Home
Subjects
Explanations
Create
Search
Log in
Sign up
Upgrade to remove ads
Only $2.99/month
Social Science
Psychology
Experimental Psychology
RDA FINAL
STUDY
Flashcards
Learn
Write
Spell
Test
PLAY
Match
Gravity
Terms in this set (37)
goodness of fit
a test for hypotheses abt the dist of a single nominal level variable
test of association
test for hypoth abt. the assoc. b/t 2 nominal level variables
no preference (goodness of fit)
Ho states that there is no pref among the diff categories; the pop id divided equally among categories
no diff from a comparison pop
Ho states that the freq list. for 1 pop (your sample) is not diff from the list. that is known to exist for another pop (a known quantity)
causual analysis
we infer that even x is a cause of event y
the principle of temporal precedence
x preceds y
the principle of covariation
x and y covay; if x occurs then y occurs
the mechanism principle
there is a plausible mechanism linking x and y
the parsimony principle
we can eliminate other possible causes that may be cofounded with x
strong inference
1. devising alternate hypoth
2. devising a crucial experiment with alternative possible outcomes, each of which will, as nearly as possible, exclude one or more of the hypoths
3. carrying out the experiment so as to get a clean result and
4. recycling the procedure, making suubhypoths or sequential hypoths to refine the possibilities that remain
ethical principles
1. seek advice - carefully evaluate ethical acceptability
2. determine risks
3. assume responsibility - researcher + assistants are obliged to see that ethical practices are followed
4. obtain informed consent
5. seek alternatives to concealment of deception
6. allow participants to withdraw, free at any time
7. protect participants from harm, discomfort, and danger
8. debrief participants - explain study after data collection
9. correct undesirable consequences, including long term effects
10. assume confidentiality
criteria for evaluating quality of measurement
1. sensitivity - does the measuring instrument display variations in recordings when the behavior being observed varies? are the variations in recordings usefully differentiated?
2. uniqueness - does the measuring instrument only reflect variations in the target behavior or characteristic?
3. reliability - how constant are measurements a cross examiners (inter-judge agreement), across occasions (stability), across indicators (internal consistency or multi-method convergence)?
4. construct validity - does the measuring instrument adequately represent a theoretical standard?
testing situation
types of q's:
1. relevant q
2. irrelevant q
3. control q (CQT) - not related but arousing
4. concealed infö q (CIT) - only person w/ direct knowledge would react
random assignment
-assigning participants to the conditions of an experiment such that each participant has an equal chance of being in a given condition
-increases confidence that experimental and control groups will not differ with respect to pre-info (pre-treatment) levels of the DV
-increases confidence that groups will not differ in their level of any extraneous casual variable
block randomization
-the most common technique for carrying out random assignment in the random groups design. each block includes a random order of the conditions, and there are as many blocks as there are participants in each condition of the experiment
random variable
any variable that is statistically unlikely to be correlated with an IV bc participants were randomly assigned to levels of that IV
random selection/sampling
-sampling procedure in which every individual in a pop being studied has an equal chance of being selected
-increases confidence that characteristics of the sample accurately represent the pop
counterbalancing
used to distribute the influence of sequence equally across all stimuli; each stimulus/condition has an equal chance of occurring at each position in a sequence
-order effects: primacy and recency in memory experiments
-carry-over effects: discrimination trials
-practice effects: performance improves w/ repitition
complete counterbalancing
-all possible sequences are presented to each participants several times
-problematic if many stimuli are of interest (if IV has several levels)
incomplete counterbalancing
-not every possible sequence must be presented, but each stimulus/condition occurs in each position, and each stimulus is both preceded and followed by each other stimulus. each sequence (and therefore each condition) is administered to each participant only once
randomized counterbalancing
-stimuli/conditions are randomly assigned to positions within a sequence, such that each stimulus has an equal chance of occurring at each position
-does not work well with few stimuli
matching
the equal dist. of relevant variables across levels of the IV, so that the influence of the relevant variable(s) is constant across levels of the IV
matching by constancy
holding an extraneous casual variable constant (when studying effects of 2 drugs on depression, limit participant to those not currently on medication)
matching by factoring
random assignment to 1 IV within levels of a second IV
matching by pairing
a pair is formed by matching on the basis of a similar value of a relevant variable before being randomly assigned to levels of the IV
advantages of multi-factor experiment
-multi factor designs allow the study of interaction effects.
