PHILIPPINE NORMAL UNIVERSITY

The Indigenous Peoples Education Hub

North Luzon Campus
Alicia, Isabela

PROFESSIONAL EDUCATION

AREA: Assessment and Evaluation of Learning - 2

EXPECTED COMPETENCY: Apply Principles of Evaluation in Classroom Testing and Measurement
Recipient Classes: Third Year BEE, BPHE, BSSE 2nd Term 2017
Lecturer: Dr. Carlito G. Miguel

PART 1. CONTENT UPDATE

BASIC CONCEPTS

A. Definition of Terms:

1. Statistics – the science of the classification and arrangement of facts based on number of
occurrences for the deduction of general assertions
2. Descriptive Statistics – one category of statistics which provide methods that concerns with
summarizing and describing numerical data. For easy understanding, tables, graphs and charts
that display data are used.
3. Data – the collection of any number of relates observations on one or more variables.
4. Data Array – the arrangement of raw data by observations in either ascending or descending
order.
5. Raw or Ungrouped Data – those information that have not been classified, arranged or organized.
6. Sample – it is a collection of some, but not all, of the elements of the population under study in
which the statistics is interested.
7. Tabulation – the process of classifying or grouping of scores in a systematic arrangement.
8. Class Frequency – it is the number of measures or observations in an interval.
9. Class Interval – these are the classes or ranges of values that the observations can assume. Each
class interval has a lower limit and an upper limit.
10. Class Boundaries/Real or Exact Limits – these are the exact values of class limits by at least 0.5. It is
the upper limit of class and the lower limit of succeeding class.
11. Class Mark - it is the midpoint of a class interval in a frequency distribution and taken as the
average of the lower and upper limits.
12. Class Size – it is the width of class intervals and measures the interval between the first value of
one class and the first value of the next class.
13. Graph – are picture of the given numerical data.

14. Frequency Distribution – a tabulation where the test scores are ranked and scores having the
same values are combined. It reports the number of observations or scores that are included in
 each of the class intervals in a table. It permits the researcher to see at a glance how these
measurements are distributed.
B.
Sample Scores
100 106 104 103 105
79 92 91 89 105
110 95 96 108 113
93 102 80 90 101
110 78 107 75 127
99 86 113 98 95
101 103 124 100 87
122 118 104 84 99
109 106 116 111 115
109 88 114 72 102

B.1 Steps in constructing a Frequency Distribution

1.1 Determine the Range

*RANGE = HIGHEST SCORE – LOWEST SCORE + 1
R = 127 – 72 = 55 + 1 = 56
1.2 Determine the number of Categories or Classes by dividing the range with the desired number of
class size.
No of Scores Categories
25 – 30 5–6
31 – 49 7–9
50 – 99 10 – 14
a. I 56/ 12 = 4.66 or 5 b. I 56/ 4.66 = 12.017 or 12

1.3 Compute for the Lower or Higher Limits

a. Lower Class Limits
Divide the lowest score by the interval and multiply by the interval. Remember that if
lower limits is odd it should be equal to the number exactly divisible by the interval. If the lower limits
is even, start with the lowest group of scores with the lowest score.
Example: 72/5
=14.5 = 72
LL + Ci – 1 = LL 70 – 74
b. Higher Class Limits
Divide the highest score by the class interval size.
b.1 Subtract the highest score by the remainder.
Example: b.1 127/5 = 25 remainder 2
b.2 127 – 2 = 125
HL = 125 -129
C. Graphical Method of Statistical Data
1. Frequency Polygon 2. Histogram
2. Cumulative Frequency Distribution
3. Kinds of Cumulative Frequency
a. Cumulative Frequency “less than” (cf<)
b. Cumulative Frequency “greater than” (cf>)
4. Cumulative Percentage Curves or Ogive –
 a. “less than ogive”
b. “greater than ogive”

