Z - Scores: Why Is This Impor Tant?

Z - Scores
Why is this Impor tant?
How do you compare apples and oranges?

Are you as good a student of French as you are in Physics?
How many people did better than you on a test?
How many did worse?
Are you abnormal or deviant?

Dr So and So, are you going to curve the exam?
If you are going to understand a situation, you will need to able

compare things. You will need to present the comparisons in a
way that facilitates decision-making. You may be asked:
s What percentage of people falls below a given score?

s What is the relative standing of a score in one distribution versus
another?
s What score or scores can be used to define an extreme or deviant
situation?
You can do all of these with the Z - Score!
Data Types
For our discussion of the Z-scores, we will be using measurement data (numerical
scores).
We Will Need:
Measures of central tendency:
Measures of variability
The Mean
Standard Deviation
Our First Intuitive Example

You enter your dorm room and find that one of your roommates has a happy
face:
But your other roommate is sad:
What's happened! Why are you so happy! But why

are You so sad?
They reply:
Hurray, I got a
60 on the
Physics test!!!
Wah!, I got a
60 on the
French test!!!
They have the same score! Why the different

reactions!
As a student, you probably have figured out that the 58 on the French test is
bad compared to every one else and that the 58 on the Physics test may be
pretty good compared to everyone else.
Our physics major roommate, says "That great old Prof. - she curved the
test!".
We need to be more precise - To do that we need to review the Gaussian
Distribution!
That's how we will compare the French and Physics tests.
Finding Yourself on the Gaussian Curve

Remember the Gaussian curve. It is the shape that quite a lot of
measurement data distributions have. That means that the
frequency of the scores tend to fall in a very specific shape.
Standard Deviation
Mean
The Gaussian curve is described by its mean (average) and standard deviation.
Hints:
This is the Gaussian curve and the line in the middle is the mean.
Mean -
X/N or the average.
Many frequency distributions in Psychology are normal.

A little box on the graph represent a score.
In our case it will be an
exam score of a particular individual.
The red arrow represents the standard deviation.
A measure of the width of a measurement data distribution
[(X-mean)2/N]
SO WHAT DOES THIS HAVE TO DO WITH THE FRENCH

AND PHYSICS TEST!
THE BIG IDEA!!

WHERE YOU ARE IN THE DISTRIBUTION DETERMINES
HOW WELL YOU DID!
60 French Mean
70
50 60
Physics
Mean
Because the Physics test was very difficult with a mean of

50, one roommate did better than most (score of 70) in the
class as was quite pleased with herself. You can see that
most people in the distribution are below her.
It is the opposite for our French-taking roommate. She is
scored below most.
THE 2ND BIG IDEA!!

We can quantify your place in the distribution. You dont have
to draw a picture every time.
We will do this with the Z-Score.
THE 3RD BIG IDEA!!

The Standard Deviation is a Ruler.
Consider the Z - score as the Inch Marks of the Ruler.
A Z-Score Equals the Number of Standard Deviations, the

score is from the Mean
Z
Deevviiaattiioonn
Meeaann))// SSttaannddaarrdd D
Z == ((SSccoorree -- M
THE CONCEPT
Rulers
Your School Bag
Normal Curve
12 Inches Long
100% of the Scores
Marked in Inches
Marked in Z-Scores
Center is 6 "
Center is the Mean
L
Teesstt!!
Phhyyssiiccss T
Looookk aatt tthhee P
Leettss L
We know its mean is 50.

The Standard Deviation is calculated to be 10 (someone used their
calculator).
Remember the SD is the distance from the Mean from to the Point
of Inflection (change in direction of the curve)
10 Points =
1 SD
2 SDs = 20 points
Mean = 50
Roommate score = 60
THE 4TH BIG IDEA!!

Whats our Physics Freak Roommates Z - score
And Why is She Happy!
1. The roommates score is 10 points above the mean or one SD above

the mean.
2. Her Z - Score Equals 1 because:
Z
D
Meeaann))//SSD
Z
Z == ((6600 -- 5500)) // 1100
Z
Z == 11
3. Lets do it again! The student down the hall runs in to our gangs
room. She yells - I got a 84!! Yahoo!
2. Her Z - Score Equals 3.4 because:
Z
D
Meeaann))//SSD
Z
Z == ((8844 -- 5500)) // 1100
Z
Z == 33..44
This is a bright young person. Lets look at the Z-Score ruler on the
Gaussian Curve. She gets a very big happy face
Shes done better than almost

everyone in the class!
Oh, Dear - Now we know we our French Taking Room

mate is Sad!
Z
Z == ((6600 -- 7700)) // 1100
Z
Z == --11..00
Her Z - score in negative and below the mean.
N
Nuuaanncceess A
Abboouutt tthhee ZZ--SSccoorree::
1. The Z - score can be positive or negative.
Positive is above the mean.
Negative is below the mean.
2. The Mean of the Z scores is Zero.
If you got the class average on a test - your Z -score = 0.
Because Z would be (Mean-Mean)/0 = 0.
3. The Standard Deviation of the Z distribution = 1. We will discuss this
later.
A REALLY BIG CONCEPT!
Z -SCORES CAN BE TURNED INTO

