Home
Subjects
Create
Search
Log in
Sign up
Upgrade to remove ads
Only $2.99/month
Social Science
Psychology
Experimental Psychology
psychology 12 final
STUDY
Flashcards
Learn
Write
Spell
Test
PLAY
Match
Gravity
Terms in this set (214)
the job of a researcher
-Act as empiricist by systematically observing the world
--Test theories through research; adapt these theories based on resulting data
-Use research to examine both basic and applied questions
-Test why, when and for whom an effect works
-Make our work public by submitting to academic journals
-Publicize our work
empiricism
using evidence from the sense or from instruments that assist the senses as the basis for conclusions
- can literally watch or listen to people
- can use surveys, timers, fMRI, eye-tracking
theory-hypothesis- data cycle
figuring out what might be happening and how would you figure it out
theory
a set of statements that describes general principles about how variables relate to one another
example: competing theories of attachment
hypothesis
a prediction; a way of stating the specific outcome the researcher expects to observe if the theory is correct
data
a set of observations
HARKing
Hypothesis After Results are Known
characteristics of a good theory
- supported by data
- falsifiability
-parsimony
a good theory is supported by data
evidence accumulates over many studies
- were often slow to give up our theories
a good theory is falsifiable
a feature of scientific theory in which it is possible to collect data that will prove the theory wrong
a good theory is parsimonious
all things being equal, the simplest solution is best
basic research
goals is to enhance the general body of knowledge
example: understanding the capacity of the human memory
applied research
conducted with a practical problem in mind, with the hopes of improving things
example: taking a drug and trying it on
teachers improving their teaching methods
translational research
uses the lessons from basic research to develop and test applications
example: derek and his alzheimers trail
mediating variables
explain why A is related to B
moderating variables
identify when or from whom A is related to B
example: waiting makes people anxious
-does uncertainty always make people anxious (WHEN)
-does uncertainty make everyone anxious (WHOM)
whats wrong with just learning from experience
NO COMPARISON GROUPS: need to compare what would happen both and without the thing you are interested in
USUALLY HAS CONFOUNDS: alternative explanations for the same effect
experience is insufficiently PROBABILISTIC: cant take into account enough cases to be sure about cause and effect
whats wrong with using our intuition
biased by good stories
biased by the AVAILABILITY HEURISTIC: things that easily come to mind seem more likely
biased by CONFIRMATORY HYPOTHESIS TESTING: asking can i think this for desirable conclusions and must I think this for undesirable conclusions
biased by our belief that none of this applies to us
biased by the tendency to avoid or reject answers we dont like
why cant we just trust authories
sometimes you can, but it's hard to tell....
variables
anything the researcher measures or manipulates that has an ability to vary or change
levels
the values of a variable in a study
constants
something that could vary but only has one level in a particular study
measured variables
a variable that is observed and recorded as it occurs naturally
dependent variable
manipulated variable
a variable that researcher controls
independent variable
dependent variable
an outcome variable; always measured, never manipulated
independent variable
a predictor variable; always manipulated in a true experiment, but can also be measured
conceptual definiton
a precise description of your construct (abstract concept)
operational variable/ definition
the specific way a variable will be manipulated or measured in a specific study
validity
how good are your measures
how good is your claim
it refers to the quality of a research method or a research claim
construct validity
how well did the researcher operationalize (measure) each variable?
example: using a valid scale
external validity
how well do the results generalize people or contexts other than those used in a specific study?
example: 44% of Americans struggle to be happy
- were participants chosen to represent different age groups and demographics
internal validity
in an experiment, how sure is the researcher that the independent variable caused in their dependent variable?
- no other variable can explain the changes in the resulting variable
*MUST BE A TRUE EXPERIMENT
example: the IV can only explain the DV
confound
something that differs between experimental groups other than the independent groups
example: do kids learn better from watching educational videos when parents actively watch with them?
- kids watch videos with their parents then take test
-OR take the test without watching the videos
statistical validity
how well do the numbers (data) support the researcher's conclusion?
