Cj Research Methods Test

Front 1

historically unethical research experiements

Back 1

Stanford Prison Study simulate a prison in basement for two weeks, with 6x9 cells (w\ a cot), a prison yard/corridor, and a 2x2x7 solitary confinement cell. Install cameras and microphones.
Subjects: assign 21 normal, healthy males as guards and prisoners. Pay each $15 per day for participation. Provide basic instructions about roles. Guards are to believe that main role of study is to observe prisoners.
Procedure for prisoners: arrest subjects at home, take to police station, search and book, blindfold, take to prison, delouse, stand naked, get uniform, get sent to cell and ordered to remain quiet.

Front 2

Fraud and plagerism

Back 2

Scientific research must be honest, accurate, and referenced.
Research Fraud: fabricate/misrepresent research
Due to publish or perish
Pressure to obtain/retain grants
Plagiarism: presenting work not one’s own

Front 3

How do we aviod research fruad?

Back 3

Scientific books and journals are peer-reviewed.
Journals are “not for profit.” What implications does this have, versus “for profit” publications?
What is the peer-review process like?
Double-blind (what does this mean?)
Most manuscripts require revisions. Editors and reviewers ensure that there is enough detail for a reader to verify factual claims and replicate the research.
Editorial corruption?
Also, authors are expected to list their affiliations as well as their sources of funding (why?)
This process is not perfect, but the discovery of fraud will fatally tarnish a researcher’s career.

Front 4

the researchers role

Back 4

Role as researcher might conflict with roles like good citizenship, and so forth.
One must be objective and value-free in doing research. But why must we be?
A neutral observer (but why might this be a problem?)
Not law enforcement (why might this be a problem?)
Not to be a CJ practitioner (why might this be a problem?)
Primary role is scientist!

Front 5

Ethical issues (summary)

Back 5

Avoid procedures that may harm subjects.
Participation must be voluntary
Deception
Honor commitments to respondents and respect reciprocity.
Exercise objectivity and professional integrity in doing and reporting research.
Protect confidentiality and privacy.
Most ethical questions involve weighing the possible benefits of research against potential harm to subjects.

Front 6

Do no harm

Back 6

Do nothing that will harm, especially if potential harm has not been explained to subjects and their informed consent given.
Researcher is personally responsible during all phases of research for harm

Front 7

voluntary participation/Honor commitments

Back 7

Cannot use force or threaten participation
But can sometimes trick people (i.e., use deception)
Sounds like a good idea, but how can we get results we can generalize?
I.e., are people who volunteer the same as people who don’t volunteer?
Keep any promises you make!
Reciprocate trust and obligation!

Front 8

protect confidentiality

Back 8

Confidentiality: researcher can connect data to individual subjects, but subject identity is screened in research report.
Anonymity: researcher cannot link individual subjects to their data.

Front 9

HOw do we promote ethical research?

Back 9

Fed. Regulation: Institutional Review Boards
Informed consent: involves relating research procedures and getting consent. Subject must understand purpose of research, risks and side effects, benefits, and procedures.
Problem 1: what if deception is needed?
Problem 2: can subjects appreciate/understand risks?
Special populations: youths, homeless, mentally retarded, and inmates. Why are these different from “regular” people?
No coercion to participate!
If one cannot avoid an ethical problem, consider using another research strategy

Front 10

The experiemental model

Back 10

Experiments minimally involve: (1) taking action, and (2) observing the consequences

Front 11

scientific notation

Back 11

X (treatment; independent variable);
Y (outcome; dependent variable);
Z (rival causal factor);
O (observation; measurement of dependent variable);
E (equivalence; randomization or matching);
1, 2 (time points)

Front 12

How do we show causation

Back 12

First, we must show a relationship between X and Y.
Second, we have to show that X appeared before Y did.
Sometimes, we can infer time order through logic (for example, it makes more sense that age causes crime than the other way around).
Third, we have to rule out rival explanations (spuriousness).