Sample Frequency Distribution

1 2 3 4 5 6 7 8 9 10 11 12
F(MP-
c.i Tally f D fd Cf< Cf</N MP MP-M d2 fd2
M)
X-X Fd1
d1
125-129 I 1 5 5 50 100 127 26.3 691.7 691.69 26.3
120-124 II 2 4 8 49 98 122 21.3 453.7 907.38 42.6
115-119 III 3 3 9 47 94 117 16.3 265.7 797.07 48.9
110-114 IIII-I 6 2 12 44 88 112 11.3 127.7 766.14 67.8
105-109 IIII-III 8 1 8 38 76 107 6.3 39.69 317.52 50.4
100-104 IIII-IIII 10 0 0 30 60 102 1.3 0.169 16.9 -13
95-99 IIII-I 6 -1 -6 20 40 97 3.7 13.69 82.14 -22.2
90-94 IIII 4 -2 -8 14 28 92 8.7 75.69 304.76 -34.8
85-89 IIII 4 -3 -12 10 20 87 13.7 187.7 750.76 -54.8
80-84 II 2 -4 -8 6 12 82 18.7 349.7 699.38 -37.4
75-79 III 3 -5 -15 4 8 77 23.7 561.7 1685.07 -71.1
70-74 I 1 -6 16 1 2 72 28.7 823.7 823.69 -28.7
TOTAL(X)
50 -13 7840.5 498
= 100.70

D. Measures of Central Tendency – a single measure in a series which represents all the scores made by
a group.
1. Mean – the best known and most reliable measure of central tendency. It is the sum of the
separate measures divided by their number.

Characteristics of Mean
1. An interval statistics
2. A calculated average
3. Value is determined by every case in the distribution
4. Affected by the extreme measures
5. Sum of deviations about the mean is zero
6. Can be subjected to numerous mathematical computations
7. Most widely used
8. Represents average quantities

a. Computation of the mean, the scores are added and their sum divided by the sum of
frequencies/ number of scores

b. For UNGROUPED DATA

X = ∑x/N ∑ = summative notation

x= scores
N = total number of scores
Example:
 X=520+450+300+600+430+380+320/7
X=3000/7
X=428.57

c. For GROUPED DATA

How to compute Mean for Grouped Data?
Steps:
1. Given a frequency distribution, enter in column 2, the midpoint of each interval and
these are designated as (X)
2. Multiply these (X) values by its corresponding ƒ (frequency)
3. Place the product in column 3.
4. Add the value in the column 3 and divide the sum by the number of cases (N) to
obtain the mean.

Formula:
X = ∑ƒ(x)/N where: X = mean
∑ = summative notation
ƒ = class frequency
(x) = class mark
N = sum of frequency
Example:
Method I
X = 5035/50
= 100.70
Method II
X=AM+ fd i
N

X = 102 + ( - 13) 5
50
= 102 + ( - 0.26) 5
= 102 + ( - 1.30)
= 100.70

d. Weighted Arithmetic Mean

Formula:
WX = NX + NX + NX Where: WX = weighted mean
N1+N2+N3 X = mean of the distribution
N = frequency of the distribution
Section Mean N
I 86 32
II 84 45
III 88 40

WX = 32 (86) + 45 (84) + 30 (88)

32+45+40
= 2752 + 3780 + 3520
117
= 10052/7
 WX = 85.91

e. When to use the Mean

1. When we want to use measure of central tendency with the greatest stability.
2. When we want a balance point in a given data.
3. When the scores are distributed symmetrically around a central point.
4. When other statistical methods like standard deviation and coefficient of correlation are to be
computed. These statistics are based upon the mean.
5. When the value of each score is desired
6. When an interval interpretation is appropriate

2. Median – it is the point on the scale such that exactly one – half of the cases in the distribution are above it
and the other half of the cases are below it.

Characteristics of the Median

1. An ordinal statistics
2. A rank or position average
3. Value is determined by scores near the middle
4. Not affected by extreme scores/values
5. If the direction or the signs of the differences of the terms in the distribution from the median is
ignored the sum of these differences is its minimum

When to use the Median?