- is it statistically significant? (how likely are the observed results be due to chance)
type 1 and 2 errors
TYPE 1: concluding that a relationship exists when one actually doesn't (false positive)
TYPE 2: concluding that no relationship exists when one actually does (false negative)
margin of error
indicates where the true value in the population probably lies
nominal scale
values are just labels or categories
example: are you happy (1=yes, 2=no)
ordinal scale
rank ordering
example: rank order the happiness of your closest friends
interval scale
equal difference between numbers reflect equal differences on the dimension being measured BUT no true zero point
example: 1= highly likely, 5= not likely
ratio scale
interval scale plus a true zero score
example: how many times did you feel happy today?
reliability of measures
degree to which a measure is CONSISTENT, stable, dependable
correlation coefficients
measures the strength and direction of association between two variable
strength
how well can you predict one thing by knowing about the other
ranges from 0 to +/- 1
direction
as one thing goes up, does the other go up or down
test-retest reliability
consistency over time
inner-rater reliability
consistency over observers
internal reliability
consistency across scale items
- self report
example: happiness scale
cronbach's alpha
Step 1: compute all possible correlations between items (item 1 with item 2, item 5 with item 7...)
Step 2: take the average of these correlations
Step 3: fancy math with the average correlation and the number of items= cronbach's alpha
.70 standard
Meaning the items are roughly correlated to each other
- The closer to 1.0 the better the scale's reliability
Subjective happiness scale ranges from 0.79 to 0.94
If LESS THAN 0.70, might have to revise items, or only select items with strong correlations
face validity
does the measure seem like a plausible one, given the construct of interest
example: face value
content validity
does the measure capture all parts of the construct of interest?
criterion validity
is the measure related to relevant objective outcomes?
- correlations between scores on your measure and objective outcomes
known groups paradigm
take groups you know to be different, give them the measure
convergent validity
is the measure related to other measures that assess similar construct
- correlations between scores on your measure and scores on other subjective, related measures
- what other measures should happiness be related to?
discriminant validity
is the measure NOT related to other measures that assess different construct
open- ended questions
allow people to respond to questions in their own ways
ADVANTAGES:
- allows people to tell you what is important to them
DISADVANTAGES: might never get to the stuff you care about: time consuming for participant and researcher
closed- ended questions
provide people with specific rating dimensions of interests
closed- ended questions: FORCED CHOICE
are you in favor of legalization
yes or no
closed- ended questions: LIKERT SCALE
i favor legalizing weed?
strongly agree to strongly disagree
closed- ended questions: SEMANTIC DIFFERENTIAL SCALE
legalizing weed is
good: 1 2 3 4 5
bad: 1 2 3 4 5
likable: 1 2 3 4 5
leading questions
make one answer seem clearly better than the other
example: do you think it is about time for weed to finally be legalized?
double barreled questions
asking two questions at once
example: do you think legalizing weed will decrease crime rate and lead to a happier, healthier population?
