Use LEFT and RIGHT arrow keys to navigate between flashcards;
Use UP and DOWN arrow keys to flip the card;
H to show hint;
A reads text to speech;
211 Cards in this Set
- Front
- Back
Ways to know things
|
tradition (aka tenacity), authority, observation
|
|
problems with “knowing”
|
sometimes our “knowledge” is wrong
we allow “experts” to give us info |
|
types of OHI
|
Causal reasoning, Probabilistic reasoning
|
|
errors in observation
|
inaccurate observtion, over-generalization, selective observation (streetlight effect), illogical reasoning
|
|
Descarte’s dilemma
|
how do I know anything? what if it’s an hallucination? (and thus the scientific method was born)
|
|
Correcting for inaccurate observation
|
Observe as a conscious, purposeful activity
use instrumentation |
|
Correcting for Over-Generalization
|
Use sufficiently large samples
replicate results |
|
Correcting for selective observation
|
specify in advance number and types of observations to be made
pay attention to deviant cases - make sure to notice all the times the streetlight doesn’t change state |
|
Correcting for illogical reasoning
|
make logical reasoning a conscious, purposeful activity
|
|
Name Movements in Epistemology
|
Pre-Modern, Modern, Post-Modern
|
|
Describe Pre-Modern Movement
|
“Naive realism,” assuming your observation is reflective of the objective truth
|
|
Describe Modern Movement
|
Descarte and his cronies
there is a truth out there, but we can have different opinions (through different observations) that can be misleading |
|
Describe the blind men and the elephant
|
Thanks! Note that this is an example of the Modern Epistemological Movement
|
|
Describe Post-Modern Movement
|
how do we know there is an objective truth at all? we have a complicated system that makes the world sensible, but it’s not meaningful to assess “truth value”
we make reality through consensus, independent of any objective truth - |
|
The two foundations of scientific research
|
Logic, Observation (together forming Theory)
|
|
Theory
|
the collected understanding of a topic - is different from philosophy because it requires observation, and is different from belief because it requires logic
|
|
Describe Variable/Attribute
|
A variable has attributes... if Variable = Profession, Attribute = plumber, politician, etc...
|
|
Different types of research
|
Exploratory, Descriptive, Explanatory
|
|
Idiographic vs. Nomothetic Research
|
idiographic - an attempt to find all causal elements in a single case
nomothetic - gives a generalizable (but incomplete) explanation |
|
Inductive vs. Deductive Research
|
Inductive - start with observations, reason an explanation, develop a theory
deductive - develop a theory, formulate an experiment, observe if results match theory |
|
Qualitative vs. Quantitative Research
|
qualitative - description of things as you see them, what was it like in human terms, how did you feel? quantitative - description of amounts
|
|
Epistemology
|
the study of how we know - the theory of knowledge
|
|
Paradigm
|
a “set of instructions” - a constellation of beliefs, values, techniques shared by the members of a given community
|
|
“Inventor” of the paradigm shift
|
Kuhn (in his “Structure of Scientific Revolutions” (1970)
|
|
Paradigm shift
|
Kuhn’s concept is that the evidence against one paradigm will build up until the paradigm collapses and makes way for another -- like the Copernican revolution
|
|
Bacon
|
“Knowledge is Power” - understanding itself is not good enough, we want knowledge that will allow us to manipulate the world around us
|
|
Comte
|
Positivism!
|
|
Positivism
|
Comte’s theory: you can know objective reality
the way is science |
|
Three stages of Positivism
|
Theological (authority)
Metaphysical (associated with the Enlightenment - philosophical) Positive (we can change or fix things regardless of doctrine or philosophy - to know, we need something both logically consistent and empirically demonstrable - thus, science |
|
Wundt
|
founder of psychology - followed the positivistic condition to form psychology
|
|
Problems with positivism
|
The human factor - we can create models based on rational human behavior, but people just aren’t rational, which is why economic prediction cannot be precise - but we can still predict why a person does not behave rationally (the post-positivistic)
|
|
why do we think objectivity is so preferable?