1. An ordinal position is needed
2. The middle score is desired
3. We want to avoid the influence of extreme values

A. How to compute the Median for UNGROUPED DATA?

Steps:
1. Arrange the scores in the order from lowest to highest, for odd number of scores, the median is the
middle score.
2. If the number of scores/cases is an even number, the median is the point halfway between the two
middle scores.

A.1 in the series

Girls 1 2 3 4 5 6 7
Ages 13 14 15 16 18 19 19

Formula:
X=N+1
2
=7+1
2
= 4th
X = 16

A.2 in the series

Rooms 1 2 3 4 5 6 7 8
No. of Patients 18 21 25 27 30 33 47 56

Formula:
X = N/2 + N + 2/2
2
 = 8/2 + 8 + 2/2
2
= 4+5
2
= 9 = 4.5th
2
X = 28.5 or 28 or 29

B. How to compute the Median for GROUPED DATA?

Steps:
1. Record/compute the cumulative frequencies (cf).
2. Divide N of 50 by 2 (N/2) or taking 50 percent of the 50 cases. Hence we are interested in
finding that point in the distribution with 25 cases above and 25 below it.
3. Find the class interval in which the middle cases falls, and determine the exact limits of this
interval.
4. Interpolate to find the value on the scale above and below where one – half of the total
number of cases falls. This is the median.
Formula:
X = LL + (N/2 – cf <) i where: X = median
fm LL = Lower Limit
cf< = cumulative frequency of the class
= 99.5 + (50/2 – 20) 5immediately preceding the median class
10 fm= frequency of the median class
= 99.5 + (25 – 20) 5
10
= 99.5 + (5) 5
10
= 99.5 + (.50) 5
= 102
3. Mode – it is the score that occurs most frequently in a distribution.

Characteristics of Mode:
1. A nominal statistics
2. An inspection average
3. The most frequently occurring value
4. Usually occurs near the center of the distribution
5. Cannot be manipulated mathematically
6. Some distribution have more than one mode
7. Most popular score

When to use the MODE?

1. A nominal interpretation is needed
2. A quick approximation of central tendency is desired
3. The most frequently occurring score is needed

How to compute the MODE?

a. For UNGROUPED DATA
 The mode is not established through calculation;
 It is determined by looking at a set or a graph of scores and seeing which score occurs most
frequently.

Consider the following data:

Set I 10 10 19 17 10 16 18 15 11 11 Mode is 10 Unimodal
Set II 2 3 4 5 7 36 8 9 Mode: 3 & 4 Bimodal
 Set III 25 22 29 28 24 24 21 25 29 20 Mode: 24,25,29 Trimodal

b. For GROUPED DATA

With data grouped in the form of frequency distribution, the mode is taken as the point of the
class interval with the largest or highest frequency. For example, the mode of your grouped is –

Formula:
X = L mo + (d1) i where: X = mode
(d1 + d2) L mo = lower mode of the assume mode
d1 = frequency of the modal class minus the frequency below it
d2 = frequency of the modal class minus the frequency directly above it.
I = class interval
Example:
X = 100 + (10 – 6) 5
(10 – 6 + 10 – 8)
= 100 + (4) 5 = 100 + (.6) 5
(4+2) = 100 + 3.33
= 100 + (4) 5 X = 103. 33
(6)

Interpretation of Measures of Central Tendency as a Whole

 Given any value (mean, median, or mode) it is the point on the scale around which scores tend to
group.
 It is the average or typical score which represents a given group of pupils/ students.
 Given two or more values of central tendency one can define who performed bad, good, better, or
best.
 The computed value represents the average performance of a given group and becomes the basis in
determining whether one group performed better or poor than the other groups.

