58 Cards in this Set
Front | Back |
---|---|
Under what conditions (or for what reasons) might a researcher use a quasi-experimental design (QED) instead of a traditional lab experiment?
|
Used when we CAN'T:
1: answer research question in the lab
2: manipulate levels of IV alone
3: eliminate potential confounds
|
How do quasi-experimental designs affect external validity?
|
Increase external validity by studying behavior in a more natural setting
|
How do quasi-experimental designs affect internal validity?(3)
|
decrease internal validity by:
1: having less control over participant selection/assignment
2: less control of potential confounds
3: less confidence in IV/DV relationship
|
What are the 5 main types of quasi-experimental designs?
|
1- time series
2- interrupted time series
3- multiple time series
4- non-equivalent before-after design
5- retrospective and ex post facto design
|
Describe a time series design
|
1- examines effect of IV on one group of participants (within-subjects)
2- utilizes a pre-test post-test design
|
What is a potential problem with time series designs and what is a potential solution?
|
problem- we don't know the normal fluctuations of performance when there is no treatment imposed
solution- use an interrupted time series design
|
Describe an interrupted time series design
|
1- same as a time series design except it uses SEVERAL pre-test post-test measures
2- allows for better estimation of normal fluctuations in behavior over time
|
What are two problems with interrupted time series designs? What is a solution to these problems?
|
1- no true control group
2- Unclear whether variables other than IV cause differences in pre‐&post‐test measures
solution - multiple time series design
|
describe a multiple time series design (2)
|
1- similar to interrupted time series design except it includes a control group
2- use of a control group changes the design from within-subjects to between subjects
|
What is a problem with a multiple time series design?
|
an unknown variable may still have an effect on the experimental group, control group, or both.
|
describe a non-equivalent before-after design (3)
|
1- used to make comparisons between 2 groups we think differ before the study begins
2- if 2 groups differ before treatment, there's a risk for confusing initial differences with the effect of the IV
3- this is controlled for by measuring differences in pretest and posttest scores for e…
|
When using a non-equivalent before-after design, what threat to internal validity is ruled out? What threat to internal validity may still be present?
|
history effects is ruled out, but maturation effects may still remain
|
describe a retrospective/ex post facto design (2)
|
1- determine whether there are meaningful relationships between events that have occurred in the past
2- often used to examine education techniques laws, disease, war, economic changes and psychopathology
|
What are three limitations of a retrospective/ex post facto design?
|
1- non-random selection of participants
2- no control group for comparison
3- no true manipulation of IV by the researchers
|
What are the two primary functions of observational studies?
|
1- acquire knowledge where little is known (observing and describing a phenomenon)
2- gain insights regarding possible patterns or lawful relationships (no manipulation/cause effect relationships)
|
What are the potential threats to validity associated with data collection in an observational study? (4)
|
1- concealing who is being studied to avoid reactive behavior/hawthorne effect
2- accuracy of observations (research bias)
3- boredom of researcher
4- data recording/developing clear operational definitions
|
What are two advantages of observational studies?
|
1- describes real-world behavior
2- see how behavior unfolds in real time
|
what are some disadvantages of observational studies? (4)
|
1- simply describes, can't explain
2- vulnerable to subjectivity of researchers
3- limited representiveness of sample, may be inappropriate to generalize
4- ethical concerns, no informed consent or right to privacy
|
How are correlations evaluated?
|
1- graphically, using a scatter plot
2- statistically, using a correlation coefficient (r)
|
describe the correlation coefficient (r)
|
ranges from -1 (perfect neg. correlation) to +1 (perfect positive correlation)
the closer r is to zero, the weaker the correlation
|
What are 4 advantages of survey research?
|
1- studies behaviors which are difficult to observe directly
2- accesses self-report data
3- relatively easy to administer
4- not very expensive
|
What are two disadvantages of survey research?
|
1- designing a good survey is challenging
2- self-report data can be unreliable
|
What are the 6 steps of survey research?
|
1- state a broad objective (why is the survey necessary)
2- narrow it down to a specific objective
3- determine the population and sample that will complete the survey
4- select a survey protocol (email, phone, face 2 face)
5- choose types of questions and responses
6- select analysi…
|
What are 2 advantages of open ended questions in surveys?
|
1- they don't impose any particular point of view
2- may provide info the researcher hasn't considered
|
What's a disadvantage of open ended questions in surveys?
|
difficult to analyze/translate answers into categories
|
What's an advantage of fixed alternative questions in surveys?
|
easyto analyze and translate answers into categories
|
What are two disadvantages of fixed alternative questions in surveys?
|
1- limits the number of responses
2- doesn't tell us the reasons behind responses
|
Whats a disadvantage of combining open ended and closed ended questions in surveys?
|
open ended questions may be difficult to classify. however, the close ended piece serves as a guide to classification
|
describe funneling
|
beginning with an open ended question and following up with more specific response alternatives
|
Why might funneling be used in surveys?
