PSYCH 210: Exam #1
68 Cards in this Set
Front | Back |
---|---|
A characteristic or condition that varies and has different values in different people or at different times
|
Variable
|
Measuring variables
|
data
|
Scales of Measurement
|
NOIR
|
to classify
-Diagnosis, Genotype, Dead/Alive etc.
|
Nominal (Qualitative)
|
Order-unequal intervals
-social rank, race finish, birth order, intervals can be different
|
Ordinal (Qualitative)
|
Distance, magnitude
-equal intervals, arbitrary "0", temperature (C,F), opinion
|
Interval (Quantitative)
|
Origin
-Natural "0", temperature in K, brain volume, # of convictions, time
|
Ratio (Quantitative)
|
Describes data
|
Descriptive statistics
|
Lets us infer using samples to study populations, uses probability
|
Inferential statistics
|
Help us understand populations
-represents population in a study
-has "n" individuals
|
Sample
|
The entire set/group of individuals you could potentially observe
|
Population
|
characteristics of populations
|
Parameters
|
characteristics of samples
|
Statistics
|
(Hopefully small) discrepancies that almost always occur
-Larger sample should have smaller sampling error from populations values
|
Sampling error
|
Links samples and populations
-replication is important
|
Probability
|
What measurement scale? Inches
|
Ratio
|
What measurement scale? Weather conditions
|
Nominal
|
What measurement scale? Rank order of favorite NCAA teams
|
Ordinal
|
What measurement scale? Weight in pounds
|
Ratio
|
What measurement scale? Religious affiliation
|
Nominal
|
What measurement scale? Distance
|
Interval
|
An equal chance for inclusion in sample
-each set of members is equally likely
-discrepancy of M from Mu is random
|
Random Selection
|
Sample where demographics are in proportion (Good)
|
Stratified random sample
|
Sample of only Intro Psych students (randomly selected)
(Not as good)
|
Convenience sample
|
Least useful sample, be cautious (Bad)
-biased
|
Self-selected sample
|
-Pre-existing patterns
-preexisting groups compared (difference in mean?)
-don't know if causal or not
|
Correlation
|
-Artificially create groups that differ on a variable
-record differences in another variable
-what's the advantage? Causation
|
Experiment
|
Measure naturally occurring associations
-quantify pattern in 2 observed random variables (x,y)
-compares means
|
Nonexperimental, Correlational Studies
|
Finding cause
1) Manipulation of Independent variable
2) Control of confounding variables
3) Observe dependent variable
|
Experimental Study
|
A factor manipulated by the experimenter. The focus of the study.
|
Independent variable
|
Factor that may change in response to an independent variable. Can be multiple
|
Dependent variable
|
3 Preventing confounds practices
|
1) Random assignment - to treatment/control condition
2) Holding constant - same treatment for everyone (e.g. all surgeons right handed)
3) Matching (participants) e.g. anxiety level, R/L handed
|
Internal attributes or characteristics that can't be directly observed but are useful for explaining/describing behavior
|
Constructs
|
Identifies an observable variable theoretically related to a construct
-defines the construct in terms of those measurements
|
Operational Definitions
|
Typically whole, countable numbers with no intermediate values possible
|
Discrete variables
|
Free to choose precision, indefinitely divisible categories
-real limits=-1/2 unit, +1/2 unit
|
Continuous variables
|
Shows the number of individuals (f) in each category (x) on a scale of measurement
-organizes and reveals trends
|
Frequency Distribution
|
"Intervals" of X values
-summarize, condense measurement categories
-lose precise frequencies
-O/I/R scales
|
Grouped Frequency Distributions
|
Unequal intervals that don't touch
-Nominal, Ordinal
|
Bar Graph
|
Quantitative scale, extends to real limits
-Ratio, Interval
|
Histogram
|
Has to start and end at bottom (0)
-Ratio, Interval
-rarely used, except published research
|
Polygon
|
Positive skew: Too hard
Negative skew: Too easy
|
Floor Effect
|
Higher percentiles always equal higher _______
|
scores
|
The sum of the scores divided by the number of scores
|
Mean
|
Distance from the Mean
|
Deviation score
|
The sum of the Deviation scores =
|
0
|
M=20
-add 1 to every score what is the new mean?
|
21
|
(SumX1 + SumX2) / (n1 + n2)
|
Weighted Mean
|
Mean is not appropriate for _____ & _______ scales
|
Nominal, Ordinal
|
Cares more about rank order than number values
-midpoint in ordered list of scores
-value dividing distribution in half
|
Median
|
Most frequent category or score
-always the peak in distribution
-always actual, observed value
-can have multiple
-N,O,I,R
-highest frequency
|
Mode
|
Positive skew: left to right (M,Mdn,Mo)?
|
Mean, Median, Mode
|
Negative skew: left to right (M,Mdn,Mo)?
|
Mode, Median, Mean
|
For a negatively skewed distribution with a Mode of X=25 and a mean of M=20, the Median is probably?
|
Between 20 and 25
|
Describes a central point of the distribution
|
Central Tendency
|
How scattered/spread out the scores are
-used to describe/summarize
-used as a basis for most inferential stats
|
Variability
|
Low variability =
|
Better representation of sample
|
High variability =
|
Unrepresentative sample common
-harder to detect relationships
|
Max distance between scores
Continuous: Xmax - Xmin
Discrete: Xmax - Xmin +1
|
Range
|
Distance covered by middle 50% of the distribution
-useful with vary skewed distribution
|
Interquartile Range
|
The average distance between a score and the mean
|
Standard Deviation
|
Average squared distance from the mean
|
Variance
|
The square root of variance
|
Standard deviation
|
Sum of Squares (SS)
|
SS=Sum(X-M)^2
|
Convey each score's distance from the mean in standard deviation units
-rarely used in applied settings (decimals, negative numbers are confusing)
|
Z-Score
|
Body-percentile rank
Tail-proportion higher than the score
|
Positive Z-Score
|
Body-proportion higher than the score
Tail-percentile rank
|
Negative Z-Score
|
p + q=
|
1
|