PERCENTAGES
Heres the deal - if you know a Z-score - you know what
percent or proportion of scores are above and below that score.
Its like slicing a pie.
Cut it in quarters and you know that each quarter as 25% of the pie.
Lets slice the Gaussian Curve
The distribution holds 100% of the scores or a proportion of 1.0

50% below
The gray line cuts it in half. So - half are below and half above.
If you received the average score on the test (Z = 0.0), you did better
than 50% of the class.
What about cutting the distribution in another place?
Calculus types have calculated the percents above and below any ZScore.
We mentioned this is Intro to Stats #1.
You statistics book has a table with this information in it.
SSoom
mpplleess::
Exxaam
mee E
Our Physics Students:
1. The score of 60 was 1 Standard Deviation (Z= 1.0) above the
mean.
The table would tell me that there are 84.13% of scores
below.
2. The score of 84 was 3.4 SDs above the mean (Z = 3.4).
The table tells me that there were 99.97% scores below it.
No wonder these kids are happy.
99.97%
below me
>84.13%
Our Poor French Student:

The score of 60 was1 SD below the mean (Z = - 1.0)
The table would tell me that there were 84.13% scores above it
You can do this with any score if you have a

Gaussian distribution.
More Nuances:
1. There are many variants of this procedure:
A. Deviation Problem Whats in Your Tails? - Huh?
Let's go back to the Physics Test.
Your are the professor and the chair of the department
wants to know what scores cut off the top 5% of the students
and the lowest 5 %.
The former get an award and the latter get a scolding!
Heres the Answer

I looked up the Z - score that cuts off 5 % in the tails. The
score is 1.64 as you can see in most Z tables.
Thus, on my Z-score ruler, the test scores would have to be
plus or minus 1.64 Z units from the mean. One Z unit (our
inch, so to speak) equals 10 points.
So I get the picture below and my two cut-offs of 33.6 (low)
and 66.4 (high)
2. Cutting of percentages in the tail of a distribution will develop the

idea of significance testing:
3. You can find the raw score that corresponds to a percentage by:
a. Looking up the Z score that cuts off the percent you want
b. Use the following formula:
X = Mean + Z (SD) - For example:
4. The Z-Score is crucial to understanding:

a. t- test
b. Correlation Coefficient (Pearsons r)
5.. If the distribution isnt Gaussian thats for more advanced
discussions.
Bottom Line:
If you are have a set of measurement scores on different measures using
Z-scores you can tell how the scores are placed in their distributions. Then
you can compare them. Also, knowing the mean and standard deviation of
a distribution that is reasonably Gaussian allows you to determine the
percentile rank of that score.
So curving a test - means that a professor marks you on a z-score basis but what if the mean is high? Huh? Think about it.

Z - Scores: Why Is This Impor Tant?

Uploaded by

Copyright:

Available Formats

Z - Scores: Why Is This Impor Tant?

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Z - Scores: Why Is This Impor Tant?

Uploaded by

Copyright:

Available Formats

Z - Scores

Why is this Impor tant?

How do you compare apples and oranges?

Are you abnormal or deviant?

If you are going to understand a situation, you will need to able

s What percentage of people falls below a given score?

You can do all of these with the Z - Score!

Our First Intuitive Example

But your other roommate is sad:

What's happened! Why are you so happy! But why

They have the same score! Why the different

Finding Yourself on the Gaussian Curve

X/N or the average.

Many frequency distributions in Psychology are normal.

SO WHAT DOES THIS HAVE TO DO WITH THE FRENCH

THE BIG IDEA!!

Because the Physics test was very difficult with a mean of

THE 2ND BIG IDEA!!

We will do this with the Z-Score.

THE 3RD BIG IDEA!!

Consider the Z - score as the Inch Marks of the Ruler.

A Z-Score Equals the Number of Standard Deviations, the

Your School Bag

100% of the Scores

Center is the Mean

We know its mean is 50.

THE 4TH BIG IDEA!!

1. The roommates score is 10 points above the mean or one SD above

Shes done better than almost

Oh, Dear - Now we know we our French Taking Room

A REALLY BIG CONCEPT!

Z -SCORES CAN BE TURNED INTO

Its like slicing a pie.

Lets slice the Gaussian Curve

The distribution holds 100% of the scores or a proportion of 1.0

What about cutting the distribution in another place?

Our Poor French Student:

You can do this with any score if you have a

Heres the Answer

2. Cutting of percentages in the tail of a distribution will develop the

4. The Z-Score is crucial to understanding:

You might also like