E. What is Measures of Variability?

 Measures of variability is a numerical index which indicates the extent to which the scores of a group
tend to scatter (disperse, spread) above and below a central point in the distribution.
 The larger the measures of variability the more spread the scores are and the group is said to be
HETEROGENEOUS.
 Conversely, the smaller the measures of variability the less spread the scores are and the group is said
to be HOMOGENEOUS.
 The three most frequently encountered measures of variability are the RANGE, the STANDARD
DEVIATION, and the QUARTILE DEVIATION.

1. What is RANGE?
 Range is simply the difference between the highest score and the lowest score in a
distribution.
Example:
99,105, 72, 119, 127, 114

Range = HS – LS
= 127 – 72
= 55
*When to use Range?
When one wants a quick and simple approximation of the spread or variability of scores/
values.
Interpretation:
If the range is small the scores are close together whereas if the range is large the scores are more
spread out.

2. Standard Deviation – it is the square root of the mean of the squares of the deviations taken from the
arithmetic : mean.

a. When to use the Standard Deviation?

 When it is desired to have a measure of variability which has the highest reliability.
 When it is desired to calculate the coefficient of correlation and measures of reliability.

b. Computation of Standard Deviation for UNGROUPED DATA

Steps:
1. First compute the MEAN.
X = ∑X
N
2. Compute the deviation of each score from the mean.
d=X–X Where : d = the deviation of each score from the mean
X = a raw score
X = the mean
3. Square each of these deviations and add this column. A check of our work would be necessary that the
sum of (x) column or the sum of the deviation about the mean should be zero.
4. To find the standard deviation, we use the following formula and substitute it as it shown.
*NOTE: in some texts, the formula for the standard deviation is given as:

SD = √ ∑d2 SD = √ ∑x2 O = √(𝑥 − 𝑥)2

N N N–1

Example:
(x) Deviation from Mean (x – x) (d1) Squared Deviation (d2)
40 -30 900
45 25 625
55 15 225
65 5 25
70 0 0
75 5 25
80 10 100
85 15 225
90 20 400
95 25 625
∑x = 700 ∑(x-x)2 = 3150

x = ∑x
N SD = √ ∑d2
= 700 N
10 SD = √ 3150
X = 70 10
= √ 315
SD = 17.75

5. Computation of Standard Deviation without the X.

X X2
 40 (40)2= 1600
45 (45)2= 2025
55 (55)2= 3025
65 (65)2 = 4225
70 (70)2= 4900
75 (75)2 = 5625
80 (80)2 = 6400
85 (85)2 = 7225
90 (90)2 = 8100
95 (95)2 = 9025
∑x = 700 ∑x2 = 52150

SD = √∑𝑥2– (∑x)2
N–N
=√52150 − (700)2
10 10
49000
5215 −
=√ 10

= √5215 − 4900
=√315
= 17.75

c. How to Compute Standard Deviation for GROUPED DATA?

The computation of the standard deviation will be illustrated by using the grouped data used for the
computation of mean and median.

Steps:
1. Find the mean
2. Subtract the mean from the class mark and assign it as d1.
3. Multiply the frequency with d1 and the product will be assigned ƒd1
4. Square ƒd1 and assign it as fd12

Formula:
SD = √∑ƒd2
N = √7840.5
50 = √156.81= 12.52

3. Average Deviation/ Mean Deviation – the mean deviation ( MD) also called arithmetic deviation is the
mean of all separate measures in a series taken from central tendency which is usually the arithmetic mean,
seldom the median.

a. When to use the Mean/ Average Deviation – when one wants to measure the extent by which each
individual value distribution deviates from the mean of the distribution.

b. Computation for Mean/Average Deviation

1. Compute for the mean.
2. Subtract the mean from the score/class mark.
3. For the ungrouped data, add the deviation and divide it by the number of scores.