|
can’t total objectivity lead to exploitation? and objectivity can also obfuscate, by forcing you to ignore other possibly relevant information. and is it even possible for a scientist to be objective? aren’t you showing subjectivity just by selecting a topic to research?
|
|
the construction of a deductive theory
|
pick a topic
inventory knowledge and thought about that topic (literature review) research what is known about two variables within that topic create a propositional structure, predicting what is going to happen between the IV and DV show that a relationship exists between those variables |
|
the construction of an inductive theory
|
start with a broad question
make observations based on that question start an open ended inquiry (not making predictions) make conclusions |
|
another name for inductive theory
|
grounded theory
|
|
informed consent
|
subject must understand what is being presented - note that the process must be tailored to the participants
|
|
Value-Neutral Science
|
science for science’s sake
|
|
Belmont Report
|
destroyed the idea of value-neutral science
|
|
Belmont Report’s Criteria for human-based research
|
Beneficence (research must benefit science, humanity, and the participants)
Respect (for the autonomy of the participants) Justice (a fair distribution of costs and benefits among all segments of the population) |
|
Anonymity vs Confidentiality
|
Anonymity: researcher cannot tie a response to a participant
Confidentiality: researcher can identify the participant, but will keep that information confidential |
|
exploratory research
|
to satisfy curiosity
to develop methods to be used in a later study |
|
descriptive research
|
a more systematic presentation of data
making a more exhaustive investigation (than exploratory research) |
|
explanatory research
|
finding cause
|
|
Aristotle’s 4 types of cause
|
material (what is it made of?)
efficient (our usual idea of cause - action/reaction) formal (the shape of the thing - what function does it support?) final (the use - what is the goal?) |
|
Aristotle’s material cause
|
what is it made of?
|
|
Aristotle’s efficient cause
|
action/reaction
|
|
Aristotle’s formal cause
|
the shape of the thing - what function does it support?
|
|
Aristotle’s final cause
|
the use of the artifact - what is the goal?
|
|
Three criteria of Nomothetic Cause
|
correlation
time order (cause comes before effect) non-spuriousness (no third variable) |
|
necessary cause
|
something you must have for something else to occur
|
|
sufficient cause
|
if you have it, something will occur
|
|
Unit of Analysis
|
the very thing being studied (tricky)
remember, statistics of a larger group are not the same as the statistics of the individuals |
|
Cross-Sectional Study
|
study happening at one point in time
|
|
Longitudinal Study
|
takes place over a period of time
could be a Trend, Cohort, or Panel Study |
|
Trend Study
|
looking at an entire population over a period of time
|
|
Cohort Study
|
looking at a certain subset of people over a period of time
|
|
Panel Study
|
looking at a specific set of people each time over a period of time
|
|
Operationalization
|
deciding how to measure a conceptualized variable
because how do you put a number on, say, social class? |
|
Kaplan’s classes of things to be measured
|
Direct observables
indirect observables constructs |
|
indirect observables
|
history books, checkmarks on questionnaires, etc
|
|
constructs
|
theoretical creations
based on observations, but cannot be observed directly IQ, compassion, predjudice concepts as opposed to real “things” |
|
Reification
|
regarding constructs as real
|
|
indicator
|
a sign of the presence of a concept
visiting hospitals for Hanukkah if the concept is compassion, etc. |
|
dimension of a concept
|
a specifiable aspect of the concept
divisions such as “compassion for humans” vs. “compassion for animals” |
|
Three definitions of a concept
|
real - not useful
nominal - represents a consensus, is useful for communication operational - specifies how a concept can be measured - gives you a definition you can test, regardless of how “true” it is |
|
Conceptual Order
|
conceptualization
nominal definition operational definition real-world measurements |
|
Range of Variation
|
be aware of concealing useful datat by not allowing it to be reported
provide “oppose very much” instead of stopping at “don’t support” |
|
attributes : mutual exclusivity
|
we must define “employed” and “unemployed” in such a way that no one can be both
|
|
attributes : exhaustiveness
|
offering only “republican” and “democrat” will cause you to miss data from independents, etc.
