FINAL Flashcards
Factorial Design
2 or more factors in an experimental design in which each level of every independent variable occurs with all levels of the other independent variables.
- looking at more than one variable at a time
- separate groups of people/participants
- looking for overall effects and interactions
Ex: Gender could be a factor with two levels male and female and Diet could be a factor with 3 levels of low, medium, and high protein.
- Statistical Analysis: ANOVA
Interaction
an experiment result that occurs when an independent variable level is differently affected by levels of other independent variables. (The effects of 1 IV are not the same across the levels of another)
Ethology
The study of naturally occurring behavior
Ethogram
are observations made of different categories for the organisms under study, then recording the number of times the organisms engage in each behavior.
Ex: Chloe’s lecture: we tallied the coded behaviors of the animals.
Chi Squared Test for Independence
a statistical test often used to determine whether that data in a contingency table are statistically significant.
Contingency research
is one sort of relational research in which data of 2 variables are compared to see if they have a relationship between them.
Correlational Research
allows the researcher to determine simultaneously the degree and direction of a relationship with a single statistic.
Positive Correlation
one variable increase so does the other.
example the relationship between smoking and lung cancer.
Negative Correlation
one variable increases as the other decreases
example of smoking and lower grades
Counterbalancing
a technique used to vary order of conditions in an experiment, so practice or fatigue are not confounders
- limits order effect
- presentation of conditions: “ADDA”
Small-n design
levels are presented to a small number of participants to have a well controlled setting and are used for special participates such as cancer patients.
Control Group
does not receive the levels of interest of the Independent variable.
Control Condition
provides a baseline against of which some variable can be compared.
ex: If study is measuring the effects of caffeine, the controlled condition would have no caffeine creating the baseline.
Regression to the mean
refers to extreme scores taken from a larger group and retested members will fall near the mean.
Theory of Signal Detection
- our perception in general is controlled by evidence and decision processes.
- measured in ROC
- d’= sensitivity (sensory processes measured)
- beta=criterion (decision processes measured)
Receiver-operating characteristic (ROC)
a plot graphing hits against false alarms
d’
sensitivity process measured
the distance between signal and noise distributions
the probability that something is there
beta
criterion= decision process measured
is the slope of the ROC function at the point of interest
Statistical Prediction Rules
to increase the accuracy of decisions. The rules are based on the predictor variables.
Pavlovian conditioning
does not indicate continuous paring will lead to being classical conditioned conditioned stimulus (CS) will predict the unconditioned stimulus (US)
Operant (instrumental) conditioning
reinforcement or punishment are used to either increase or decrease the probability that a behavior will occur again in the future.
Sniffy pressing the bar to gain reinforcement.
Shaping
by waiting until the animal makes a desired response observer rewards it (with food) which reinforces greater approximations to desired behaviors.
Classical Conditioning
is a basic form of learning, in which stimuli initially incapable of evoking certain responses.
Ex: Watson’s experiment in fear response was conditioned in a boy known as Little Albert. The unconditioned stimulus was the loud, clanging sounds and the unconditioned response was the fear response created by the noise.
Postive Reinforcement
Increases the likelihood of the response that produces it. Ex: giving a child praise for good behavior
Negative reinforcement
when removed, increases the likelihood of the response that removed it.
Ex: Bob does the dishes so his mom will stop nagging him.
Instrumental Conditioning
conditioning of a subject to learn to make a response that leads to a reward or prevents punishment.
Discriminative Stimulus
Signals when a behavior will be followed by a reward.
Ex: a pigeon might be trained to peck a button for food only when there is a red light.
Extinction
when reinforcement is withheld after an organism has learned.
After several attempts of reinforcer failing, the organisms ceases to make the response.
Unconditioned Response (UCR)
A response made to an unconditioned stimulus
Unconditioned Stimulus (UCS)
unconditioned stimulus is one that unconditionally, naturally, and automatically triggers a response. For example, when you smell one of your favorite foods, you may immediately feel very hungry. In this example, the smell of the food is the unconditioned stimulus.
Law of Effect
- the principle that reinforcement of a response leads to the response being made likely to occur in the future.
- The consequences of a behavior determine it’s fate.
Latent Learning
you don’t see it unless you learn it.
Latency- amount of time needed to complete a task.
Electroencephalogram (EEG)
a recording of the electrical activity of the brain that is done by electrodes placed on the scalp.
Microelectrodes
narrow measure of activity in the brain records it is measured by stimulation chemically stimulate cooling: to find if area on brain will be damaged
ERP
a type of brain wave that is measured shortly after a specific evoking stimulus.
fMRI
images used to measure blood flow in the brain, a correlate of neural activity.
False alarm
the incorrect reporting of the presence of a certain signal, triggered by a different signal
- aka: false positive
Ex. cancer marker tripped, but no cancer
hit
the correct detection of a signal that has been presented
- aka: true positive
carry over
the effect ozone condition carries over to the other condition(s)
- eliminated/reduced by counterbalancing
Latin Square
a structured method of randomizing/counterbalancing conditions
- ( n x n ) method in which you may only use half or so of the possible counterbalanced combinations
dependent samples t-test
parametric test measuring the effect of some condition on one group
- test the difference between two means (before and after means)
- before provides a baseline
Ex. classes average weight before and after winter break
independent/between samples t-test
parametric test measuring the effect of a single condition on two different groups
Ex. two groups: males vs. females
IV: the effect of caffeine
DV: reaction time
sensitivity
probability of a ‘hit’, and detecting something when its there
- greater sensitivity = greater possibility of a false positive (false alarm), greater possibility of a true positive (hit)
- decreased sensitivity = greater possibility false negative (missed hit), decreased possibility of a false alarm
Ex. probability of exceeding a certain value on a cancer marker
- a ‘hit’ decreases with a lower percentage
specificity
the extent to which a positive result is indicative of the presence of the condition you are testing for
- introduces the possibility of a ‘false alarm’
within subject experimental design
- single groups
- all participants exposed to each condition (level of IV)
between subjects experimental design
- two or more groups
- each group of participants/subjects are being tested by a different factor/condition simultaneously
naturalistic observation
observe how animals behave in natural environment
ad libitum/narrative sampling
- informal observations
- similar to field notes
continuous focal sampling
- focal animal or small focal group
- record start & stop time of each behavior observed
- provides a complete record of the subject’s behavior
all occurrence sampling
- record every time a behavior happens (tally)
- ideal for events (not states)
- provides frequency data
instantaneous/scan sampling
- focal animal or group
- record behaviors at pre-established time intervals
- provides an estimate of time spent performing a behavior
one-zero sampling
- records whether or not behavior occurs during a previously established time interval
- “0” if no, “1” if yes
- over-estimates rare behaviors
- under-estimates common behaviors
independent variable
variable being manipulated
dependent variable
variable measured and recorded by experimenter
one sample t-test
comparing the mean of a sample to the mean of the population from which the sample was taken
one-way ANOVA
- compares two variables
- IV: categorical
- DV: numerical
two-way ANOVA
- compares two IV’s and one DV
- IV: categorical
- DV: numerical
- multiple hypotheses, as there can be sig. differences between one main effect and the DV and not the other main effect
main effect
- each IV in a two-way anova are called “main effects”
interaction effect
- the interaction of two main effects (IVs)