negatively worded questions
using negations (never, not, wouldnt, shouldnt) makes questions more cognitively difficult
example: i do not believe that weed should not be legalized
disagree or agree
question order
responses on earlier questions can affect interpretation of later questions
response sets: ACQUIESCENCE
yea- saying tendency (answering yes or agree to most questions without careful thought)
solution: reverse scoring
response sets: FENCE SITTING
staying close to the middle of the scale
solution: even numbers of scale; no mid-point
response sets: SOCIAL DESIRABILITY SCALE
concern over the impression one's response might convert
include a dew items to catch socially desirable responding
use surreptitious measures
participant observation
becoming part of the world you wise to observe as the researcher
contrived observation
observe behavior in research setting
example: speed dating
undisguised observation
participants know they are being observed
reactivity
people may change their behavior when they know someone is watching
disguised observation
participants DO NOT know they are being observed
partial concealment strategy
participants know they're being observed, but not WHY
latency
time between an event and a response
rating scales
outside parties make subjective judgments about behavior on specific measures
observer bias
observers' expectations can influence their interpretation of participants' behavior
observer expectancy effect
observers' expectations can actually influence participants' behavior
representative sample
all members of the population have an EQUAL CHANCE of being included in the sample
biased (unrepresentative) sample
some people in the population have a much better chance of being included in the sample than otehers
self- selection
sampling those who volunteer
simple random sampling
sample chosen completely at random from population
sampling frame
a full list of those in the population
cluster sampling
STEP 1: Break populations into clusters
STEP 2: Randomly sample clusters
STEP 3: Use everyone from the selected clusters
multistage sampling
STEP 1: Break populations into clusters
STEP 2: Randomly sample clusters
STEP 3: Randomly sample people from each randomly selected cluster
stratified random sampling
use demographic groups
STEP 1: identify the demographic groups for sample
STEP 2: Determine the percentage representation of each group within the population
STEP 3: Randomly sample the right number of people from that group to ensure the correct percentage representation within sample
over sampling
STEP 1: identify the demographic groups for sample
STEP 2: Determine the percentage representation of each group within the population
STEP 3: Randomly sampling MORE THAN the right number of people from that group than you would need to ensure correct percentage representation within sample
systematic sampling
pick a number and then take every th person
challenges with random sampling
CONTACTING PARTICIPANTS: over represent people who are reachable under represents those who arent
example: no phone, no mailing adress
RESPONSE RATE: even people who are reachable may not want to participate
ways to increase response rate ad how to deal with missing responses
solutions:
- follow up
- give incentives
- check hoe people who respond differs from people who dont
convenience sampling
using a sample of people who are readily available to participate
purposive sampling
Non-randomly recruiting a particular type of participant
snowball sampling
Recruitment via participants' social networks
quota sampling
pick a target number of participants in a particular category, recruit until you get a number
belmont report (1976)
A broad set of principles to guide research with human subjects. Motivated by problems with the Tuskegee Syphilis Study
principle of respect for persons
Informed consent, particularly for groups with reduced autonomy (children, prisoners, etc.)
principle of beneficence
protect participants from harm, ensure well-being
principle of justice
Fair balance of benefits and costs associated with research participation
Institutional Review Board (IRB)
committee that reviews research at universities to ensure ethical conducts
coercion and undue influence
COERCION: the explicit or implicit suggestions that someone who chooses not to participate will suffer negative consequences
UNDUE INFLUENCE: offering an incentive too attractive to refuse
informed consent
must provide participants with information about the study, particularly risks and benefits, so they can decide if they want to participate
information included and excluded: INFORMED CONSENT
Brief description of the procedure
Potential risks and benefits (including compensations)
Confidentiality
Right to withdraw
Contact information
informed consent is problematic
- people wont behave naturally
- some people cant give informed consent
- consent may be impractical or impossible to obtain
when informed consent can be waived
- behavior is public
-no more than MINIMAL RISK of harm ( an amount that would occur in everyday life)
decpetion
researchers withhold some details about the study, either through omission or commission
example: withholding true purpose of the study
confederates
an actor playing a role for the study
participants' views of deception
participants may prefer deceptive studies if conducted with respect and given a through debriefing
debriefing
informing participants about all aspect of the study AFTER the study is over
research misconduct: PLAGAIARISM
misrepresenting the ideas or words of others as one's own
data fabrication or data falsification
research with animals: REPLACEMENT
finding alternatives when possible
research with animals: REFINEMENT
minimize or eliminate animals' distress
research with animals: REDUCTION
use as few animals as possible
the publication process: 4 STEPS
manuscript is sent to one journal for consideration
editor assigns paper to associate editor
associated editor identifies 2- 5 reviews
- experts in the field
- remain anonymous to authors
- write review and decides publication
associate editor
takes reviews and makes the final decision about publication
criteria for publication
-Significance of the question (to the field, to society)
-Novelty
-Interestingness (ideterm-114a and findings)
-Methods high in construct, internal, and external validity
-Appropriate analyses and interpretation of data
-Good writing
possible outcomes of publication process
- accept as is
- accept with minor revisions
- revise for invited re-submission
-reject
bivariate correlation
a statistical relationship between exactly two variables
- measures two variables (either through self report, observation, etc.)
