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;
45 Cards in this Set
- Front
- Back
- 3rd side (hint)
sampling procedures:
|
-random
-stratified:randomly sample from groups of interest -cluster: sample group rather than individuals -convenience: sample p's that are readily available -quota: representative proportions |
|
|
statistics vs. parameters
|
stat-concrete
para-abstract |
|
|
statistics
|
a set of mathematical procedures for organizing, summarizing, and interpreting information
|
|
|
parameters
|
a value that describes the population
ex. the average household income in the us (all in us) |
|
|
descriptive vs. inferential statistics
|
descriptive - describing amt
inferential - describing an effect |
|
|
variables vs. constructs
|
..
|
|
|
descriptive statistics
|
-used to simplify data
-describing amount ex. avg. household income just saying what you see, not explain anything |
|
|
inferential statistics
|
-used to make generalizations from sample to population
-describing an effect -p value only provided for inferential |
|
|
variables
|
a characteristic or condition that changes or has different values for different individuals
-something that varies -flipside of constant -descriptive ---# of crimes committed -categorical ---age ranges ---race ---gender ---recieving drug or placebo ---exposure to violent media or no media -construct (usually psy construct) ---age ---acceptance of gender roles ---aggression level |
|
|
constructs
|
elements that the researcher claims to be measuring and manipulating
|
construct validity: the degree to which the study measures and manipulates the underlying psychological elements that the researcher claims to be measuring and manipulating
|
|
operational definitions
|
-how you measure it
-the measure |
|
|
categorical vs. continuous variables
|
..
|
|
|
categorical variables
|
-separate categories only
ex. who is still alive in the sample -all statistical analyses involve knowing the difference b/w categorical and continuous -no values in b/w -your in or not, male or female -#'s designate outcome or category -used in advanved correlational studies |
|
|
continuous variables
|
-based on a range of values
ex. hormone levels in the sample -can have whole # or something b/w #'s ex. temperatures |
|
|
categorical vs. discrete variables
|
|
|
|
discrete variables
|
-separate categories or whole numbers
ex. # of ppl. in room, counting things up -no "in betweens" |
|
|
real vs. apparent limits of numbers
|
real limit = .5 below and .5 above
apparent limit = as it appears |
|
|
real limits of numbers
|
the boundaries of scores that are represented on a continuous number line
-the real limit separating two adjacent scores is located halfway between the scores -upper limit at the top of the interval -lower limit at the bottom of the interval ex. 1 = .5-1.5 |
|
|
apparent limits of numbers
|
ex. 3, 4, 5, 6, 7
- apparent limit = 3-7, not 2.5-7.5 (real limit) |
|
|
statistical notation
|
-raw score = x
-set of raw scores = xi ---a number next to x regers to the order in which it appears in the data set ---so, xi is the first score in the set -sum of raw scores = Σxi -score multiplied by itself = x^2 -number of people in the population = N -number of people in the sample = n |
|
|
order of mathematical operations
|
PEMDΣAS
|
|
|
mode
|
-the most frequent score
-can be more than 1 -1,2,2,3,3,3,3,4,4,5,5,5 m=3 -pro: uses a real number and picks up on existence of multiple groups -cons: may not really be in the middle |
|
|
median
|
- the middle score
-50% are above it and 50% are below it -determine N, cf, and c% then find the score that contains 50 as the c% -be able to hand calculate |
-requires ordering the scores from the lowest to the highest
-if an odd # of scores, find the score the is exactly in the middle -if an even # of scores, find the midpoint b/w the 2 middle scores |
|
mean
|
-the average score
-Σxi/n -n=number of scores -be able to hand calculate |
-pro: accounts for clusters of scores
-con: affected by outliers |
|
when do you use mode?
|
-when you have a categorical variable
---uses a nominal scale (0=male 1=female) ---discrete variables (ex. # of computers in a home) |
advantages:
-most common score actually exsists (ex. can't have 2.3 computers - want to see the most freq. # of comp. in home) -helps describe the shape of the distribution -useful when multiple groupings exsist (ex. bimodal distributions) |
|
when do you use median?
|
-when weird scores exist
---extreme scores -> skewed dist. ---throws off the mean, not the median |
advantages:
-can calculate when undetermined scores exist ---ex. participant took too long to finish task so had to stop -can calculate when open-ended categories exist ---ex. # of computers:1,2,3,4 or more -more appropriate for ordinal scale (rank) |
|
when do you use mean?
|
-typical measure of central tendency
|
advantages:
-accounts for all scores in dist. -related to measure of variability (distribution) -useful for stats involving hypothesis testing (aka inferential statistics) |
|
central tendency
|
a single sore that defines the center of a distribution (i.e. the most representative score)
|
|
|
shape of distribution
|
bell-curved or 2-humped
|
|
|
create a two-group histrogram
|
-draw normal curves for each group
-interpreting mean differences |
|
|
simple vs. grouped frequency distributions
|
-simple-number of ppl with each score
-grouped-number of ppl in a particular subset ex. A's, B's, C's, D's, F's on exam |
|
|
Σxi
|
sum of scores
|
|
|
obtaining Σxi from a distribution
|
add up all f's
|
|
|
obtaining the mean from a distribution
|
-number that appears the most - biggest frequency #
create a frequency dist. table, find the largest value(s) in the f column |
|
|
proportions and percentages in distribution
|
determine N by adding up the frequencies, divide f by N to get the proportion (p), then multiply p by 100 to get the percentage (%)
|
|
|
real limits
|
-the boundries of scores that are represented on a continuous number line
-the real limit separating 2 adjacent scores is located halfway b/w the scores -upper limit at the top of the interval -lower limit at the bottom of the interval -3,4,5,6,7 =2.5-7.5 not 3-7 |
|
|
frequency distribution
|
an organized tabulation of the number of individules located in each category on the scale of measurement
|
|
|
frequency distribution graph (histogram)
|
-possible values on x-axis
-number of times calue appears on y-axis |
|
|
how can graphs be misused?
|
toying with y-axis
-making it appear big deal or no big deal |
|
|
shapes of frequency distribution
|
graph is skewed
-bell-curved -2-humped |
|
|
positively skewed
|
most scores grouped at bottome but some are extremely high _/^\_____
|
|
|
negatively skewed
|
most scores grouped at top, but some scores are extremely low ___/^^^^\__
|
|
|
percentiles and percentile ranks
|
-def
-finding them in frequency dist. -interpolating a percentile using a formula |
|
|
problem with inferential stats
|
-there will always be a difference between the sample statistic and the population parameter
-difference: sampling error (can be big or small; if big -> can't report as real finding) -most often seen in political polls ex. approval rating has a sampling error of +/- 3 percentage points (sampling error) 45% approval rating -> statistic act. b/w 48-42% -> parameter statistical dead heat -> sampling range overlap eachother |
|
|
interpolation
|
Int=(Dsco*Dper)/D%
add this to the lower SCORE, and you will have the desired percentile |
|