A. For UNGROUPED DATA

Formula:
Scores d
12 +1.2
15 +4.2
10 -0.8
8 -2.8
9 -1.8
∑d = 10.8
 MD/AD = ∑d/N
= 10.8/5
= 2.16
4. For the Grouped Data, multiply the frequency
with the deviation.
5. Add the product of ʄd and divide it by the N.
ʄ𝑑
6. Find the square root of the quotient√ 𝑁 .
B. For GROUPED DATA
Formula:
MD/AD = ∑fd/N
= 498/50
= 9.16
 Interpretation of Standard Deviation
- The computed value indicates how spreads out scores are. One standard deviation below and
above the mean includes around 68.26% of the cases and measuring of two standard deviations
units on each side of the mean includes between these points approximately 95% of the cases and
three standard deviation units over 99% of the cases is included. It helps in determining how many
students performed above the mean or average, above average or below average.

4. Quartile Deviation – it is one – half of the distance between the 75th percentile and the 25th percentile in a
frequency distribution.
 Q1 (25th Percentile) is the first quarter or quartile on a set of scores, the point lies 25 percent of the
number of cases.
 Q3 (75th Percentile) is the third quarter or quartile on a set of scores, the point below which lies 75% of
the number of cases.

When to use the Quartile Deviation?

- When one wants to measure the dispersion in the middle half of the items arrange in an array.
How to Compute for Quartile Deviation? (Ungrouped data)
1. Arrange the score from highest to lowest.
2. Find the number or cases of the cases/scores
𝑁 2𝑁 3𝑁 2𝑁
3. Apply the formula Q1 = 4 , Q2 = 4 , Q3 = 4 , Q4 = 4 respectively.
Example:
118 138 148 158
120 139 148 159
124 140 149 161
128 142 149 161
129 142 150 165
130 142 150 165
130 143 151 167
133 146 152 168
137 147 153 175
Q1 138 – 10th Score Q2 147 – 20th Score Q3 156 – 30th Score Q4 180 – 40th Score

40 2(40) 3(40) 4(40)

Q1 = = 10th Q2 = = 20th Q3 = = 30th Q4 = = 40th
4 4 4 4

a.) For GROUPED DATA

1. Compute for Q1 by dividing the total frequency by 4.
2. Find the lower limits of Q1.
3. Apply the formula of Q1 for grouped data.
 𝑁
( −𝑐ʄ𝑓<) Where:
4
Q1 =LL + i
ʄ𝑓𝑄𝑖 LL = lower limits of the first quartile class
(12.5−10) cf< = cumulative frequency just below the first quartile
= 89.5 + 5
4 class
(2.5)
= 89.5 + 4 5 fQ1 = the frequency of the quartile class
= 89.5 + 3.125 N = total of frequency
= 92.625 or 92.63
𝑁
( −𝑐ʄ𝑓<)
4
4. Apply the formula for Q3 = LL + i
ʄ𝑓𝑄3
(37.5−30)
Q3 = 104.5 + 5
8
(7.5)
= 104.5 + 5
8
= 104 + (9375)5
= 104 + 4.6875
= 109.1875 or 109.19
5. Compute for the quartile deviation using the formula:
𝑸𝟑−𝑸𝟏
Q =
𝟐
109.19−92.63
=
2
16.57
=
2
= 8.28

6. Stanines/Deciles – these are values which divide the distribution into ten parts.

a. How to compute for the deciles

1. First, like in the computation of quartiles, divide the total number of frequencies by 10.
2. If it is grouped find the lower limit of the deciles as indicated by its point.
3. Compute the desired decile using the formula for the ungrouped data:
𝑁 2𝑁 3𝑁
d1 = , d2 = , d3 = ...
10 10 10
4. For the grouped data, the formula is:
𝑁
( −𝑐ʄ<)
10
D = LL+ i
ʄ𝑑1

Where: LL – lower limit of decile class

N – total frequency
cf< - cumulative frequency just below the decile class
` fd1 – the frequency of decile class
i – class interval

Example:
50
(6( )−20)
10
D6 = 99.5+ 5
10
300
( −20)
10
= 99.5+ 5
10
30−20)
= 99.5 + 5
10
(10)
= 99.5 + 5
10
 ` = 99.5 + (1)5
= 99.5 +5
= 104.5
6. Percentile
The percentiles are the points that divide the total number of scores into exactly one hundred equal
parts. It is understood that there are ninety nine (99) percentile that determine the points below which certain
percentage of the test scores would fall. For example, the seventh percentile (P7) would indicate within or
below the test scores in the distribution lies within or below it while 72% lies above it.
The percentiles are computed exactly in the same manner as the median.

