Observational units and variables Porportion
Descriptive
·
statistics methods for
summarizing
and
organizing
data
Average of a
binary variable
-
Observational unts (cases) -
specific entities about what information is sample poportion (*) -
If observations are coded as either 1 or O
recorded & is
variable -
any characteristic of an observational unit
categorical-places units into a
group
x
M=
...
n
quantitative -
numerical values
Population purportion (t)
-
represents all observational units in a
population
=...
with population data , the mean is a parameter and the
symbol
observational unit -
the universities 1T
categorical variables name state, type , with sample data, the porportion is statistic and the
symbel is
region a
-
quantitative variables -
enrollment , tuition in , tuition out E
Distributions
Exploratory measures
Exploraty data analysis (EDA)
-
way to make sense of van data describes how the values of a variable are spread across different
and possible outcomes
by identifying patterns detecting anomalies
umvariable analysis -
examination of a
single variable Pattern of a variable over
repeated observations
central values indicate of data describes how variable behaves
can
general tendency a
Mean and median Histogram
Mean (average)
-
Sum of all observational units divided
by number Represents distribution of a
quantitative variable
of observations Groups data into intervals and shows how
many observations fall into
each interval
=...
1
n =
sample size =
sample marker
median-middle value when value is
arranged from smallest to
largest
Ifhis odd , middle value
Ifhis even take of the I middle values
,
average
Mean is sensitive to extremes
Median is more robust to extremes
Interpreting Histograms S number
summary
modes number of peaks histogram pth percentile value of P percent of
in a
quantitative variable which
greater than
- -
a is
umimodal , bimodal , multimodal the data
skewed
right (positively skewed) -
lower values are more common oth-minimum
skewed left (negatively skened) -
higher values are more common 15th (G1) -
1st
quartile
goth-median
approx
.
Symmetric 7th (Q3) 3rd
quartile
-
unimodal
mean median
100th-maximum
Range =
100th -
oth
Interquartile (IQR) 75th-19th
=
range
skewedrigthe s : [30 , 39 5 59 833
number 75 5
summary
.
,
.
, ,
mean median
Range =
S3
IQR =
36
skewed
right
bimodal
s number [S2 60 5 62 5 65 5 713
summary
.
,
.
, ,
.
,
median
Range
mean >
=
19
IQR = S
Boxplots
Based on S number summaries
Impact of unusual observations A box is drawn between the quartiles
means sensitive to lines extended from the box
are
outlying values are to the maximum and minimum
Median is marked with a line
outhers are also marked individu
upper hinge Q3t1 S IQR
-
.
symmetric skew
right shew left Lower Lunge Q, -1 5
.
-
IQR
mean emphasized for distributions while median
is
symmetric is
Sf ·
emphasized for skewed distributions as a measure of center
AA- 8 N
or
"typical" value
Standard deviation
Deviation -
how far an observed unit is from its mean
standard deviation (o) -
approximate average distance of the
observational units from their common mean value
number or
som of squared deviations
of cases- -
mean = -3
3122(720
24(630377440023725(8203
=
6 403
.
standard deviation for outcome
a
binary
=N
standardized sears (2-score)
observationexputed
· is
z =
Indicates how standard deviations the value is from the mean
many
July 1 30
.
December =7583 -1
.
. 1617
the month of July was a little more than 1 3 SDs above
.
average
If a 2 score = 0
,
measurement = M
If a I score
> 0 , measurement > M
If a z score <0 , measurement < M
Magnitude of I
a secre is the distance from the mean in units of standard
deviations
Zanarbor =
-1 .
1617
Isanfrancisco =
-1 9954
.
Because 11seK 121A) ,
St had more extreme temperature in
December