Intro+ Methods Flashcards
define psychology
study of mind [bridge between brain and world]
two kinds of psychology
experimental, clinical
experimental psychology (+ subtypes of experimental)
figuring out how it works/learning about the norm Cognitive Developmental Personality Social psychology Abnormal psychology
Cognitive psych
direct study of relationship between brain and world
Look for the recipes that translate between brain, mind and world
developmental psych
what changes between infancy and adulthood
How do our genes and experiences interact to produce adults
Personality
Interested in the differences between people–most psychologists are interested in how people are the same but these interested by differences
Social psychology
wider social society and how it impacts what we think and do
Abnormal psychology
interested in deviations and using it to answer questions about the mind rather than treating it
Clinical psychology
deviations from the norm/figuring out how to treat deviations
5 foundations of psychological science
Evolution Materialism Idealism Modularity Empiricism
1st foundation: evolution
genes make brains
Result of natural selection; 4.6mil years (small) between humans and chimps; brains similar to primates in a lot of ways
2nd foundation: materialism
brains make minds
Every feeling, experience is from brain
3rd foundation: idealism (+ why does it matter?)
minds make reality
Ideas create ways in which we see world
We view world as impinging on us–we experience something thats already there but it’s created by our mind–we see black bc our brain sees black
This matters bc sometimes our brains and the world don’t match up–visual illusions etc
4th foundation: modularity
the mind is collection of parts
module
diff parts of mind
5th foundation: empiricism
believe only what you can count–the scientific method (tools to figure out world)
Not anecdotes, intuitions–not good basis for building theories about the world
Some intuitions are right, some are not
Methods of “knowing things”
dogmatikos=belief
empirikos=experience
Empiricists:
belief that accurate knowledge of the world requires observation
history of knowledge methods
Greeks preferred empiricists, but fall of Rome killed their legacy
For much of European history dogmatism ruled
Until the renaissance
Method
set of rules and techniques for observation that allows us to avoid illusions/mistakes/etc that observation produces
Problems for psychological science
Sometimes just looking at the world doesn’t give us enough info (e.g. geocentric theory)
So this is hard with psychology bc our perspective is not necessarily reliable
Complexity
Variability
Reactivity
variability problem
we’re all super different
reactivity problem
–humans react to things around them so ppl might react in non standard ways to attempts to understand them/we change our behavior when we’re being watched
Abstract concept
unmeasurable concept u try to measure (ex: aggression)
how to measure abstract concept
Come up with operational definition for the concept
operational definition
observable conditions that define the concept that MAKE SENSE TO USE (construct validity) (Scientific conclusions depend on how property was OD and measured)
Construct validity–
tendency for clear conceptual relation to exist btwn abstract property and operational definition (facial expression for aggression makes sense; clothing doesnt)
Convergent validity–
Operational Definition’s ability to predict other things that it should be able to
Tendency for operational definitions to be related to other ODs
So we check if two ODs give same results (again, Decisions researcher makes has effect on results of the study)
Reliability:
tendency for measure to produce same result when it is used to measure same thing
Discriminant validity:
Also want to make sure converse is true: different results for different things
Good definition has what qualities
has construct and convergent validity
good measure has what qualities
Good measure has reliability and discriminant validity
theory
hypothetical account of how and why phenomenon occurs (eg e=mc^2)
No direct way to test theory
Use it for hypothesis
Hypothesis
testable prediction made by a theory
Population–
complete collection of people whose properties we want to k
Sample–
people whose properties we act measure
weird problem-
most samples are western educated industrial rich democratic
prob not accurate picture of populations that include outside of these groups
Bell curve/normal distribution–
individuals vary considerably around the mean
normal distribution–tend to get results near the mean but if u draw from ND its unlikely to get someone super diff from avg or u were wrong about distribution
distributions are capable of overlapping a lot (and if so people in groups will prob not be that different)
what are statistical conclusions based on
a) mean differences b)estimates of variability (analysis of variance)
bias
Bias–any factor distorting measurement
observer or subject
Subject bias–
research subject might be motivated by something else, like helping or hurting experimenter
We want to know how someone actually behaves when no one is observing
Avoid subject bias by:
Ensuring anonymity
Measure involuntary or nonobv behavior (a tell when someones lying)
Keep subject blind to hypothesis
observer bias
(unconscious) push on part of observer to have results come out a certain way–> has impact on results
avoid observer bias by:
Double blind technique–researcher delegates blind observer and code it so researchers will know but not observer
complexity problem
human brain most complex thing in world
strange loops
We can sit here and think about mind but im thinking about the mind thinking about the mind thinking about the mind…i am a strange loop we are studying what is doing the studying…
Why is psychology so popular now?
Advances in technology
Elements of observation
Measurement
Samples
Bias
When is generality a good assumption?