STEPS IN FINDING THE PERCENTILE FROM THE RAW SCORES ARE AS FOLLOWS:
1. Arrange the scores from highest to lowest or from lowest to highest.
2. Determine Pk where Pk is the kth percentile and k is equal to 1, 2, 3, … 99.
3. Computation of the percentile is very much similar to the computation of the quartile and decile.
4. Consider the following formulas for ungrouped and grouped data:

Ungrouped Data Grouped Data

𝑁
P= (10 (
𝑁
)− 𝑐ʄ𝑓<)
100 100
P10 = LL + 5
ʄ𝑓𝑃10
Example:
Example:
50 50
P = 100 (10(
100
)−4)
P10 = 79.5 + 5
2
500
= .5 ( −4)
100
= 79.5 + 5
2
(5−4)
= 79.5 + 5
2
(1)
= 79.5 + 5
(2)
= 79.5 + (.5)5
= 79.5 + 2.5
Interpretation of Quartile, Deciles, and Percentile = 82
 In a normal distribution, taking the median and add and subtract 1 quartile deviation on each of it
includes approximately 50% of the cases. In the same manner if quartile deviations are measures off
on each side of the median in a large sample practically all cases will be included. As a result 8 quartile
deviations cover the range, as do also 6 standard deviations, when is large.
 The quartiles, Q1, Q2, or Md, Q3, and Q4 are the points which divide the total number of cases in a
frequency distribution into four equal parts or four equal number of cases. They are points of division
when dividing a group in four subgroups. Q1 and Q3 are also used in computing the quartile deviation
from the median.
 The deciles D1, D2, D3, D4, etc. are the points which divided the number of cases in frequency
distribution into ten equal parts, or ten equal number of cases.
 The percentiles P1,P2,P3,P4, etc. are the points which divide the total number of cases in a frequency
distribution into one hundred equal parts. They indicate the percentage of pupils surpassed by a pupil,
or the number of pupils that surpassed him. They can be used as grades of pupils and are comparable
to grades in percentages, in letters, or in numbers. The computation of the median.

F. Measures of Divergence from Normality

Skewness
 Distribution which departs from normality by having more scores piling at one end of the scale
than the other are said to be skewed.
 A distribution is negatively skewed if the mean is less than the median and if the scores pile at the
upper - score end.
 A distribution is positively skewed if the mean is greater than the median and if the scores pile at
the lower - score end.
  Negative skewness represents a bright class; positive skewness represents a dull class.
How to compute for SKEWNESS
1. Compute the percentile range for P10, P50 and P90
2. Compute the skewness using the formula:
3 𝑀−2𝑀𝑑𝑛)
𝑷𝟗𝟎+𝑷𝟏𝟎 = 99.085 – 102 3. S2 =
S1 = – P50 𝑆𝐷
𝟐 = -2.915 or - 2.92 =
116.17+82 12.52
= - 102
2
198.17
= – 102
2

Mo Md M
Positively skewed distribution.
Example:
Table 1. Computation of the sum of Cubes of deviations about Arithmetic mean in a symmetrical
distribution.

SCORE (X) (X-Ẍ) (X-X)3

25 4.5 91.125
25 3.5 42.875
23 2.5 15.625
22 1.5 3.375
21 0.5 0.125
20 -0.5 -0.125
19 -1.5 -3.375
18 -2.5 -15.625
17 -3.5 -42.875
16 -4.5 -91.125
205 0 0

Table 2. Computation of the sum of squared deviation and sum of cubes of deviation times frequency above
the arithmetic mean in a positively skewed distribution.

