Documenti di Didattica
Documenti di Professioni
Documenti di Cultura
16 January 2010
What is a z-score?
Computational Formula
x X X = z= S s
Where
or
z=
= any raw score or unit of measurement standard distribution of scores mean and standard distribution of scores deviation deviation of of the the
X , s = mean and
Calculating z scores
What is the z score for your test: raw score = 80; mean = 75, S = 5?
X X z= S
z=
80 75 =1 5
What is the z score of your friends test: raw score = 80; mean = 75, S = 10?
X X z= S
80 75 z= = 0. 5 10
Who do you think did better on their test? Why do you think this?
Calculating z scores
Example: Raw scores are 46, 54, 50, 60, 70. The mean is 60 and a standard deviation of 10.
X 70 60 50 54 46 x 10 0 - 10 -6 - 14 z 1.00 .00 - 1.00 - 0.60 - 1.40
z=
X X S
70 60 10 = =1 10 10
z=
60 60 0 = = .0 10 10 50 60 10 = = 1 10 10 54 60 6 = = 0.6 10 10 46 60 14 = = 1.4 10 10
z=
z= z=
Why zz-scores?
Transforming scores in order to make comparisons, especially when using different scales Gives information about the relative standing of a score in relation to the characteristics of the sample or population Location relative to mean Relative frequency and percentile
Z-score Distribution
Mean of zero
Zero distance from the mean
Standard deviation of 1 Z-score distribution always has same shape as raw score. If distribution was positively skewed to begin with, z scores made from such a distribution would be positively skewed.
-3 20 0 200
-2 30 5 300
-1 40 10 400
0 50 15 500
+1 60 20 600
+2 70 25 700
+3 80 30 800
Transformation Equation
Transformations consist of making the scale larger, so that negative scores are eliminated, and of suing a larger standard deviation, so that decimals are done away with. Transformation scores equation:
standard score= z(new standard deviation) + the new mean
Transformation Equation
standard score= z(new standard deviation) + the new mean
A common form for these transformations is based upon a mean of 50 and a standard deviation of 10. in equation form this becomes:
standard score= z(10) + 50
above
below
Comparing distributions
scores
from
different
Arithmetic 40 36 24 22 6
Arithmetic Average 80 67 73 62 53 48
Standard score =
Arithmetic Average 80 67 73 62 53 48
80 -2 s
90 -1 s
100
110 1s
120 2s
90
80 -2 s
90 -1 s
100
110 1s
120 2s
How many standard deviations below the mean is a score of 90? How many standard deviations above the mean is a score of 120?
1 2
-3
-2 30
-1 40
0 50 50
1 60
2 70 67 70
3 80 80
scores
Student A
20
z =
1. What is the standard deviation of 50? 2. What is the standard deviation of 70?
0 ___ ___ 2
67 50 10
17 = 1 .7 10
=
=
3. What is the standard deviation of 80? ___ 3 4. What is the standard deviation of 67? 1.7 ___
-3
-2 30
-1 40
0 50 50
1 60
2 70 67 70
3 80 80
scores
Student A
20
1.7 2
Student A is at mean in Geography, 2 standard deviation above the mean in Spelling, 3 standard deviation above the mean in Arithmetic and has an average of 67 which is 1.7 standard deviation above the mean.
-3
-2 30
-1 40
0 50 50
1 60
2 70 67 70
3 80 80
scores
Student A Student B
20
0
50 62
1.7 2
73
0 Student C
36 48 53 55
1.2
2.3
-1.4
Random variation conforms to a particular probability distribution known as the normal distribution, which is the most commonly observed probability distribution. Mathematicians de Moivre and Laplace used this distribution in the 1700's
de Moivre
The shape of the normal distribution resembles that of a bell, so it sometimes is referred to as the "bell curve".
Completely Parameters
Described
by
Two
The normal distribution can be completely specified by two parameters: 1.mean 2.standard deviation If the mean and standard deviation are known, then one essentially knows as much as if one had access to every point in the data set.
.3413 of the curve falls between the mean and one standard deviation above the mean, which means that about 34 percent of all the values of a normally distributed variable are between the mean and one standard deviation above it
The normal curve and the area under the curve between units
about 95 percent of the values lie within two standard deviations of the mean, and 99.7 percent of the values lie within three standard deviations
Percentage under the Normal Curve at various standard deviation units from the mean
68.26%
2.15%
13.59%
13.59%
2.15%
-3s
-2s
-1s
+1s
+2s
+3s
In a normal distribution: Approximately 68.26% of scores will fall within one standard deviation of the mean
90%
10%
c10
z = 1.28
c90
z = 1.28
N = 1,000
Points in the normal curve above or below which different percentage of the curve lie
= 80 = 16
110 80 16
30 = 16
= 1.875
X = 80
X = 110
z = 1.875
Equation of unknown
for
different
z=
Similarly, the raw-score equivalent of the point below which 10 percent of the case fall is:
X 80 1.28 = 16
X 80 = 16(1.28) X = 80 + 20.48 X = 100.5
1.28 =
X 80 16
Acknowledgement of References:
N.M Downie and R.W Heath. Basic Statistical Methods, 5th Edition. Harper & Row Publisher, 1983 Robert Niles
http://www.robertniles.com/stats/stdev.shtml
Rosita G. Santos, Phd, et. al. Statistics. Escolar University, 1995. Leslie MacGregor. z Scores & the Normal Curve Model (presentation)