- most appropriate when both measures are continuous (ordinal, interval, or ratio)
bi-variate correlation: SCATTER-PLOTS
correlations are visually represented by these
independent sample t test: CATEGORICAL MEASURES AND BAR GRAPHS
MEASURES: nominal
BAR GRAPHS: visual representative of t test
five questions about correlations
1) Effect size (r value)
2) Statistical significance (p value)
3) Outliers
4) Restriction of range
5) Curvilinear associations
effect size
the strength of a relationship
r statistic: ranges from -1 to 1
- closer to -1 and 1= strong
- closer to zero= weak
direction: positive, negative, or zero
SMALL/ WEAL: 0.1
MEDIUM/ MODERATE: 0.3
LARGE/ STRONG: 0.5
- large effect size= more accurate predictions
statistical significance (p-value)
the likelihood that the effect would come out that strong by chance, assuming no effect in the real world
null hypothesis: assume there is zero association between the variables in the population
p value LESS than 0.05 is SIGNIFICANT
- GREATER than 0.05 IS NON SIGNIFICANT
outliers
an extreme score on either or both variables
restriction of range
when you only have part of the full scale for one or more of your variables represented in the sample
-artificially reduces the strength of correlation
curvilinear relationship
a relationship that is not well represented by a straight line (not linear)
3 requirements to determine causation: COVARAINCE
as one changes, the other also changes
3 requirements to determine causation: TEMPORAL PRECEDENCE
cause must precede effect
3 requirements to determine causation: INTERNAL VALIDITY
nothing else can explain the relationship between two variables
directionality problem
Failing on temporal precedence
third variable problem
failing on internal validity
multivariate designs
correlational studies that involve more than two variables
multivariate designs: LONGITUDINAL DESIGNS
measuring the same variables repeatedly at several points in time
multivariate design: MULTIPLE- REGRESSION ANALYSES
predicting an outcome from more than one predictor variable to narrow down the relationship with the predictor of interest
longitudinal study: CROSS SECTIONAL CORRELATIONAL
are the two variables correlated within the same point in time
comparing variable within a single time point
longitudinal study: AUTOCORRELATIONS
is each variable related to itself across time?
comparing same variable at different time points
longitudinal study: CROSS LAG CORRELATIONS
is the earlier measure of one variable associated with the later measure of the other variables?
compares a variable at an earlier time to a different variable at a later time
- addresses temporal precedence (reduces directionality problem) by being strong or really weak
goal of multiple regression
the regression has a strength and a direction that tells us how well the points fit a line in a scatterplot
multiple regression: CONTROLLING FOR
their effects on the outcome:
- would solve internal validity
- it tells us the association of one predictor while controlling for the other
multiple regression: CRITERION VARIABLES
the variable you are trying to predict; dependent variable
multiple regression: PREDICTOR VARIABLE
the variable that might be causing change in the criterion variable' independent variable
multiple regression: BETA
statistical representation of the relationship between each predictor variable and the criterion variable
-similar to the r in a bivariate correlation
-ranges from -1 to +1, has a p value
-difference is that the relationship controls for the other predictor variables
- has positive, negative, ans zero betas
mediation: 4 steps to establish mediation
1) Test relationship A to C (called path c)
2) Test relationship A to B (called path a)
3) Test relationship B to C (called path b)
4) Multiple-regression: Does c hold up after controlling for the mediator? (c')
control variables
Any variable the researcher intentionally holds constant across conditions
conditions
the levels or version of the independent variables
ways to test construct validity of manipulated variables
1. PILOT STUDY: conducted before the actual study to check the construct validity of an manipulation
2. MANIPULATION CHECK: an extra measure designed to see how well a manipulation worked
types of conditions: CONTROL GROUPS
a condition that is supposed to represent no treatment or a neutral state
types of conditions: TREATMENT GROUPS
the conditions of interest which are compared to the control group
types of conditions: PLACEBO
a control group who believes they are the treatment group with the goal of ruling out expectancy effects
confounds
anything that differs between your group OTHER THAN the levels of the independent variables
design confounds
Something that inherently varies along with the independent variable