|
|
Nominal Measures
|
have only exclusivity and exhaustiveness
provide labels for characteristics allow you to say that two things are the same or are different (but no more) |
|
Ordinal Measures
|
allows an ordered ranking, but pays no heed to intervals
|
|
Interval Measures
|
allows a measure of distances between attributes
does not allow X = 2Y often used with an arbitrary scale |
|
Ratio Measures
|
adds a zero point, allows averages
|
|
Conceptions
|
an idiosyncratic understanding of a real thing - may not line up neatly with actual reality
|
|
Concepts
|
a shared, intersubjective understanding of a real thing - may not line up neatly with actual reality
|
|
Ways to know things
|
tradition (aka tenacity), authority, observation
|
|
problems with “knowing”
|
sometimes our “knowledge” is wrong
we allow “experts” to give us info |
|
types of OHI
|
Causal reasoning, Probabilistic reasoning
|
|
errors in observation
|
inaccurate observtion, over-generalization, selective observation (streetlight effect), illogical reasoning
|
|
Descarte’s dilemma
|
how do I know anything? what if it’s an hallucination? (and thus the scientific method was born)
|
|
Correcting for inaccurate observation
|
Observe as a conscious, purposeful activity
use instrumentation |
|
Correcting for Over-Generalization
|
Use sufficiently large samples
replicate results |
|
Correcting for selective observation
|
specify in advance number and types of observations to be made
pay attention to deviant cases - make sure to notice all the times the streetlight doesn’t change state |
|
Correcting for illogical reasoning
|
make logical reasoning a conscious, purposeful activity
|
|
Name Movements in Epistemology
|
Pre-Modern, Modern, Post-Modern
|
|
Describe Pre-Modern Movement
|
“Naive realism,” assuming your observation is reflective of the objective truth
|
|
Describe Modern Movement
|
Descarte and his cronies
there is a truth out there, but we can have different opinions (through different observations) that can be misleading |
|
Describe the blind men and the elephant
|
Thanks! Note that this is an example of the Modern Epistemological Movement
|
|
Describe Post-Modern Movement
|
how do we know there is an objective truth at all? we have a complicated system that makes the world sensible, but it’s not meaningful to assess “truth value”
we make reality through consensus, independent of any objective truth - |
|
The two foundations of scientific research
|
Logic, Observation (together forming Theory)
|
|
Theory
|
the collected understanding of a topic - is different from philosophy because it requires observation, and is different from belief because it requires logic
|
|
Describe Variable/Attribute
|
A variable has attributes... if Variable = Profession, Attribute = plumber, politician, etc...
|
|
Different types of research
|
Exploratory, Descriptive, Explanatory
|
|
Idiographic vs. Nomothetic Research
|
idiographic - an attempt to find all causal elements in a single case
nomothetic - gives a generalizable (but incomplete) explanation |
|
Inductive vs. Deductive Research
|
Inductive - start with observations, reason an explanation, develop a theory
deductive - develop a theory, formulate an experiment, observe if results match theory |
|
Qualitative vs. Quantitative Research
|
qualitative - description of things as you see them, what was it like in human terms, how did you feel? quantitative - description of amounts
|
|
Epistemology
|
the study of how we know - the theory of knowledge
|
|
Paradigm
|
a “set of instructions” - a constellation of beliefs, values, techniques shared by the members of a given community
|
|
“Inventor” of the paradigm shift
|
Kuhn (in his “Structure of Scientific Revolutions” (1970)
|
|
Paradigm shift
|
Kuhn’s concept is that the evidence against one paradigm will build up until the paradigm collapses and makes way for another -- like the Copernican revolution
|
|
Bacon
|
“Knowledge is Power” - understanding itself is not good enough, we want knowledge that will allow us to manipulate the world around us
|
|
Comte
|
Positivism!
