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;
61 Cards in this Set
- Front
- Back
statistic
|
methods of describing and interpreting quantitative data; collect, organize, summarize, analyze, and draw conclusions from data
|
|
descriptive statistics
|
organizing, summarizing, presenting quantitative data; describe the group that was assessed
|
|
inferential statistics
|
inferences about a larger group of indiv. based on data collected from a smaller group or sample
|
|
variable
|
general characteristics measured
|
|
data
|
qualitative or quantitative observations made in envt answers q's: how much, how fast, how long, how often, where, what kind (values on a given set of variables, data set)
|
|
quantitative variable
|
continuous data, numerical (age, height weight, heart rate)
|
|
qualitative variable
|
categorical data, non-numerical (gender, race, ethnic background)
|
|
nominal
(sex, race, hair/eye color) |
name only
primitive categorical and qulitative cannot +, -, x, / no absolute zero or equal intervals |
|
ordinal
(place finishes in a race, grades A,B,C,D, F) |
relative position of objects
no equal intervals nor abs. zero magnitude (a>b, b>c, then a>c) cannot +, -, x, / |
|
interval
(temperature, ACT scores) |
equal intervals betwn numbers
arbitrary zero (not absolute) magnitude +, -, not x or / |
|
ratio
(heart rate) |
absolute zero
ratios are meaningful +, - , x, / |
|
random sampling
|
every subj/obj has equal/indep. chance of being selected
|
|
systematic sampling
|
every 'kth' subg/obj selected
|
|
stratified sampling
|
subj/obj divided into groups/strata accord to charact. important to study, random sample drawn form each group
|
|
cluster sampling
|
clusters of subj/obj randomly selected from intact groups, subj, or obj
|
|
convenience sampling
|
selection based on convenience
|
|
ways to misrepresent data
|
1. suspect samples
very large or very small 2. ambiguous averages several types = different results 3. changing subj/reference |
|
ways to misrepresent data
|
4. detached stats
works 4X faster, 40% markup 5. implied connections may reduce, could work for you 6. misleading graphs |
|
misleading stats
|
7. faulty survey questions
leading q's, forced response, not having not applicable option |
|
frequency distributions
|
1. raw data = original form
2. frequency = number of times value occurs 3. classes = grouping of raw data into clusters of similar values |
|
frequency distribution
|
organizing raw data in table form using classes and frequencies
|
|
categorical frequency distrib
|
data can be places in specific categories
(nominal or ordinal level data) |
|
grouped freq. distrib
|
interval or ratio data grouped into similar classes
(clusters of similar values) |
|
ungrouped freq. distrib
|
class interval on one unit
(small data set) |
|
class limits
|
boundary betwn two class intervals, includes lower and upper (similar to rounding rules)
|
|
lower class limit
|
lowest value data can take to be consid. a member of the Class
|
|
upper class limits
|
uppermost value data can take to be consid. a member of the Group
|
|
class width
|
subtract the lower (or upper) class limit from the lower (or upper) class limit of the Next Class
(10.5-5.5=5CW) |
|
cumulative frequency
|
total numb. raw scores that fall at or below the value of interest
(freq.+freq, starting from bottom,-up- until reach total # of classes) |
|
percent
|
frequency divided by the total number of observations (x100%)
(freq. / total classes(teams)x100) |
|
cumulative percent
|
cum. freq / total number of observations
(%+% all the way from bottom to top, totaling 100% at top) |
|
relative frequency
|
proportion of freq's relative to sample size (freq / total number of classes (teams) NOT x100)
|
|
RULES for freq dist
|
1. between 5 and 20 classes
2. mutually exclusive (no overlap) 3. continuous (no skip #'s) 4. exhaustive (all data w/in classes) 5. equal in width 6. lowest scores at bottom of table |
|
histogram
(bar graph) |
vertical bars of various heights representing freq's
CLASSES = X axis 9horizontal) FREQ's = Y axis (vertical) NO gaps btwn interval and ratio gaps btwn nominal and ordinal |
|
frequency polygon
(line graph) |
lines connect pts plotted for freq's at Midpoint of classes
freq's represent. by heights of points dots connected by straight lines line drawn back to x axis at beg. and end (L, H) |
|
ogive
(line graph) |
represents Cumulative Freq for classes
aka cum. freq. polygon open ended (classes on Y axis) |
|
relative frequency graphs
|
proportion of freq's relative to sample size (like a percent)
(classes on x axis, rel freq on y axis) polygon, ogive, and histogram |
|
misleading graphs
|
watch the reference points!
|
|
statistic
|
data values from SAMPLE
|
|
parameter
|
data values from a POPULATION
|
|
outlier
|
extreme value in data set,
single observation in data set |
|
central tendency
|
measure of average, or middle of data
typical response inc. mean, median, and mode |
|
mean
|
arithmetic average, affect by outliers
add up all observ's x-bar for sample u for population E for summation |
|
median
|
half-way point in data set
symbolized by MD, arrange data in order, select middle pt. not affected by outliers good measure in Ordinal data |
|
mode
|
most freq. occurring observation
best measure for Nominal data can have zero, one or more than one (bimodal) |
|
distributions
|
layout of observ's
pattern of raw data can be symmetric, + or - skewed |
|
symmetric distrib
|
data values evenly distrib on both sides of mean
mean, median and mode all at cntr (IQ scores, heights for adult males) |
|
pos. skewed distrib
|
maj of data fall to L of mean, cluster at lower end of distrib (tail points to higher values (pos))
cent tend values from low to hgh: mode, median, mean (household income in US) |
|
neg. skewed distrib
|
maj of data fall to R of mean and cluster at higher end of distrib (tail points to lower values (neg))
cent tend values from low to hgh: mean, median, mode (4th grd spel test to univ studnts) |
|
standard deviation POPULATION
|
square root of variance
(/N) |
|
standard deviation SAMPLE
|
not as much variance as pop
(n-1) |
|
range
|
symbol R=highest value - lowest value
affected by outliers doesn't take cent tend into consideration |
|
variation
|
how spread out are observ's
range, variance, stand deviation |
|
empirical rule
|
dist. is symmetrical (bell curve)
68% data values fall in 1 stnd. dev of the mean 95% data values fall in 2 99.7% data values fall in 3 |
|
parameter
|
characteristic of Population
given greek letters |
|
statistic
|
characteristic of a sample
given latin letters |
|
central tendency
|
typical score
|
|
mean
|
INTERVAL and RATIO
|
|
median
|
ORDINAL
|
|
mode
|
NOMINAL
|
|
most affected by outliers
|
central tend = MEAN
variability = RANGE |