data analysis week 2

measurement levels

micro

meso

macro

data types

categorical

quantitative

types of measurement

noir

nominal

orginal

interval

ratio

what types of measurements under categorical

nominal - no equal distance - dummy variables -

ordinal- ranking but no eual distance between variables

mean

effected by extreme values

median

relatively unaffected by outliers

mode

more than one mode - bimodal,multimodal

what ct to use in nominal data

mode

ct for ordinal data

median

ct for interval and ratio and what it depends on

mean and median but it depends on the skewness

no skewed mean

is skwed median

sknewness

ralked about the shape of the bell shape symmetry

when skewness=0 symmetric when not eual to 0 it is sknewed

values outise -1,1 are SUBSTANCIALLY SKEWED

when mean>median - positively skewed right side

when median>mean - negatively sknew left side

when median > mean

common for variables to take a value concenrtaed around larger values

kurtosis

degree to which scores cluster at the tails

mesokurtosis

kurtosis =3 chilled

leptokurtosis

k>3

heavy tails

pointy

platykurtosis

k<3

flat

lighter tails

range

measures variability

upper range - lower range

affected by outliers

inter quartile rang e

q1 p25

q2 p50

q3 p75

Q3-Q1

7th decile

better than 70 percent

90th percentile

better tahn 90 percent

variance and sd

squared avergae of points that are at a distance from the mean ]

sd- underoot of the variance - homogenity

68 95 99.7

TO CREATE FRQUENCY DISt table

control shift and enter at the end