the standard normal distribution fr. chris a.p.statistics st. francis high school

21
The Standard Normal Distribution Fr. Chris A.P.Statistics St. Francis High School

Upload: amie-garrison

Post on 16-Dec-2015

214 views

Category:

Documents


1 download

TRANSCRIPT

Page 1: The Standard Normal Distribution Fr. Chris A.P.Statistics St. Francis High School

The Standard Normal Distribution

Fr. Chris A.P.Statistics

St. Francis High School

Page 2: The Standard Normal Distribution Fr. Chris A.P.Statistics St. Francis High School

Requirements for Any Probability Distribution

• Always Positive

• Total area under the curve must be 1(Since 1 represents 100%)

Page 3: The Standard Normal Distribution Fr. Chris A.P.Statistics St. Francis High School

The Standard NormalDistribution

1

2πex2

Any value of x will produce a

POSITIVE result

Page 4: The Standard Normal Distribution Fr. Chris A.P.Statistics St. Francis High School

But is the area below equal 1?

-3 -2 -1 1 2 3

0.1

0.2

0.3

0.4

„- x 2ÄÄÄÄÄÄ2

ÄÄÄÄÄÄÄÄÄÄÄÄÄè !!!!!!!2 p

Page 5: The Standard Normal Distribution Fr. Chris A.P.Statistics St. Francis High School

In other words, does

∫−∞

∞1

2πex2 dx=1? YES!

Page 6: The Standard Normal Distribution Fr. Chris A.P.Statistics St. Francis High School

Lets work it out…

∫−∞

∞1

2πex2 dx=∫

−∞

∞12π

e−x2

2 dx

=12π∫

−∞

e−x2

2 dx

Page 7: The Standard Normal Distribution Fr. Chris A.P.Statistics St. Francis High School

But there is a problem…

∫−∞

e−x2

2 dx Has NO antiderivative!

But Karl Friedrich Gauss figured a tricky way around this!

Page 8: The Standard Normal Distribution Fr. Chris A.P.Statistics St. Francis High School

Let us introduce an “I” such that:

I = e−x2

2 dx∫−∞

So the area under the “Standard Normal Curve”

would beI2π

Page 9: The Standard Normal Distribution Fr. Chris A.P.Statistics St. Francis High School

But let’s concentrate on the I

I = e−x2

2 dx∫−∞

∞Since x is merely avariable of integration, we can also express I as

∫−∞

e−y2

2 dyI =

Page 10: The Standard Normal Distribution Fr. Chris A.P.Statistics St. Francis High School

So why not express I2 as

e−y2

2 dye−x2

2 dxI 2 = ∫−∞

∫−∞

or

−∞

∫−∞

∫e−x22 e

−y2

2 dxdy

Page 11: The Standard Normal Distribution Fr. Chris A.P.Statistics St. Francis High School

−∞

∫−∞

∫e−x22 e

−y2

2 dxdy=

−∞

∫−∞

∫e− x2+y2( )2 dxdy

This double integral is actually the volume under a 3-D “bell”

-20

2

-2

0

2

0

0.25

0.5

0.75

1

0

0.25

0.5

0.75

1

Page 12: The Standard Normal Distribution Fr. Chris A.P.Statistics St. Francis High School

Rectangles aren’t the only thing that integrates!

We can now do a clever change of variableby converting to Polar coordinates!

x + y = r

x

2

ry

2 2

−∞

∫−∞

∫e− x2+y2( )2 dxdy=

e− r2( )

2 rdrdθ∫∫π

0 −∞

Page 13: The Standard Normal Distribution Fr. Chris A.P.Statistics St. Francis High School

How polar coordinates work,and how to make the switch...

Any integral can be computed by the limit of Riemann sums over Cartesian rectangles or Riemann sums over polar rectangles. The area of a Cartesian rectangle of sides dx and dy is dx*dy but the area of a polar rectangle of sides dr and dt is NOT just dr*dt

Page 14: The Standard Normal Distribution Fr. Chris A.P.Statistics St. Francis High School

Rather dA is a bit more...

Asector =12dθ r2

dA=12b

2dθ−12a

2dθ

=12 (a+b)(a−b)dθ

Since 12 (a+b) =r and (a−b)=dr

dA=rdrdθGetting back to our story...

Page 15: The Standard Normal Distribution Fr. Chris A.P.Statistics St. Francis High School

Recall the Pythagorean Theorem?

We can now do a clever change of variableby converting to Polar coordinates!

x + y = r

x

2

ry

2 2

−∞

∫−∞

∫e− x2+y2( )2 dxdy=

e− r2( )

2 rdrdθ∫∫π

0 −∞

Page 16: The Standard Normal Distribution Fr. Chris A.P.Statistics St. Francis High School

So now we DO have an antiderivative!

e− r2( )

2 rdrdθ=∫∫π

0 −∞

∫π

0

−e− r2( )

2

−∞

Page 17: The Standard Normal Distribution Fr. Chris A.P.Statistics St. Francis High School

∫π

0

−e− r2( )

2

−∞

dθ= ∫π

0

e− r2( )

2

−∞

And because of symmetry

=2∫π

0

e− r2( )

2

0

dθ=2 1

e02−

1e

∞2dθ∫

π

0

We can now start evaluating the integral from negative to positive infinity…

Page 18: The Standard Normal Distribution Fr. Chris A.P.Statistics St. Francis High School

Almost there!

2 1

e02−

1e

∞2dθ∫

π

0

=2 11

−limb→ ∞

1

eb2

dθ∫π

0

1−0 dθ dθ∫π

0

=2 =2∫π

0

=2π

Page 19: The Standard Normal Distribution Fr. Chris A.P.Statistics St. Francis High School

But that wasn’t “I”

I 2 =2πSo

I = 2π

...that was “I” squared!

Page 20: The Standard Normal Distribution Fr. Chris A.P.Statistics St. Francis High School

Recall the area under the Standard Normal Curve

I2π

So… =2π2π

=1

Page 21: The Standard Normal Distribution Fr. Chris A.P.Statistics St. Francis High School

So the area under the curve is 1!

Wasn’t Professor Gauss clever?

It is no accident that many Mathematicians still prefer to call

the “Standard Normal” distribution

The Gaussian Distribution