|
funneling gets the participant to think about the answer and how it fits into specific alternative responses before answering
|
Know the basic guidelines recommended when developing fixed alternative questions for a survey
|
...
|
What is a likert like scale and what is it used for in surveys?
|
a 5 point scale ranging from strongly agree to strongly disagree. used in surveys to discover the participants attitudes/beliefs
|
What is a semantic differential?
|
a seven point scale using bipolar adjectives
ex:
good + + + + + + + bad
|
How do single subject design arrangements differ from other design arrangements? (3)
|
1- random sampling/assignment aren't necessary
2- focus on individual behavior, not average behavior
3- interested in functional change rather than statistical change (statistical significance is not so important)
|
What is the goal of case study designs?
|
to provide a detailed DESCRIPTION of an unusual condition/trauma/disorder
|
What is the goal of experimental single subject designs?
|
to determine the functional relationship between environmental events (IV) and behavior (DV)
|
what are the two subtypes of experimental single subject designs?
|
1- intra-subject replication (within)
2- inter-subject replication (between)
|
describe a one-shot case study (2)
|
1- a more experimental version of the case study
2- involves direct observation of a participant after some imposed change/treatment (tooth in eye surgery)
|
What are the advantages of reversal and multiple element designs?
|
1- focus on changing behavior under specific circumstances
2- frequently reversing treatments allows for replication of effects (increase internal validity)
3- multiple element designs allow for several manipulations in the same study
|
What are the disadvantages of reversal and multiple element designs? (2)
|
1- only works when the effects of treatment can be reversed (otherwise, carry-over confounds can affect baseline recovery rates)
2- order of treatments may determine the effectiveness the treatment(s) (order effects), but counterbalancing can help avoid this
|
describe a multiple-baseline design (3)
|
1- used when treatment is believed to permanently change behavior or be long lasting
2- monitors several participants simultaneously
3- applies treatment to one participant after finding a baseline, then apply treatment to additional participants and see if behavior changes are similar…
|
Whats a limitation of multiple-baseline designs?
|
each participant must be independent of each other in order to avoid cross contamination of the IV across groups
|
What are the 5 major ethical issues involved in conducting human research?
|
1- nature of risk
2- informed consent
3- use of deception
4- privacy
5- risk/benefit ratio
|
How can risk be reduced/managed when conducting human research? (3)
|
1- anonymity: no way to link data with a participant's identity
2- confidentiality: keeping identity data separate from study data
3- debriefing
|
what must researchers do in order to fulfill the requirements of informed consent?
|
1- inform participants of any aspect of the experiment that may effect their willingness to participate
2- answer participant's questions regarding the study
3- let participants know their participation is voluntary and they can quit at any time w/out a penalty
|
What 3 groups of individuals can not legally give consent for an experiment?
|
1- mentally ill
2- developmentally delayed
3- people under 18
|
what is the most controversial ethics issue in regards to conducting human research?
|
deception
|
describe deception in terms of conducting human research (2)
|
1- contradicts the practice of informed consent
2- can occur by omission or commission
|
What term(s) did Milgram(1977) use in place of deception?
|
masking/technical illusions
|
what are the 3 dimensions of privacy?
|
1- sensitivity of information
2- setting (public vs private behaviors)
3- degree and manner of dissemination of results
|
what is the general rule regarding the risk/benefit ratio of conducting human research?
|
if the benefits to individuals and society are clear and immediate, greater risk is generally tolerated
|
what general principles guide the practice of science? (6)
|
1- seeking knowledge
2- carrying out competent research
3- accurately + completely reporting results
4- acknowledging other's contributions
5- consider potential impacts of work
6- speak out publicly on societal concerns
|
Ethical issues arise in both the process and products of research. What does this mean?
|
process- what actually happens to the participants before and after the study
products - how the results of the study will be used
|
What resulted from the National Research Act of 1974?
|
1- the department of agriculture and the public health service formulated guidelines for the care and use of lab animals
2- required all institutions seeking federal funding for research to create a committee to review research proposals (Institutional Review Boards)
|
What are the requirements of Institutional Review Boards?
|
1- must have at least 5 members of varying backgrounds
2- at least one member must have no affiliation with the institution
3- must have both males and females
|
Must all research be reviewed by an Institutional Review Board?
|
No, projects posing minimal risk are exempt from being reviewed
|
What is undue inducement?
|
involves the debate about whether or not individuals such as prisoners and psych 105 students can truly give voluntary consent
|
What regulations are in place to protect animals used in laboratory research?
|
1- use of animals should be in accordance with the Animal Welfare Act
2- procedures should consider relevance to human or animal health or seeking knowledge
3- minimum # of animals should be used to obtain results
4- pain/discomfort should be minimized
5- painlessly kill an animal tha…
|