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;
32 Cards in this Set
- Front
- Back
Francis Bacon
|
(1561-1626) inventor of modern science; casual observation is not enough to establish relationships
|
|
comparitive expirement
|
an expirement comparing two groups of people
|
|
randomized expirement
|
expirement using a random device
|
|
expirement
|
active manipulation of a subject and then careful oberservations/recording of the events that result; helpful in establishing causal relationships
|
|
causal relationship
|
establishing that one thing is the cause of something else
|
|
observational study
|
no intervention, just observations of a certain situation; more difficult to establish causal relationships; descriptive studies; suggest relationships but do not provide proof
|
|
descriptive study (purpose)
|
to observe and gather data; suggest relationshiups for future study
|
|
randomness required for what inferences
|
1)inferring characteristics from a sample to a whole population
2)inferring a cause-effectrelationship |
|
random selection
|
also called random sampling; necessary to infer from sample to population; tends to produce smaples that represent the diversity of the population
|
|
random assignment
|
necessary to be able to infer a cause-effect relationship; assign treatments randomly so that extraneous variables will not give any group an advantage/disadvantage
|
|
goals of an expirement
|
1)demonstrate/explain the causes of an observed outcome
2)assess the effect of manipulating a (known) causal variable w/ the intent of a particular effect |
|
how to extablish a causal relationship
|
1)random assignment 2)sound expiremental methodaology 3)logic
|
|
big saying
|
correlation does not imply causation
|
|
requirements (Mill) for extablishing causal relationship btwn X & Y
|
1)X preceeds Y
2)X is related to/associated with Y 3)No plausible alternative explanation for Y, other than X, can be found |
|
prospective study
|
looks forward. gather data; consider a variable that may affect the data; collect follow-up data.
easy to establish time sequence |
|
retrospective study
|
looks backward. notice the effect first; rewinds to find the potential cause of the effect. difficult b/c there are many possible causes for a single event
|
|
Statistics
|
study of variation; how to quantify it, control it, and draw conclusions from it.
|
|
treatments
|
set of circumstances manipulated by the expirementor
|
|
expiremental unit
|
smallest unit to which a treatment is applied at random and a response is obtained
|
|
design of expirement
|
overall plan for conducting the expirement. good design makes it possible to give unambigious answers to questions
|
|
scope of inference
|
what the investigator can ultimately infer from the expirement; determined by sampling procedure
|
|
strategies for designing an expirement
|
1)eliminate sources of variability
2)isolate sources of variabity so that their effect can be later seperated 3)ensure that the remaining sources of variability produce only chancelike variability |
|
confounding variable
|
extraneous variable that makes us unable to distinguish which variable produced the result.
|
|
how to eliminate sources of variability
|
1)direct control-holding the potential source constant
2)blocking or grouping 3)randomization 4)replication 5)control group 6)use a placebo 7)conduct single-blind or double-blind expirements |
|
blocking
|
creating groups that are similar as possible with respect ot one or more variables thought to be sources of variability
|
|
blocks
|
expiremental groups similar in some feature we would like to control
|
|
grouping
|
same as blocking expect the group size need not be equal to the number of treatments
|
|
matched-pairs design
|
experiments are paired based on the value of some vairable thought to be related to the response variable
|
|
randomization
|
random assignment to expiremental groups
|
|
replication
|
using several trials and runs; multiple observations for each expiremental condition
|
|
single-blind experiment
|
subjects don't know what treatment they've received; human mind odesn't affect outcome
|
|
double-blind
|
subjects and those measuring the response don't know what treatment the subject has received
|