24 Descriptive Stats Flashcards
way of summarizing/organizing and interpreting data
statistics
number from math treatment of data
statistic
ex. of descriptive stats?
interrelationships between values, typical values, how many cases fall in particular category
nominal and ordinal data are ___
discrete
gender of patients undergoing gall bladder surgery at hospital and tabulating frequency distribution is ex. of:
nominal data
taste of new food vs. conventional, count number of cases of each ordered rank of scale and then tabulate as frequency distr is example of ____
ordinal data
y axis also called ____ is used to plot ____
ordinate; frequency
x axis is also called ____ and used to indicate ____
abscissa; categories
in interval/ration, can have increasingly accurate values depending on ___ of msmt
sensitivity
useful when continuous data involves large # varied measurements
grouped frequency distr
how to do grped frequency dist?
organize data into ordered array of score frequencies, find range of scores, decide on width, note limits of each class interval, count frequency of scores in each class interval
how to find range?
highest-lowest + 1 (1 is added to include real limits of continuous data)
how to find width?
range / groups
max ___ groups
9 (>9 is diff. to inspect)
if too few groups, what is prob?
meaning of varied data lost
how is histogram diff from bar graph?
bars touch to show continuity of data
how is frequency polygon diff from histogram?
plot line thru midpoint of each class interval
why use frequency polygon?
allow interpolate values between points
in positive skew, mean is ___ than the median
larger
after summarize data in frequency distr, often compare _____ of diff categories
relative frequencies
why find relative frequencies?
useful to understand trends in data on any scale
number resulting from manipulation of raw data
statistic
relative frequency of one set of frequencies to another
ratios
frequency of one category relative to total sample/pop
proprtions
number to make statistic more understandable
base
how to calc. incidence rates?
new cases over time = (new cases/total pop at risk) x base
how calc. prevalence rates?
total cases over time = (total cases/total pop at risk) x base
measures of _____ are most typical or representative scores in a distr
central tendency
measures of ______ are extent scores spread out
dispersion/variability
for nominal data this is appropriate measure of central tendency
mode (most frequently occurring score)
can be used for ordinal/interval/ratio scale data, score that divides distr in half
median
x bar is used for mean for ____, mu is used for ___
sample; pop
gen. most appropriate for interval or ratio data but not suitable for nominal data since no average of discrete categories:
mean
for skewed continuous data, _____ may be more suitable than ___
median; mean
what are 3 stats for dispersal?
range, variance, standard deviation
easy to calculate; diff between highest and lowest scores–>distorted by outliers, based on extreme and not typical
range
average deviation about the mean would be convenient measure of variability if solve prob by squaring deviations, called:
sum of squares (however, overstates spread)
sum of squared deviations divided by # cases is called:
variance
for variance for sample, use (n-1) instead of (n)…why?
otherwise too small estimate as lost a degree of freedom calculating x bar
what is sq root of variance?
standard deviation
what is standard error of mean?
when calc SD from means of several samples of pop. (SD / sq rt of n)