- creates systematic validity (big PROBLEM)
design confounds: UNSYSTEMATIC VARIABILITY
created when something differs among participants but does NOT systematically co-occur with the independent variable
random assignment
each participant has an equal chance of being placed into any group
a true experiment
manipulation of one or more independent variables
random assignment to conditions
selection effects
When the kind of person in one condition are systematically different from the ones in other conditions
types of experiments: BETWEEN SUBJECT DESIGN
each participant is only one experimental condition
also called independent group designs
between subject design: POSTTEST-ONLY DESIGN
participants undergo the manipulation (one condition) and then complete the measures (once)
between subject design: PRETEST POSTTEST DESIGN
participants first complete the measures, then the manipulation, then the measures again
pretest posttest design: ADVANTAGES AND DISADVANTAGES
ADVANTAGES:
-test and control for selection effects
-test and control for failures of random assignment
DISADVANTAGES:
-might create demand characteristics
- people might think they should be consistent in their responses
types of experiment: WITHIN SUBJECT DESIGN
each participant is in all experimental condition
within subject design: ADVANTAGES
-guarantees equivalence of groups (no SELECTION EFFECTS)
- STATISTICAL POWER: ability of a study to get a statistically significant effect, assuming the effect is real
between subject design: ADVANTAGES
ORDER EFFECTS: a confound that occurs when experiencing one condition changes how participants react to subsequent conditions in a within subject design
order effects: PRACTICE EFFECTS
participants get better at the measures
order effects: FATIGUE EFFECT
participants get worse at the measures
order effects: CARRYOVER EFFECT
effects of one condition contaminate subsequent responses
order effect: SENSITIZATION EFFECT
participants become suspicious or clued in from earlier conditions
order effect: COUNTERBALANCING
randomly assigning participants to experience the conditions in different orders
demand characteristics
something in the experiment allows participants to guess what the study is about and change their behavior accordingly
avoid demand characteristics: DOUBLE BLIND STUDY
neither the participants nor the researcher knows which condition the participants are in`
one-group pretest-posttest design
One group of participants complete the dependent variable, then experience a "treatment", then complete the measures again.
threats to internal validity: MATURATION
a change in the dependent measure that occurs naturally over time
threats to internal validity: HISTORY
a change in the dependent measure that occurs because of an intervening event
threats to internal validity: REGRESSION TO THE MEAN
a natural shift away from extreme scores at the pretest
threats to internal validity: ATTRITION
participants dropping out of the study in a non- random way
threats to internal validity: TESTING
includes practice effects and fatigue effetcs
threats to internal validity: INSTRUMENTATION
occurs when the measurement "tool" changes over time
null effects
the independent variable did not make a difference in the dependent variable
5 sources of null effects:
1. hypothesis or theory is wrong
2. weak manipulation
3. measures wasn't sensitive enough
-includes ceilings and floor effect
4. reverse design confound
5. measurement error: noise caused by imperfect measures
ceiling effect
Scores started out at the top of the scale (no room for an increase).
floor effect
scores started out at the bottom of the scale (no room for a decrease)
factorial design
an experiment with 2 or more independent variables
factorial design: FACTOR
another name for an independent variable
factorial design: CONDITION
one level of an independent variable
factorial design: CELL
in a between subject factorial design, a particular combination of the conditions of each independent variable
numbering notation: NUMBER OF NUMBERS
refers to the total number of factors (IV) in a design
2 x 2 = 2 factors
2 x 2 x 2 = 3 factors
numbering notation: NUMBER VALUES
refers to the number of levels of each factor (IV)
2 x 2 = 2 factors with 2 levels each (4 levels total)
2 x 3 = 2 factors, with 2 levels and one with 3 levels
**order of numbers makes no difference
numbering notation: PRODUCT OF ALL NUMBERS
indicates the number of cells
2 x 2 = 4 cells
2 x 3 = 6 cells
main effect
the overall effect of one independent variable on the dependent variable
main effect: MARGINAL MEAN
the mean for ONE level of ONE independent variable
interaction effect
A test of whether the effect of one independent variable depends on the level of the other independent variable.