Class ʄ X (X-Ẍ) (X-Ẍ)2 ʄ(X-Ẍ) 2 (X-Ẍ)3 ʄ(X-Ẍ)3

93.5-96.5 1 95 23.35 642.6225 642.6225 16290.48037 16290.48037
90.5-93.5 1 92 22.35 499.5225 499.5225 11164.32787 1164.32787
87.5-90.5 1 98 19.35 374.4225 374.4225 7245.075375 7245.075375
84.5-87.5 1 86 16.35 267.3225 267.3225 4370.722875 4370.722875
81.5-84.5 1 83 13.35 178.2225 178.2225 2379.270375 2379.270375
78.5-81.5 2 80 10.35 107.1225 214.245 1108.717875 2217.43575
75.5-78.5 2 77 7.35 54.0225 108.045 397.065375 794.13075
72.5-75.5 3 74 4.35 18.9225 56.7675 82.312875 246.938625
69.5-72.5 4 71 1.35 1.8225 7.29 2.460375 9.8415
66.5-69.5 6 68 -1.65 2.7225 16.335 -4.492125 -26.95275
63.5-66.5 7 65 -4.65 21.6225 151.3575 -100.544625 -703.815625
 60.5-63.5 5 62 -7.65 58.5225 292.6125 -447.697125 -2238.485625
57.5-60.5 4 59 -10.65 113.4225 453.69 -1207.949625 -4831.7985
54.5- 57.5 2 56 -13.65 186.3225 372.645 -2543.302425 -5086.60425
Total
40 3635.1 31830.56999
X=69.6

∑f(X-X)3 above the X = 44,718.22349

∑f(X-X)3 below the X = 12.887.65350

(𝑋−Ẍ)3
∑ʄ Wherein: (X) – midpoint of the class limit
𝑁
Skewness =
∑ʄ
(𝑋−Ẍ)2
√∑ʄ(𝑋−Ẍ)2 f- frequency
𝑁 𝑁
X - arithmetic mean
This is the computation of skewness in table 2. N – number of cases

(𝑋−Ẍ)3
∑ʄ
𝑁
Skewness = 795.7642492
(𝑋−Ẍ)2
∑ʄ √∑ʄ(𝑋−Ẍ)2 = Given:
𝑁 𝑁 90.8775(9.53297) ∑fX-X)3/N = 31830.56999
795.7642498
31830.56999
= ∑f(X-X)2 = 3635.1
866.3324812 N = 40
=
40 = 0.92
3635.1 3635.1
√
40 40
A distribution can be positively or negatively skewed.

Normal

Negatively Skewed
Positively Skewed

Figure 1. Skewed and Normal Distribution

2. Kurtosis
 The peakness or flatness of a curve in a distribution as compared with the normal.
 The kurtosis of a normal curve is .263. A distribution that is more than peaked than the normal is
leptokurtic, flatter than normal is platykurtic, and the normal is mesokurtic.
a. Compute the quartile deviation of the scores/ intervals
b. Find the percentile ranges of P10 and P90
c. Compute for Kurtosis using the formula:

𝑸.𝑫
K=
P90 – P10
Where: Q.D = Quartile Deviation of the score/class interval
P90, P10 = Percentiles range of the scores/ class interval
Example:
8.28
K=
116.17−82
 8.28
=
34.17
= .242317
= .24
G. Standard Scores
These are linearly divided scores that represent the area under the normal curve.
1. How to compute for STANDARD SCORES?
a. Find the difference between the individual’s raw score and the mean of the normative group.
b. Divide the difference by the standard deviation of the normative group.
Formula:
𝑋−Ẍ
Z= 𝑆𝐷
Where: X- any raw score
SD – standard deviation of the group
Ẍ - mean of the group
Z- standard score

Example I.
Which score is better in an intelligence test: a score of 145 on Test I or a score of 60 on Test II?

Test I Test II
X= 100 X = 40
SD = 15 SD = 5

145−100 60−40
Z= Z=
15 5
45 20
= =
15 5
=3 =4

Example II. Which subject has the higher score?