|
|
Positivism
|
Comte’s theory: you can know objective reality
the way is science |
|
Three stages of Positivism
|
Theological (authority)
Metaphysical (associated with the Enlightenment - philosophical) Positive (we can change or fix things regardless of doctrine or philosophy - to know, we need something both logically consistent and empirically demonstrable - thus, science |
|
Wundt
|
founder of psychology - followed the positivistic condition to form psychology
|
|
Problems with positivism
|
The human factor - we can create models based on rational human behavior, but people just aren’t rational, which is why economic prediction cannot be precise - but we can still predict why a person does not behave rationally (the post-positivistic)
|
|
why do we think objectivity is so preferable?
|
can’t total objectivity lead to exploitation? and objectivity can also obfuscate, by forcing you to ignore other possibly relevant information. and is it even possible for a scientist to be objective? aren’t you showing subjectivity just by selecting a topic to research?
|
|
the construction of a deductive theory
|
pick a topic
inventory knowledge and thought about that topic (literature review) research what is known about two variables within that topic create a propositional structure, predicting what is going to happen between the IV and DV show that a relationship exists between those variables |
|
the construction of an inductive theory
|
start with a broad question
make observations based on that question start an open ended inquiry (not making predictions) make conclusions |
|
another name for inductive theory
|
grounded theory
|
|
informed consent
|
subject must understand what is being presented - note that the process must be tailored to the participants
|
|
Value-Neutral Science
|
science for science’s sake
|
|
Belmont Report
|
destroyed the idea of value-neutral science
|
|
Belmont Report’s Criteria for human-based research
|
Beneficence (research must benefit science, humanity, and the participants)
Respect (for the autonomy of the participants) Justice (a fair distribution of costs and benefits among all segments of the population) |
|
Anonymity vs Confidentiality
|
Anonymity: researcher cannot tie a response to a participant
Confidentiality: researcher can identify the participant, but will keep that information confidential |
|
exploratory research
|
to satisfy curiosity
to develop methods to be used in a later study |
|
descriptive research
|
a more systematic presentation of data
making a more exhaustive investigation (than exploratory research) |
|
explanatory research
|
finding cause
|
|
Aristotle’s 4 types of cause
|
material (what is it made of?)
efficient (our usual idea of cause - action/reaction) formal (the shape of the thing - what function does it support?) final (the use - what is the goal?) |
|
Aristotle’s material cause
|
what is it made of?
|
|
Aristotle’s efficient cause
|
action/reaction
|
|
Aristotle’s formal cause
|
the shape of the thing - what function does it support?
|
|
Aristotle’s final cause
|
the use of the artifact - what is the goal?