interaction effect: CROSSOVER INTERACTION
the effect of one independent variable on the dependent variable REVERSES across levels of the other independent variable
- simple effects in opposite direction
interaction effect: SPREADING INTERACTION
the effect of one IV on the DV is STRONGER at one level of the IV than it is at the other level
- simple effects are in the same direction OR only one simple effect
quasi experiment
researcher identify an independent variable but do not manipulate it or randomly assign participants to conditions
- necessary when the iv cannot or should not be manipulated
repeated measures quasi experiment
1. INTERRUPTED TIME SERIES DESIGN: measures the DV repeatedly before, during, and after some event
2. INTERRUPTED TIME SERIES WITH REVERSAL: in which measurements continue after things have gone back to their original state
types of quasi experiments: NONEQUIVALENT GROUP DESIGN
includes a control group for which the key event does not occur
quasi experiment: SELECTION EFFECTS
the major potential confound in nonequivalent groups designs
quasi experiment: MATCHED GROUP
a particularly useful strategy in quasi experimental designs but also relevant to true experiments
small N design
gathering a lot of information from a small sample instead of a small amount of information from a big sample
single N design
a study of a single person or animal's experience
often called a CASE STUDY
advantages of a smaller or single N design
- allows researchers to study rare people or events
- provides rich data about a narrow span of experience (depth)
- avoids problems associated with averaging across participants
primary disadvantage of small or single N design
-finding may not generalized (low external validity)
replicability
If you did the same study again, would you get the same results?
2 important about replicability
1. how should we define same study
2. how should we define same results
direct replication
repeating the methods of a study as closely as possible to see you get the same results with a different sample
direct replication: GOALS
1. to make sure the effect was not specific to those participants, that lab, those researchers, etc.
2. to make sure the researchers did not engage in questionable practices, or even outright found
conceptual replication
testing the same research question with different methods
conceptual replication: GOAL
to make sure the effects was not specific to a particular operational definition of the variable
replication-plus-extension study
repeating the original methods and adding something new
replication-plus-extension study: GOALS
to determine boundaries of the effect
2 definitions of a successful replication
STATISTICAL SIGNIFICANCE: was the effect significant before? is it this time?
EFFECT SIZE: what was the effect size before? is the new effect size similar?
meta- analysis
a mathematical compilation of studies that all tested the same effect
meta- analysis: OVERALL EFFECTS AND MODERATORS
OVERALL EFFECTS: priming ---> behavior
MODERATORS: time between priming and behavior measure, nature of prime (words or pictures)
meta- analysis: FILE DRAWER PROBLEM
meta- analyses tend to overestimate effects because null and contradictory effects are difficult to publish
file drawer problem: SOLUTION
contact colleagues to collect "failed" studies
PRE-REGISTRATION: documenting your planned study before conducting it and making results available regardless of the success of the study
ecological validity
does the situation you create in your study resembe the real world?
experimental realism:
Is the situation you create in your study engaging, eliciting "real" emotions, motivations, and behaviors?
YOU MIGHT ALSO LIKE...
Psych 12 Midterm 3
73 terms
Psy341 Research Design Final
97 terms
Morling: Research Methods in Psychology - Carroll…
96 terms
Intro to Research: Chapters 8-10
62 terms
OTHER SETS BY THIS CREATOR
psych 110: vision system to motor
12 terms
Psychology 12: chapter 8-10
98 terms
Psychology 12: Chapter 1-3
51 terms
starbucks shots and syrups
53 terms