethical when the benefits and relevance of research balance costs in time and risks of harm to participants, when their interests and well being are respected and when they are informed about nature of research and voluntary nature of participation. |
research ethics |
guidelines to research decision making |
informed consent and voluntary participation |
what is study about? what will it cost? risks?
an agreement with participants that clarifies the nature of the research and the responsibilities of each party.
should be in writing and ensures participation is voluntary. |
elements of a consent form |
topic domain-what is study about
one of N participants
time
risks and costs (can't shift costs to participants)
benefits
confidentiality or anonymity of responses
withdrawal and associated penalties
IRB review if required
researcher contact info -IRB contact info -signed copy by both researcher and part.icipant for both
|
deception |
first and primary concern should be the welfare of participant.
is deception absolutely necessary?
weigh potential costs against potential gains
when deception is used- debriefing and desensitizing needed as |
debriefing |
debriefing participants about any deception that was used in the study. also increases their understanding. |
desensitizing |
eliminating any undesirable influences that the experiment may have had on participants- debriefing participants about their behavior. don't do if it will cause more harm than not doing so. |
ethical dilemma |
investigator's conflict in weighing potential cost to participant agaunst potential gain to be accrued from research project. |
truth in reporting |
plagirism
accurate reporting |