|
|
Three criteria of Nomothetic Cause
|
correlation
time order (cause comes before effect) non-spuriousness (no third variable) |
|
necessary cause
|
something you must have for something else to occur
|
|
sufficient cause
|
if you have it, something will occur
|
|
Unit of Analysis
|
the very thing being studied (tricky)
remember, statistics of a larger group are not the same as the statistics of the individuals |
|
Cross-Sectional Study
|
study happening at one point in time
|
|
Longitudinal Study
|
takes place over a period of time
could be a Trend, Cohort, or Panel Study |
|
Trend Study
|
looking at an entire population over a period of time
|
|
Cohort Study
|
looking at a certain subset of people over a period of time
|
|
Panel Study
|
looking at a specific set of people each time over a period of time
|
|
Operationalization
|
deciding how to measure a conceptualized variable
because how do you put a number on, say, social class? |
|
Kaplan’s classes of things to be measured
|
Direct observables
indirect observables constructs |
|
indirect observables
|
history books, checkmarks on questionnaires, etc
|
|
constructs
|
theoretical creations
based on observations, but cannot be observed directly IQ, compassion, predjudice concepts as opposed to real “things” |
|
Reification
|
regarding constructs as real
|
|
indicator
|
a sign of the presence of a concept
visiting hospitals for Hanukkah if the concept is compassion, etc. |
|
dimension of a concept
|
a specifiable aspect of the concept
divisions such as “compassion for humans” vs. “compassion for animals” |
|
Three definitions of a concept
|
real - not useful
nominal - represents a consensus, is useful for communication operational - specifies how a concept can be measured - gives you a definition you can test, regardless of how “true” it is |
|
Conceptual Order
|
conceptualization
nominal definition operational definition real-world measurements |
|
Range of Variation
|
be aware of concealing useful datat by not allowing it to be reported
provide “oppose very much” instead of stopping at “don’t support” |
|
attributes : mutual exclusivity
|
we must define “employed” and “unemployed” in such a way that no one can be both
|
|
attributes : exhaustiveness
|
offering only “republican” and “democrat” will cause you to miss data from independents, etc.
|
|
Nominal Measures
|
have only exclusivity and exhaustiveness
provide labels for characteristics allow you to say that two things are the same or are different (but no more) |
|
Ordinal Measures
|
allows an ordered ranking, but pays no heed to intervals
|
|
Interval Measures
|
allows a measure of distances between attributes
does not allow X = 2Y often used with an arbitrary scale |
|
Ratio Measures
|
adds a zero point, allows averages
|
|
Conceptions
|
an idiosyncratic understanding of a real thing - may not line up neatly with actual reality
|
|
Concepts
|
a shared, intersubjective understanding of a real thing - may not line up neatly with actual reality
|
|
Hermeneutic circle
|
winnowing down
look at a piece of a larger phenomenon and the conceptualization of that piece changes your conception of the larger construct which leads you to another small piece, etc |
|
range of Variation
|
size of the spectrum of attributes
|
|
variation between extremes
|
what size gradiations?
how specific does the information need to be? |
|
Reliability
|
vs. validity
precision - getting the same results each time |
|
assessing reliability
|
test-retest method
split-half method use of established measures |
|
validity
|
vs. reliability
accuracy - how does it measure up to reality? |
|
categories of validity
|
face validity
predictive validity construct validity content validity |
|
face validity
|
is it valid on the face of it -- unlike the shoe-size/math skills study?
|
|
predictive validity
|
how well does it actually predict reality?
|
|
construct validity
|
does the measure rate as expected with other variaboes?
|
|
content validity
|
how well does it establish a true understanding?
|
|
reliability vs. validity
|
as reliability rises, validity may decrease.
|
|
Nonprobability Sampling
|
generally no control over representativeness
Types: reliance on available subjects purposive sampling snowball sampling informants quota sampling |
|
Grounded theory
|
inductive theory
|
|
purposive sampling
|
aka judgmental sampling
not interested in representativeness seen in Grounded Theory a nonprobability technique |
|
snowball sampling
|
asking subjects to recommend other subjects - useful when you cannot select freely of a, say, secretive group
a nonprobability technique |
|
informant sampling
|
based on a subject’s willingness to talk - not unlike mafia informants
a nonprobability technique |
|
quota sampling
|
when you think you know what the larger population looks like, and attempting to choose sample based on percentages in the larger population
a nonprobability technique |