If something is being presented as universal truth of human nature, u need to check sample and see if it’s accurate representation of population
Sometimes generality may be a good assumption based on circumstances
Generality can be investigated (compare rural chinese to american undergrad for example)
This is all around concept of variability
(ex: different education might mean bad assumption, same genes might mean good assumption)
two parts of a claim
observation and explanation
elements of explanation
correlation and causation
Correlation
two variables are correlated when variation in the value of one is synchronized with variation in the variable of the other
Correlation enables prediction
correlation is evidence in the direction of causation but it’s not concrete proof for causation
variable
properties that can change
line of best fit
Use r=correlation–line of best fit
Correlation can be anywhere from -1 to 1
perfect correlation line of best fit
r=1 is perfect correlation
(looks like y=x)
(perfect=variables perfectly predict each other)
r=-1 negative perfect correlation
(looks like y=-x)
perfect correlations dont rlly exist irl–we’ll see positive slope but it’s not perfect
no correlation line of best fit
Correlation of flat line is =0, no correlation
3 ways we might think correlation equals causation when it doesn’t:
1) reversed causation (firefighters –> size of fire)
2) Confounding variable
3) unrelated
Does correlation establish causation? Does causation establish correlation?
Correlation DOES NOT establish causation.
Causation does establish correlation.
Goal of research experiments
a) establish internal validity which allows us to b) establish causation
holding constant
making the proposed third variable the same in all tests to check for notwithstanding causation
Third variable problem
means that correlation can never completely establish causation because there might always be a third (confounding) variable we are unaware of
how to experiment in spite of third variable problem
Experiment: technique for establishing causal relationships btwn variables
Manipulate one variable and see effect on other variables
If it changes it implies causation
Look for change in a way that gets around the third variable
confounding variable
third variable/some other thing causing relationship to exist (3rd variable problem)
Key ingredients of experiment:
manipulation
control
measurement
- Effectively manipulate IV
- Randomly assign subjects to groups that manipulation created
- Measure DV w valid, powerful reliable device while avoiding bias
Random assignment
assigning members to each sample group randomly so as to avoid skewed results
How does random assignment help experiment?
Little things that could be third variable will most likely get mixed up in both groups–w/ sufficiently large sample avg composition of both groups will be equivalent along all confounding variables!!
Tools for control:
Holding constant
Random assignment
independent variable
variable you change
dependent variable
variable you measure
Internal validity
characteristic of experiment that allows one to draw accurate inferences about causal relationships between independent and dependent variable
Replication crisis
realization over the history of psych (and other sciences) not everyone was running perfectly designed experiments leading to errors
Researchers take famous study and try to rerun and they find different results
So experiments are not replicating
most common error–replication crisis
Most common error here is false positives
Goal of replication/proving causation
if there is real effect→ conclude real effect
If there is not→ conclude not
Type I error
False positive–Type I error: testing effect that doesnt exist but by chance you find a result that seems like it exists
If there is not real effect→ conclude there is
Type II error
False negative–Type II error: testing effect that does exist but by chance you find a result that seems like it doesn’t exist
there is real effect→ conclude not real effect
Confirmatory testing
when u have a hypothesis and you test it and the answer confirms OG hypothesis
Exploratory testing
You don’t have clear hypotheses so you do a bunch of tests to see if u find something
It can lead to real things but u can’t draw strong conclusions for something u weren’t directly testing; likely that one result will happen by chance
Solving replication crisis
Solving replication crisis:
Distinguish exploratory from confirmatory
Preregister central tests prior to data collection
Replicate promising results (a lot) from exploratory analysis w preregistered followup
why do we preregister central tests prior to data collection
Proves confirmatory bc u intended to from beginning
If u do different experiment than preregister ppl will know
How did humans become the dominant species?
50k years ago Started in savannahs of Eastern africa, moved to middle East + Europe and wiped out competing species in few thousands years
Then spread to asia, pacific rim, americas, in tens of thousands of years
Little to no other organisms in places like south pole and space–we mastered globe and beyond
We’ve taken over animals, domestication
Humanity doesn’t have special abilities–venom, flying, etc–why are we dominant?
Human brain is special ability!
Human brain components as special ability
basic info
100 billion neuron cells
each have 1k synapses connected to 1 each=100 trillion connections
We are only animals to use tools, improve on them and pass them on
Complex language system
Math
Teaching (other species imitate, but this is explicit)
Engage in arts
Brain uses 20% of bodily resources
2-3lbs
Brain doesn’t directly influence world like special abilities usually do! How does brain count as special ability?
What mediates between brain and environment? how brain influences world? through the mind! Central problem in psychology!
Mind--allows brain interact w world Mind similar to computer Mental algorithm Cognitive processes Mind is software Brain is hardware
Demand characteristics
those aspects of observational setting that cause ppl to behave as they think observer wants/expects them to