Scribd
Upload
41 views

Introduction and Descriptive Statistics - Vinjar Fønnebø PDF

The document provides an introduction to statistics and statistical testing for beginners. It uses the example of a girl named Mary who is shorter than her peers to illustrate how to calculate measures of central tendency, measures of spread, the normal curve, z-scores, and how this relates to determining if something is statistically significant or not. The key aspects are determining the mean, standard deviation, and seeing how many standard deviations something is from the mean.

Uploaded by

Gabriel
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
41 views

Introduction and Descriptive Statistics - Vinjar Fønnebø PDF

The document provides an introduction to statistics and statistical testing for beginners. It uses the example of a girl named Mary who is shorter than her peers to illustrate how to calculate measures of central tendency, measures of spread, the normal curve, z-scores, and how this relates to determining if something is statistically significant or not. The key aspects are determining the mean, standard deviation, and seeing how many standard deviations something is from the mean.

Uploaded by

Gabriel
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/ 37

Statistics for beginners

!  It is fun
!  You can understand it
!  You use it intuitively every day
A DAY AT THE OFFICE

Meet Mary. She


is 12 years old.

!  Mary is very short


!  What makes you say that she is short?
Mary’s peers

!  You think Mary should be like


everybody else.
Mary’s peers

!   You compare Mary with everybody else


!   Is Mary’s height and ”everybody else’s”
height all you know in order to make a
conclusion?
Mary’s peers

!  NO, you also have an idea in your mind that


Mary’s height is farther away from
everybody else than random variation
Mary’s peers

!  You make an estimate that Mary is


more different than random
variation should account for
Formula:
Mary’s height
Formula:
Mary’s height - Everybody else’s height
Formula:
Mary’s height – Everybody else’s height
Random variation
Formula:
Mary’s height - Everybody else’s height
Random variation

!  A high figure leads you to think that


Mary can not be like everybody else
!  You start looking for a specific «reason»
for her low height, maybe a medical
condition.
Girls with Turner’s syndrome

!   Mary is not like everybody else. but she is


like everybody else if you only compare
with other girls with Turner’s syndrome
Girls with Turner’s syndrome

!   Mary is not like everybody else, but she is


like everybody else if you only compare
with other girls with Turner’s syndrome

Mary’s height- Everybody else’s height


for girls with Turner’s syndrome

Random variation
for girls with Turner’s syndrome
Important!!!

!   THIS IS BASICALLY ALL THERE IS TO


STATISTICAL TESTING
!   REMEMBER, you do it every day just by
«eye-balling»
Formula:

Mary’s height – Everybody else’s height

Random variation

!   This measurement carries few difficulties


Formula:

135 cm - Everybody else’s height

Random variation
Formula:

135 cm - Everybody else’s height

Random variation

!  What is everybody else’s height?


Formula:
Measures of centrality:
!   Mode
!  The value occurring most often
!   Median
!  The middle value when all values are ranked
!   Mean
!  Sum of all values divided by number of values
!  Σx/N = µ"
Formula:
Measures of centrality:
!  If Mary’s peers are a statistical
”population” of 10 twelve-year olds :

10
Formula:

135 cm - 154 cm

Random variation
Formula:

135 cm - 154 cm

Random variation

!  What is random variation?


Formula:
147 Measures of spread:
151 !  Difference between highest and
151
152
lowest value: 162 cm - 147 cm =
15 cm
153
154
155 Τhis is called RANGE
156
159
162 !   Utilizes only a small part of the
available information"
Formula:
Measures of spread:
!  Mean distance from the mean"

Σ(x − µ)
N

€
What is the value of this mathematical expression?
Formula:
Measures of spread:
!  Mean distance from the mean:

Σ(x − µ) 0
= =0
N 10

€
Formula:
Measures of spread:
!  Mean distance from the mean:"

Is it possible to ignore the sign?


Formula:
Measures of spread:
!  Absolute mean distance from the
mean:"

Σ | (x − µ) | 32
= = 3,2
N 10

€
Formula:
Measures of spread:
!  Mean distance from the mean:"

Is there another way of getting rid of


the sign?
Formula:
Measures of spread:
!  Mean squared distance from the
mean:"

Σ(x − µ) 2 166
= = 16.6
N 10

This is called:
€ VARIANCE
Formula:
Measures of spread:

!   Variance is in this case expressed in cm2


!   This is not very practical when the mean
height is expressed as cm
Formula:
Measures of spread:
!  Square root of the mean squared
distance from the mean:"
Σ(x − µ) 2
= 16.6
N
Σ(x − µ) 2
= 4.1
N
€
This is called:

€ STANDARD DEVIATION
Formula:
Measures of spread:
Σ(x − µ) 2
N

The STANDARD DEVIATION


€
carries the symbol σ."

This is the measure of spread in a population


Formula:

135 cm - 154 cm

SO WHAT??
4.1 cm
- 4.63

!   Mary’s height is four and a half times further


away from the mean than the standard
deviation (random variation)
!   This value is called z
Normal curve:
Histogram of the height of Mary’s peers

Most of the girls fall in the middle, a few further out.


Normal curve:
Histogram of the height of Mary’s peers
-2 σ" µ" +2 σ"

Most of the girls fall in the middle, a few further out.


Normal curve:

68.3%

95.4%

99.7%

-3 σ" µ" +3 σ"


Normal curve:

The exact probability of


68.3%
any number of standard deviations
from the mean can be calculated
and are given in statistical tables or in your
computer’s statistical packages.

95.4%

99.7%

-3 σ" µ" +3 σ"


Normal curve:
Histogram of the height of all Mary’s peers
and Mary µ"
-2 σ" +2 σ"

Most of the girls fall in the middle, a few further out.


Mary is way out there: The statistics agree with our
previous «eye-balling»
Was that all?
Yes, the principles
of statistical testing
are
not more than this.

You might also like