Statistics Interval level of measurement

- a branch of mathmatics that involves - specifies that the distances between

collecting, organizing, summarizing each interval on the scale are
and presenting and presenting data equivalent along the scale from low
- a discipline concerned with the interval to high interval
analysis of data and decision making
based upon data Ratio level of measurement
- can be used to spot trends or isolate - the observations. in addition to
causes having equal iintervals can have a
- based upon a solid edifice of value of zero as well, the divisions
mathematical theorems proven between the points on the scale
through unassailable laws of logic have an equivalent distance
between them
Descriptive statistics
- branch that deals with the Arithmetic mean
description of data collected - set of measurements is the sum of a
sample of measurements divided by
Inferential statistics their number of data points
- deals with examining the
relationships between variables Median
within a sample then making - a set of data arranged according to
generalizations about how those size is the value of the middle data
variables will relate to a larger point
population
Mode
Population - simply the value in a data set that
- if a measurement is gathered for occurs with the highest frequency
every experimental unit in the entire and more than once
collection the resulting data set Range
constitutes the _ of interest - the difference between the highest
and lowest observations in a set of
Sample data
- any smaller subset of measurements
Standard deviation
Nominal data or categorical data - the square root of the variance
- numbers in the variable are used
only to classify the data. use of Variance
words letters numeric symbols - mean of the standard deviations
from the mean
Ordinal level of measurement Standard score/ z score
- depicts some ordered relationship - indicates how many standard
among the variable's observations deviations an element is from a
mean
Normal distributions variance and standard deviation
- "Bell Curve", shape of the - the tendency for individual
distribution when graphed, The observations to deviate from the
mean, mode, and median turn out to mean
be the same score in symmetrical
distributions coefficient variation
- used when comparing two or more
Properties of a normal distribution sets of variables specially when the
- the mean median mode are all equal units are different, results are
- the curve is symmetric at the center percentages
- exactly half of the values are to the
left of center and exactly half the Quartiles
values are to the right - divide a rank ordered data set into
- the total are under the bell curve is 1 four equal parts

Frequency histogram box and whisker plots

- quantitative data, to see distribution - help see the distribution in the data
and the spread of data
Relative frequency histogram
- quantitative data, to see distribution, another way to look at distribution
good for comparing data - histogram and stem and leaf
Stem and leaf display
- quantitative data, to see distribution, outlier
easier to make by hand than - an observation that lies an abnormal
histogram distance from other values in a
random sample from a population
Time-series graph
- for graphing a variable that changes a five-number summary statistic can be
over time and is measured at regular used to determine if there is an outlier in a
intervals data set
bar graph
- qualitative data or quantitative data, linear regression
for displaying frequency or - attempts to model the relationship
percentages between two variables by fitting a
pareto chart linear equation to observed data
- frequencies of rare events in
descending order sampling frame
pie graph - list of individuals from which a
- mutually exclusive categories sample is actually selected / the part
qualitative or quantitative of the population which you want to
draw a sample
variance and standard deviation
- measures of how spread out a
distribution is
sampling error (fact of life error) Skewness
- the population mean will be different - Measure of the symmetry of a
from your sample mean/sample distribution
percentage kurtosis
- measure of steepness
non sampling error (error you want to
avoid) Q1 corresponds to P25
- using a bad list, poor sample design, Q2 corresponds to P50
bias in data collection Q3 corresponds to P75
Q2 median value of the set
simulation
- numerical facimilie or representation Percentiles
of real world phenomenon - The values that divide a
rank-ordered set of elements into
simple random sampling 100 equal parts
- a simple random sample of n
measurements from a population is linear correlation
a subset of the population selected - relation between two variables that
in such a manner that every sample shows up on a scatter diagram as
of size n from the population has an the dots roughly following a straight
equal chance of being selected line

stratified sampling
- the list is divided into groups or
strata

systematic sampling
- can be done with or without a list

cluster sampling
- geography

convinience sampling
- using results that are conviniently or
readilty obtained fill out survey

convinience and multi stage sampling

- used when resoueces are low used
in govt studies

Coefficient of Variation (CV)

- shows how much the data varies
compared to the mean