-increase both in the precision/sensitivity and generalizability of results. whenever an extraneous causal variable (ECV) is made into a random variable, experimental results increase in generalizability but decease in precision. When an ECV is held constant, precision increases by generalizability decreases
-have a statistical advantage of reducing random/unexplained variation in the DV
interaction effect
a relationship b/t at least 2 IVs and 1 DV, in which the effect of factor A on the DV depends on levels of factor B
disadvantages of multi-factor experiments
-multi-factor designed may produce interaction effects that are difficult to interpret
-demand greater resources (more participants, more time, more materials)
-the statistical procedures for multi-factor designs become more complex and more difficult to master, than those for single factor designs
correlation
a measure of the linear association b/t 2 variables; assesses the extent to which 2 variables covary
correlation coefficient
quantitative index of the direction and magnitude of the relationship
direction
the way in which 2 variables covary
-positive: x increases, y increases; direct relationship
-negative: x increases, y decreases; inverse relationship
magnitude
strength of the relationship, can vary from 0 (no relationship) to plus/minus 1 (perfect relationship, we can correctly predict y from x 100% of the time)
survey sampling
-nonprobability sampling - accidental & purposive samples
-probability sampling - simple random sample & stratified random samples
-margin of error - factors that influence margin of error: how confident do we want to be (z)? what size is our random sample (n)? whats the proportion in the pop (pi)?
-sampling frame: population (define conceptually), sampling frame (define the pop operationally), sample (select group from sampling frame), element (unit of analysis)
the third variable problem
-when deaths in new delhi increase, the # of ice cream cones eaten increase
quasi-experiments
-experiments tha have treatments, outcomes measures, and experimental units, but do not use random assignment to create the comparisons from which treatment - caused change is inferred
-quasi-experiments are used when random assignment to treatments is feasible or ethical
types of validity
-validity of measurement/construct validity - does a measuring instrument adequately represent a theoretical standard?
-statistical conclusion validitiy - are the assumptions underlying a statistical procedure adequately met when the procedure is applied?
-internal validity -if we observe covariation b/t 2 variables, how well can we justify the inference that one variable was a cause of variation in the other?
external validity - wha are the limits to the generalizability of a demonstrated casual relationship? (to other types of participants, other measures of constructs, other research settings
threats to internal validity
classes of extraneous casual variables which, if not controlled or assessed in the quasi-experimental design, might produce effects cofounded with the effects of the treatments
- history - specific events occurring b/t the 1st and 2nd measurement in addition to the treatment of interest
-maturation - processes within respondants operating as a function of the passage of time per se (not specific to particular events), including growing older, hungrier, more tired
-testing - the effects of taking a test upon the scores of a second testing
-instrumentation - changes in the calibration of a measuring instrument or changes in the observers or scorers used
-statistical regression - the tendency for extreme score to converge a pop average
-selection - biases resulting from systematically diff characteristics of comparison groups
-experimental mortality - differential loss of respondents from the comparison groups
-selection-by-maturation interactions - rates of internal change within the comparison groups may differ systematically
Sets with similar terms
Psych 12 Midterm 3
73 terms
RDA Test 3
69 terms
Experimental Psych Exam 2
79 terms
Psych 301W quiz 4 (not mine, but really good study…
44 terms
Other sets by this creator
SPANISH EXAM 2
68 terms
CHILD DEVELOPMENT
18 terms
PERCEPTION EXAM 1
55 terms
RDA exam 1
30 terms