Front 13

rival causal factors

Back 13

Two types: internal (problems within the study that invalidate its results) and external (factors that limit generalizability)
Can we trust our results (I.e., does our research have validity)? There are two types of validity:
Internal: did a variable other than X cause a change in Y?
External: what limits how far I can generalize my results?

Front 14

internal treats to validity- history

Back 14

other specific events that took place during the period of the study that the researcher did not intend, but could have produced the results.

Front 15

internal threats to validity- maturation

Back 15

biological or psychological changes in the subjects not due to the treatment variable. Ask yourself: “Is the change due to natural causes or the experiment?”

Front 16

internal threats to validity- testing effects

Back 16

pretesting can lead to bias and foreknowledge that can lead to different results.

Front 17

internal threats to validity- instrumentation

Back 17

occurs when the researcher changes the measurement strategy. Changing the measures can influence the results.

Front 18

internal threats to validity- statistical regression

Back 18

refers to the tendency of groups selected because of extremely high or low scores to move back toward the average on a second testing.

Front 19

internal threats to validity- selection bias

Back 19

occurs when the researcher chooses nonequivalent groups for comparison.

Front 20

internal treats to validity- experimental mortality

Back 20

refers to the loss of subjects during the study, which might bias the results if the subjects lost are different from those remaining.

Front 21

internal treats to validity- selection-maturation interaction

Back 21

maturation can be more pronounced for some groups relative to comparison groups, which can affect results.

Front 22

external treats to validity- testing effects

Back 22

occurs when pretests make subjects unusually sensitive to the study (which means that the findings would not apply to those who did not participate in the study).

Front 23

external treats to validity- selection bias

Back 23

nonrepresentative samples cannot be generalized to larger populations

Front 24

external treats to validity- reactivity

Back 24

awareness of being studied

Front 25

external threats to validity- multiple treatment interferences.

Back 25

when more than one treatment is used on the same subjects (I.e., what caused the change?)

Front 26

list internal threats to validity

Back 26

History, maturation, testing, instrumentation, statistical regression, selection bias, experimental mortality, and selection-maturation interaction

Front 27

list external threats to validity

Back 27

These affect the representativeness of our results
Testing effects,Selection bias, Multiple-Treatment Interferences, reactivity

Front 28

list Related Rival Causal Factors

Back 28

The Hawthone effect, Post Hoc Error, Halo effect, placebo effect

Front 29

Related rival causal factors- hawthorne effect

Back 29

refers to atypical behavior by subjects because of the awareness of being studied.

Front 30

rival causal factors- post hoc error

Back 30

where people attribute causation to X (an earlier event) after becoming aware that Y has changed

Front 31

rival causal factors- placebo effect

Back 31

where subjects react to known stimulus in a predicted manner.

Front 32

The best way to rule out threats to internal validity

Back 32

have an individual receive the treatment (and see what happens)—then go backward in time—have the same individual not receive the treatment (and see what happens).

But it can’t be done! (or can it?)

Front 33

experimental designs

Back 33

characterized by random assignment to experimental and control groups (include classical, posttest only control group, and Solomon 4-group studies

Front 34

quasi experiments

Back 34

use precision matching to assign subjects (includes time-series and counter-balanced designs).

Front 35

preexperimental designs

Back 35

lack equivalence across groups and include one-and two group ex post facto and one-group before/after designs.

Front 36

the classic experimental design

Back 36

equivalence, pretests and posttests, experimental and control groups.

Front 37

equivalance
pre/post test
experiment/control group

Back 37

necessary to ensure that experimental and control groups are alike; researcher employs random assignment (where all subjects have an equal probability of being selected into a particular group).
Matching is an alternative way of achieving equivalence, with both groups being matched on all characteristics the researcher believes to be important.Pretest: observation prior to treatment; posttest: observation after treatment
Experimental group: group of subjects that receive treatment; control group: group of subjects that do not receive treatment