Test I – ENGLISH Test II – GEN. SCIENCE

X = 141 X = 112
X= 148 X = 113
SD = 3.2 SD = 4.3

141−148 112−113
Z= Z=
3.2 4.3
−7 −1
= =
3.2 4.3
= -2.19 = 0.23
The test in GEN. SCIENCE has a higher score.

H. The Normal Curve

The data from a test administration are graphed in the form of a curve; the shape of the curve will
depend on a way in which the scores are distributed. The normal curve has the following characteristics:
Symmetrical – mean, mode, and median are the same
Bell shaped
Are under the shaped presents 100%.It is possible to calculate any area under the curve.
The score above the mean of distributed the same way as the score below the mean.
On the base line of the curve, a mark is placed in the center to represent the center of distribution.
Three marks are placed equidistant from one another above the center and three below the mean.
They represent 3 standard deviations both above and below the mean.
 The normal curve represents frequency distribution for a population.
In a normal curve most of the scores are concentrated in the middle of the distribution, with a few
scores at each end.

OTHERS FACTS:
Distribution that contains at least 30 scores will approximate the normal curve.
An important use of the normal curve for physical educators is in determination of percentile ranks.
When distributions are not normally distributed the curve is skewed.
Positive – tail to the right
Negative – tail to the left

AREAS UNDER THE CURVE:

Total are is 1.00 and 0.50 is the area at the left of the mean and 0.50 is the area at the right of the
mean.
68% of the total area is under the curve between a s of -1 and a s of +1.
95% of the total area under the curve between a s of -2 and a s of +2.
99% of the total area is under the curve between as of -3 and s of +3.

.13% 2.14% 13.59% 34.13% 34.13% 13.59% 2.14% .13%

30 -20 -10 X +10 +20 +30

Example: Problems involving the Normal Curve

o What percentages of the general population have deviation IQs lower than 85?
𝑋−Ẍ
Z = 𝑆𝐷
To solve this problem, we have to convert the 85 IQ to a z – score. Since, where X is a score, X is the
mean, and 15 is the Standard Deviation. So our problem the Z – score associated with an IQ of 85 is
𝑋−Ẍ 85−100
Z= = = −1.00
𝑆𝐷 15

So an IQ of 85 is equivalent to a z- score of -1 and our question becomes what proportion of the normal
curve is lower than the z-score of -1. Appendix A only shows the proportion of scores between a z-score and
the mean (we can ignore the negative sign as the table only uses positive values). So we must look at the
proportion of scores between a z-score of -1 and the mean and then subtract this value from 50 (50%). So 50 -
.3413 (Appendix A Value) =.1587 or 15.87% of the population has IQ scores lower than 85.
We can also solve this problem by using the normal curve figure we have seen in this lesson. Just add up
the proportion of the curve below -1 standard deviation, which is 13.59 + 2.14 + 0.13 = 15.86%.

SCORE AREA
0.0 0.0000
0.5 0.1915
0.86 0.3051
1 0.3413
1.52 0.4357
2.5 0.4988
3.99 0.5000
Z = 1.00 Z = 0.5
0.3413 0.1915
 0 1 2 0 1 2

Z = 1.52
Z = 2.5

0 1 2 3 -4 -3 -2 -1 0 1 2 3 4

Comparison of Standard Score Deviation

Relationship between percentile score, z –scores, T – scores and stanine scores in a normal distribution.

Percent of cases
under portion
34.13% 34.14%
the normal curve
13.59% 13.59%

0.13% 2.14% 2.14% 0.13%

-40 -30 -20 -10 0 10 20 30 40

Standard Deviations

Percentile Scores

0.1 2 16 50 84 98 99.9

z-scores

-4.0 -3.0 -2.0 -1.0 0 +1 +2 +3 +4

T- scores

20 30 40 50 60 70 80
Stanines

1 2 3 4 5 6 7 8 9