Scribd
Upload
7 views5 pages

Nothing HHHHHHH

The document provides an overview of descriptive measures in statistics, focusing on survival times of mice after a deadly injection. It includes calculations for sample mean, variance, standard deviation, and coefficients of variation, as well as methods for determining median, quantiles, and percentiles. Additionally, it explains how to visualize data distribution using box plots and discusses the implications of skewness in data.

Uploaded by

vfhgsjykz4
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
7 views5 pages

Nothing HHHHHHH

The document provides an overview of descriptive measures in statistics, focusing on survival times of mice after a deadly injection. It includes calculations for sample mean, variance, standard deviation, and coefficients of variation, as well as methods for determining median, quantiles, and percentiles. Additionally, it explains how to visualize data distribution using box plots and discusses the implications of skewness in data.

Uploaded by

vfhgsjykz4
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 5

Course: Stat 102

Instructor: Prof. Ahmad Soltani

Lecture 4:
Descriptive Measures.

Example 1 Continues: A number of 23 mice are subjected to a deadly injection. The


variable of interest is:
x: the survival time, in minutes, for a mouse.

stem Leaf
0 1 4 5 7 7 7 8
1 0 1 4 8
2 2 2 2 6 8
3 1 1 6 6 8
4 0 8

Data (sample):
31, 40, 1, 22, 22, 7, 4, 10, 18, 38
5, 8, 22, 14, 26, 48, 36, 31, 28, 7
36, 11, 7

Ordered Data:
1, 4, 5, 7, 7, 7, 8, 10, 11, 14, 18, 22, 22, 22, 26, 28,
31, 31, 36, 36, 38, 40, 48
sample size: n = 23

P
sum of x : x = 13768, sample mean: x̄ = 20.5217,

1/n
geometric mean: ( ni=1 xi ) = 14.9699
Q

sample variance: s2 = 185.5336, sample standard deviation: s = 13.6211

coefficient of variation: cv = 100 x̄s % = 100 13.6211


20.5217
% = 66.37%,
coefficient of variation is used in the making comparison between variations of two data sets.
Data with larger cv is more varying.
cv has no unit.
Quick demonstration: Assume for two samples from two populations,
Sample 1: x̄ = 30, s = 6, Then cv = 20%
Sample 2: x̄ = 10, s = 3, Then cv = 30%
Therefore there is more variation within sample 2 in comparisons to sample 1. So do not
make just based on the SD values.

standard error: √s = 13.6211


√ = 2.8402
n 23

This is indeed called the standard error for the sample mean.

Median, Quantile, Percentiles

Measures on the distribution of data:


• Median: Q2
• Quantiles Q1 , Q2 , and Q3
• Min, Max and Range: R = max − min
• IQR= Q3 − Q1 : Inter Quantile Range
• Percentiles
Median is the center of the distribution for data. This means that 50% of values, in
size, are below median and 50% of values, in size, are above it.

To find median:
put data in increasing order. This means look at the ordered data. Then
If the sample size n is odd, then the median is the datum at the center of the ordered data.
If the sample size n is even, then the median is the average of the two values at the center
of the ordered data.

In Example 1, n = 23: odd, so the median Q2 = 22.


For the data: 6,2,4,1,10,12,
the median is what?
ordered data: 1,2,4,6,10,12; n = 6: even
4+6
Two center values: 4 and 6. Hence median= 2
= 5.

The First Quantile Q1 : is the 25% distribution cut value.


25% of values are below Q1 , and 75% are above Q1 .

To Find Q1 : Find the median of the values that are to the left side of the median Q2 of
data.
Q2 is not included.

In Example 1, there are 11 values to the left side of Q2 = 22, and the median of these values
is Q1 = 7

The Third Quantile Q3 : is the 75% distribution cut value.


75% of values are below Q3 , and 25% are above Q3 .

To Find Q3 : Find the median of the values that are to the right side of the median Q2
of data; Q2 should not be included.
In Example 1, there are 11 values to the right side of Q2 = 22, and the median of these
values is Q3 = 31
Min is the smallest values in size, the first value in the ordered data.

In Example 1, min x = 1

Max is the largest values in size, the last value in the ordered data.

In Example 1, max x = 48

Range; R = max x − min x.


IQR: Inter Quantile Range; IQR = Q3 − Q1 .

In Example 1, R = 48 − 1 = 47, IQR = 31 − 7 = 24

min x, Q1 , Q2 , Q3 , max x: Five values summary.

measure min x Q1 Q2 Q3 max x IQR R


value 1 7 22 31 48 24 47

Box Plot: The five values Summary are nicely displayed by ”Box Plot”.

Put the five values on the horizontal axis, then build up a rectangle above the horizontal
axis, as follows.

Index 0

0 10 20 30 40 50
The Box-Plot gives a good picture on:
1- How data are spread, from min to max.
2- It shows if the distribution is symmetric, skewed to the right or skewed to the left.
Symmetric: The Box-plot is completely symmetric about Q2 . For symmetric distribution
always Q2 = x̄

Skewed to the right: The distribution has short left tail but long right tail. So Box-Plot is
condensed on left and is wide on right. For skewed to the right distribution, always Q2 > x̄
skewed to the left: The distribution has short right tail but long left tail. So Box-Plot is
condensed on right and is wide on left. For skewed to the right distribution, always Q2 < x̄

Percentile: We note that Q1 is the 25% lower percentile of data, and also 75% upper
percentile.
Also Q2 is the 50% lower percentile, lower or upper.
Also Q3 is the 75% lower and 25% upper percentile.

For any value γ between 0 to 100, one can locate the lower or upper percentile.

To find the γ% lower percentile:


multiply γ by sample size n divide by 100.
count = γn/100.
this count roughly gives the position of the desired percentile in the ordered data.
γ = 95, corresponds to 95% lower or 5% upper percentile.

In Example, count= (95)(23)/100=21.85. So 5% upper percentile is 39.

You might also like