|
probability sampling
|
attempting to choose samples that wil be representative of the larger population
|
|
sampling bias
|
when the sample does not have the same aggregate characteristics as the larger population
|
|
EPSEM
|
Equal Probability of Selection Method - making you more likely to randomly choose a representative population
|
|
Element
|
the unit of analysis before the actual analysis
|
|
Population
|
aggregate of elements from which the sample is selected (made of theoretical and study populations)
|
|
theoretical vs study population
|
theoretical - the population you’re interested in studying - such as “overweight americans”
study - the population you actually study - such as “overweight americans from the contiguous states” |
|
Parameter
(Probability Theory) |
the actual mean in the population
|
|
The Normal Curve
|
the bell shape curve created when taking enough samples around a value
|
|
sampling error
|
how likely we are right
|
|
Standard Error
|
S = sqrt[(p*q)/n]
in the bell curve, 68% of the population (34% on either side of mean) falls within S |
|
Sampling Frame
|
the device by which you choose your sample
if you’re picking names out of the phone book, the phone book is the sampling frame |
|
SRS
|
simple random sampling
probability design everyone gets a number, and you choose subjects randomly based on number |
|
Systematic Sampling
|
choosing every kth element
k = sampling interval = pop.size/samp.size sampling ratio = 1/k |
|
Periodicity
|
a problem with systematic sampling
when the list is arranged in such a way that the systematic sampling gives you a nonrepresentative study like the sergeants |
|
Sergeant Study
|
an example of periodicity
in a study of enlisted men, they chose every 10th member of the population for the study, but since there is one sergeant for every 9 privates, only sergeants were chosen, and were thus non-representative |
|
Stratified Sampling
|
a modification of systematic sampling
organize the sampling frame into layers |
|
Sampling error
|
how likely we are representing the population
|
|
Cluster Sampling
|
aka multistage cluster sampling
List, Sample, Repeat |
|
correcting for disproportionate sampling
|
oversample an underrepresented population
weight the result by inverting the resultant data |
|
Physiological IV
|
inserting a direct physiological element
|
|
experience IV
|
putting a subject in a context
|
|
stimulus IV
|
applying a stimulus to judge effect
|
|
unwanted variables
|
extraneous (confounding)
nuisance variable |
|
extraneous variable
|
works between groups, moving the control and experimental groups closer or further apart
|
|
nuisance variable
|
spreads out scores within a group - doesn’t really move the distribution around the mean, just makes the curve flatter and wider
|
|
ways to measure DVs
|
measure correctness of responses - questionnaire
measure rate or frequency degree/amount latency/duration - how fast the response, how long it lasts |
|
Randomizing
(controlling for unwanted variables) |
if everyone has an equal chance you may neutralize the variable
|
|
ellimnation
(controlling for unwanted variables) |
removing the variable altogether
|
|
balancing
(controlling for unwanted variables) |
distribute the variable evenly among each group
such as spreading gender evenly amongst the subjects - make sure we use both male and female researchers for each subject, for example |
|
counterbalancing
(controlling for unwanted variables) |
such as correcting for order effects - make sure each subject is exposed to each order, or within groups making sure each order is represented in the research
|
|
incomplete counterbalancing
|
when we don’t have enough subjects to properly counterbalance, randomly selecting sequences in hopes of giving a representation of all possible sequences
|
|
Type I Error
|
Incorrectly rejecting the Null Hypothesis
|
|
Type II Error
|
Incorrectly failing to reject the Null Hypothesis
|
|
Significance Level
|
indicates confidence... .05 confidence level means 95% surety
aka Alpha |
|
Rosenthal Effect
|
“gifted kids” performed better
|
|
Controlling for Experimenter Effects
|
balancing (difficult)
standardizing (such as giving a script) |
|
Controlling for Experimenter Expectancies
|
standardization
automation - use a machine single/double blind experiment |
|
Participant Effects
|
Demand Characteristics
Good participant effect response bias (yea saying / nay saying) response set |
|
Controlling for Demand characteristics
|
double-blind experiment
deception (but IRBs don’t like it) |
|
Controlling for Response Bias
|
word questions such that “no” is sometimes a positive response
|
|
Controlling for Response Set
|
Be careful about the context - including the wording of questions
|
|
r
|
pearson product-moment correlation coefficient
-1 to 1 |
|
problems with r
|
restriction of range will limit data
curvelinear plots can’t be described with r |