2) Ethics, Stats Flashcards
What are the three core ethical principles?
- Beneficence & concern for welfare
- Autonomy & respect for persons
- Justice
Beneficence & Concern for welfare
Consider the Risks & Benefits
- maximize benefits and minimize harm
- Benefits of participation
- Consequences of not doing research
- Risks of participation
- Confidentiality
Autonomy & Respect for persons
Informed Consent
- given in advance, all aspects of research that might influence decision to participate
- ensure autonomy (ability to make one’s own decision about participation)
- protect minors, cognitively impaired, and disabled
- decision is free from coercion
All aspects of research given in informed consent
- purpose
- procedures
- risks & benefits
- rights to refuse & terminate participation
- compensation
- confidentiality
- contact info
- ethics
Justice
Selection of Participants
- Treat people fairly & equitably by distributing benefits / burdens of participating in research
- Inclusion & exclusion of groups must be scientifically justifiable
- Can’t have one group bear all the burden of research
- Group that is studied must receive benefits from the study
ex/ drug to cure cancer study, ppl in study should get the treatment first if successful
Debriefing
process where researchers inform participants what the study was about
- Sometimes explain hypotheses
- Needed especially if study involves use of deception or withholding hypothesis
- Research study should be a learning experience for participants too!
Describe the controversy around animal research
Clear benefits in our understanding of learning, brain parts, treatments
No good alternatives to using animals
Emphasize the need for adequate housing and feeding conditions
Question is whether animal research offers enough external validity to justify its use
What is the purpose of descriptive statistics?
- Summarizes mass of data points
- Help with understanding and interpretation
- Visual displays and make appropriate calculations
Descriptive statistics in experiments vs correlational designs
compare averages between groups
vs
describe magnitude and size of relationship between two variables
What are the measures of central tendency?
Mean, median, mode
Mode
most frequently occurring score
- used for ordinal variables
- don’t take numerical value but can count them
Median
- Score that divides group in half
- Useful when an equal number of people score higher and lower
Mean
Arithmetic average
- what we usually use
What are the pros and cons of the mean?
pros
- maximizes use of all of data (unlike median and mode)
- has mathematical properties that allow us to use in statistical analysis
cons
- affected by outliers
What measure of central tendency is most affected by outliers? What can we use instead?
- mean is most affected
- use median if there are outliers
What can we do so that the mean is less affected by outliers?
increase the sample size
Can there be more than one mode?
yes
- we can have none, one, or more than one mode
Variability
Spread in a distribution of scores
i.e. differences from average
Measures of variability
Range
- more intuitive measure of variability but can be deceptive
Standard deviation
- better measure of variability but more difficult to calculate
Are there units in standard deviation?
nope
standardized means removed units of measure
A group with greater spread will have a ____ standard deviation
higher
graph is lower / more spread out
What is a normal range?
within one standard deviation of the mean
(that’s 68% of scores on a normal distribution)
Identify the 3 skills for evaluating psychological claims
- consider the source
- Be on lookout for excessive sharpening & levelling
- Seemingly balanced coverage / pseudosymmetry
- unreliable experts talking to both sides of issue
What is sharpening & levelling?
Sharpening is the tendency to exaggerate the central message or gist of a study
Levelling is tendency to minimize the less central details of a study
This brings out most important facts into sharper focus (good in moderation)
pesudosymmetry
appearance of scientific controversy when none exists
Replicability vs reproducibility
Replicability is obtaining consistent results using same methods but collecting different data
Reproducibility is reanalyzing the same data to see if u get the same results
Which correlation suggests the strongest relationship?
A) +0.01
B) -0.99
C) +0.87
D) -0.35
B
A well-designed experiment has ____ internal validity
high
People who exercise more will score higher on math test
Conditions: run 6 hrs outside vs run 3 hrs in gym
What are the IV, DV, and confound?
IV: Amount of exercise
DV: test performance
Confound: running outside or inside
Which design(s) is sample size less of a concern?
Case Study
Which design(s) uses random assignment?
Experimental only!
Which design(s) may use random sampling?
Experimental
Correlational
maybe quasi-experimental
Which design has the highest internal validity?
Experimental
Which design has the highest external validity?
Naturalistic observation
In order to examine the effects of malnutrition, dental services were withheld for a number of years. This violates the principle of:
A) Respect for persons
B) Concern for welfare
C) Justice
D) Autonomy
E) Two of the above
B
Findings from the study suggesting the importance of vitamins and nutrition never benefited the participants of the studies. This violates the principle of
A) Respect for persons
B) Concern for welfare
C) Justice
D) Autonomy
E) Two of the above
C
Subjects in the vitamin experiments were not permitted a choice concerning participation.
A) Respect for persons
B) Concern for welfare
C) Justice
D) Autonomy
E) Two of the above
E
What is the key to generalizability?
random selection
Research design make use of [ ] thinking because scientific reasoning requires us to question our [ ]
analytical // intuitions
