1- Research Methods Flashcards
Authority
- the weakest form of knowledge because it is based on trust, not on individually collecting information
- it can be useful
- but is often exploited
- humans are designed to work together, which leads to small cues which can influence their validity/authority
the bullshit issue with authority and how it differs from lies
a LIAR cares about truth…
but a BS-er doesn’t care about truth, they just want to convince
many authority figures don’t care about truth
intuition
effortless recognition of pattern
- commonly a starting point in science
- draws form life (anecdotal) experience
- sometimes correct
- often wrong
problems with intuition?
illusory correlation
- we are always looking for patterns!
correlation =/= causation
- third variable problem!!
susceptible to bias
overconfidence
- the dunning-kruger effect
fuck probability we struggle with probability
the dunning-kruger effect
- confidence is highest when you know NOTHING about the subject!!
- so there is no link between confidence and correctness!
gambler’s fallacy
believe that there relationship with probability when there IS none
types of bias (problems with intuition)
- confirmation bias
- hindsight bias (creeping determinism)
- post-hoc explanations (logical fallacy of assuming that one thing caused another merely because the first thing preceded the other)
Observation
- critical to good science: empiricism
- works best with objective measures
- still not enough to acquire the best information about the world…
- example: Naturalism – collecting doesn’t give you information; observation is not enough
problems with observation
- bias/limited explanatory power
scientific skepticism
- question authority and intuition
- question your senses
- question knowledge, beliefs
- systematic doubt and continual testing
but be careful of extreme skepticism/extreme postmodernity
the scientific method
- observation
- idea
- consult past research
- hypothesis
- design study
- ethical approval
- collect data
-analyze data - modify and repeat (if hypothesis wrong)
- consider implications of results, build theories
aspects of good science
- materialism
- everything can be measured objectively because everything is physical in nature
- why? because how would materialist and non-materialist stuff even interact
- universalism
- systematic objective observations in a format that we all agree on
- communality
- methods and results freely accessible and available to everyone
- allows for you and other people to reproduce
- Disinterestedness
- don’t care about the outcome of the experiment, if it fails it fails
- avoid confirmation bias
- Organized skepticism
- weighing it on the science, not authority
- peer review!!
UM OCD
theory
explanations that organize and explain a variety of facts/descriptions/observations
- falsifiable
- generate new knowledge
- parsimonious
- simple and elegant
pseudoscience
- often relies on authority
- emphasis on scientific-sounding jargon
- not falsifiable
- poor/no methodology
- poor/no anecdotal evidence
- not peer reviewed
- ignores/conflicts with known/existing evidence
- vague claims
- often reinforces status quo or a worldview
- does not facilitate further resarch
goals of psychological research
- describe behaviour
- predict behaviour
- determine causes of behaviour
- influence/control behaviour
types of research
- basic research
- description
- prediction
- causes and mechanisms
- applied research
- attempts to find practical solutions to existing problems
conceptual variable
any construct/idea/concept/variable that we can conceptualize but not completely measure
operational variable
proxy/indirect measure of conceptual variable
four general categories of variables
- independent variable
- dependent variable
- situational variables
- participant variables
situational variable
all aspects of experimental situation
- room
- time of day
- temperature of room
- environment!!
try to keep the same for all individuals
participant variables
what the participant brings in with them
- collect enough participants to reduce the noise
curvilinear
describing an association between variables that does not consistently follow an increasing or decreasing pattern but rather changes direction after a certain point
causality
- covariation of cause and effect
- temporal precedence (a has to happen before b!)
- no more plausible alternative explanation
confounding variable
- a variable that varies along with the independent variable
placebo
dose of the drug without the acutal drug
- explanation will shape experience
single-blind
participant doesn’t know whether they’re in the control group or the experimental group
double blind
neither the participant NOR the researcher knows which participants are in the control group or the experimental group
open label
everyone knows which condition they’re in
for ethical reasons
animal models for research
- understand basic mechanisms, genetics
- because there are no viable alternatives
- because they have a high predictive value in regard to humans
- have a number of protections in place
human research
ethical considerations must be made!!
- federally mandated committees review human research proposals
- informed consent
why r^2 instead of r
it gives SHARED variance in the sample
- how much variance in one is related to variance in another variable
standard regression equation
y = mx + b
y = dependent variable
m = coefficient, rate, and slope of line
x = independent variable
b = where line crosses the y-axis
if you have this equation and you have x, you can predict y
standard deviation
tells you how wide the normal curve is
- how much variance there is in your data
exceptions are 2 SD away from the mean!
standard error
how far the sample mean is from the population mena
statistical significance
p < 0.05
gives likelihood of correctness, not effect size
effect size
cohen’s d
tells us magnitude of effect of I on DV
STRENGTH of the relationship
P-hacking
if you do enough comparisons, you will find SOME significance
compare in many different conditions
- If you do 20 of them, there will likely be a false positive or a false negative
- Can deliberately manipulate
- Taking advantage of multiple comparison to make a fake relationship look like a real one
construct validity
do your operational definitions accurately describe what you want it to define?
did you do a good job at operationalizing it?
indicators of construct validity
- face validity
- content validity
- concurrent and convergent validity
- discriminant validity
- predictive validity
face validity
intuition, basically
the degree to which a procedure, especially a psychological test or assessment, appears effective in terms of its stated aims.
content validity
did we capture all the different aspects and content of the conceptual variable
Content validity evaluates how well an instrument (like a test) covers all relevant parts of the construct it aims to measure.
concurrent and convergent validity
does score on test line up with rest of the data with other existing measures?
discriminant validity
Discriminant validity specifically measures whether constructs that theoretically should not be related to each other are, in fact, unrelated.
make sure that there is no 3rd variable issues!!
predictive validity
do variables predict the subject’s future behaviours well?
types of validity
internal validity
external validity
internal validity
you designed a good experiment!
- no confounding variable
- can we trust the relationship between the independent variable and dependent variable
a measure of how well a study is conducted (its structure) and how accurately its results reflect the studied group.
external validity
- good representation of what’s happening in the outside world?
- generalize well?
- need to consider if our participants are WEIRD?
WEIRD participants
Western
Educationalized
Industrialized
Rich
Democratic
may not generalize to other cultures! need cross-cultural psychology
accuracy (validity)
measures what you mean to measure
reliability
does it do the same thing every time?
- Margaret flow Washburn
○ 1871-1939
○ First woman to earn a PhD in psychology
Research focused on animal behaviour and cognition
- Mary Whiton Calvin’s
○ 1863-1930
- first female prez of the apa
○ Preeminent first-generation american psychologist
○ opposed the behaviourist movement
○ conducted significant research into memory
Established one of the earliest experimental psychology labs in the US
- Francis Sumner
○ 1895-1954
○ First African american to receive a PhD in psych in 1920
○ Dissertation focused on issues related to psychoanalysis!
○ Interested in
§ Racial bias
§ Educational justice
○ One of the founders of Howard Unviersity’s department of psychology
“Father of Black Psychology”
- Inès Beverly Prosser
○ 1895-1934
○ First African American woman to receive a PhD in psychology
○ Highlighted issues related to education in segregated vs integrated schools
○ Her work was very influential in the Brown v Board of Education
§ Ruled that segregation of public schools was unconstitutional
Horatio pinero
○ Founded some of the first experimental psychology labs in South America
2 in buenos aires, argentina
- Gunamudian David Boaz and Narendra nath Sen Gupta
○ India
1st independent departments of psych at the U of Madras and the U of Calcutta, respectively
- Mary Whiton Calkins
○ First female prez of the APA
By 1946, nearly 1/4 of american psychologists were female!!
deductive reasoning
ideas are testedi n the real world
Inductive reasoning
real world observations lead to new ideas
○ May or may not be correct
Used to form theories -> hypotheses… tested with deductive reasoning
theory
well developed set of ideas that propose an explanation for observed phenomena
- Hypothesis
testable prediction about how the world will behave if our idea is correct
Clinical or Case Studies
- Focus on one person or just a few individuals
- Pros:
○ Big insight, rich information - Cons
○ Hard to generalize
Ability to apply the findings of a particular research project to larger segments of society
- Pros:
Naturalistic Observation
- Observing behaviour in it’s natural setting
- Observer has to be unobtrusive and inconspicuous!!
○ If people know that they’re being watched, they are less likely to behave naturally
Not limited to research about humans
- Observer has to be unobtrusive and inconspicuous!!
pros and cons about Naturalistic observation
- Pros
○ High validity/accuracy
○ Higher degree of ecological validity —> realism
○ Better able to generalize the findings- Cons
○ Hard to set up
○ Hard to control
○ Requires a lot of time, money, and luck
○ Observer bias
§ Observers might unconsciously skew their observations to fit their research goals or expectations
- Cons
how to protect against observer bias
□ Clear criteria established for the types of behaviours recorded and how they should be classified
□ Inter-rater reliability—measure of reliability that assesses the consistency of observations by different observers
Surveys
- Lists of questions
- Can be completed in short time, easy to administer
Makes it easy to collect data from a large number of people - Pros
○ Info from a large sample of people —> better generalizability - Cons
○ Less depth of information on each person than a case study
Not always accurate responses
- Can be completed in short time, easy to administer
Archival Research
- Relies on looking at past records/data sets for interesting patterns or relationships
Pros
○ Researcher never directly interacts with research participants -> Less investment of time and money
Cons
○ No control over what info was originally collected
§ Research questions have to be tailored to fit this
○ No guarantee of consistency between the records
§ Difficult to compare and contrast different data sets
Longitudinal
- Data-gathering is administered respeatedly over an extended period of time
- Pros
○ Less concern for differences among cohorts
○ Incredible time investment by researcher and research participants
○ Financial investment
○ Research participants must also be willing to continue their participation for an extended period of time
§ High attrition rates
Reduction in number of research participants
- Pros
Cross-sectional research
- Researcher compares multiple segments of the population at the same time
- Pros
○ Shorter term investment compared to longitudinal - Cons
Limited by differences that exist between the different generations
- Pros
illusory correlation
when people believe that relationships exist between two thigns when no such relationship exists
operational definition
precise description of our variables
Important in allowing others to understand exactly how and what a researcher measure sin a particular experiement
Quasi-experiemental
when you cannot directly control the independent variable
○ Cannot make cause-and-effect claims
Ethical constraints
Statistical analysis
determines how likely any difference found is due to chance
Internal consistency
degree to which different items on a survey that measure the same thing correlate with one another
Test-retest reliability
degree to which the outcomes of a particular meausre remain consistent over multiple administrations
ecological validity
degree to which research results generalize to real-world applications
organizations for ethics
institutional review board (IRB)
Institutional Animal Care and Use Committee (IACUC)
