You are on page 1of 67

Part 1 of Lecture in

Advanced
Educational Statistics

Functions/Uses of Statistics
Helps in providing a better understanding and
exact description of a phenomenon of nature.
Helps in proper and efficient planning of a
statistical inquiry in any field of study.
Helps in collecting and classifying an
appropriate set of quantitative data.
Helps in presenting complex data in a suitable
tabular, diagrammatic and graphic form for an
easy and clear comprehension of data.

Helps in understanding the nature and


pattern of variability of a phenomenon
through quantitative observations.
Helps in drawing valid inference, along
with a measure of their reliability about
the population parameters from the
sample data.
Helps in understanding statistical
techniques for the purpose of making
informed decisions that affect our lives
and well-being.

Statistics

is a branch of
mathematical science that deals with the
scientific collection, classification,
organization, description, analysis and
interpretation of data obtained from
surveys and experiments. It also deals
with prediction and forecasting based on
data.

Two Types of Statistics


1. Descriptive Statistics - used to describe
the basic features of the data in a study. It is
used to present quantitative descriptions in
a manageable form. It provides summaries
about the sample and measures and form
the basis of virtually every quantitative
analysis of data. This type includes the
Measures of Central Tendency and
Measures of Dispersion.

2. Inferential Statistics used to make


inferences or generalizations on a
population based upon a sample and
used to test hypothesis and evaluate
estimates. It is used to make judgments
or conclusions on the probability that an
observed difference between/among
groups is a dependable one or one that
might have happened by chance in a
study.

Differences between Descriptive and


Inferential Statistics:
With descriptive statistics, one is simply
describing what is or what the data shows while
with inferential statistics one tries to reach
conclusions that extend beyond the immediate
data.
Descriptive statistics uses graphical
numerical descriptions to give a picture of a data
set while inferential statistics uses mathematical
probabilities to make generalizations about a
large group based on data collected from a small
sample of that group.

Population portion of the


universe that can be reached by a
researcher. It includes all individuals
with certain specified characteristics
in a target community/locale or
setting. It is the group to which the
researcher would like the results of a
study to be generalizable.

Sample the group or


portion of a population on
which information is
obtained, preferably selected
in a such a way that it
represents the population.

Data are facts,


observations, and
information that come
from investigations

Types of Data:
1. Measurement data sometimes
called quantitative data the result of
using some instrument to measure
something (e.g. test score, weight).
2. Categorical data also referred to
as nominal or qualitative data
provide categories for sorting or
classifying objects or events on the
basis of some quality like major field
(Sociology, Psychology, etc.)

3. Ordinal Data - provide a system for ranking


observations from most to least or least to most.
It includes the position of individuals finishing
a race (first, second, third, etc.); social class
position (upper, middle, lower) and any data
involving scales that organize observations in
terms of categories (ex. Very favorable,
favorable, neutral, unfavorable, very
unfavorable)
* The rule for assigning numerals to ordinal data
categories is based on ordering observations in
a descending or ascending order.

4. Interval data - provide categories which


consist of equal intervals which indicate
that the distance of each interval is known.
Example of categories which show interval
data are: ruler (the distance between 1 inch
and 2 inches is exactly the same as the
distance between 7 and 8 inches), age,
number of years of formal education, weekly
earnings, etc.

Variable property of an object


or event that can take on
different values. For example,
college major is a variable that
takes on values like
mathematics, computer
science, English, psychology,
etc.

Types of Variable:
1. Continuous a variable that consists of an
infinite continuum of points and provides
exact measures of the amount of a
characteristic present, in theory, any value
between the lowest and highest points on the
measurement scale.
2. Discrete provides counts of the number of
observations appearing in a finite set of
categories (e.g. gender, (male/female), college
class (freshmen/sophomore/junior/senior),
etc.)

3. Independent also referred to as


experimental or predictor variable that is
manipulated, measured or selected by the
researcher as an antecedent condition to an
observed behavior. In a hypothesized causeand-effect relationship, the independent
variable is the cause and the dependent
variable is the outcome or effect.
4. Dependent also referred to as criterion
variable that is not under the experimenters
control the data. It is the variable that is
observed and measured in response to the
independent variable.

5. Control Variables which are


not measured in a particular study
must be held constant,
neutralized/balanced or eliminated
so that these will not have a biased
effect on the other variables

6.Extraneous - factors in the research


environment which may have an effect
on the dependent variable/s but which
are not controlled. These variables may
damage a studys validity making it
impossible to know whether the effects
were caused by the independent
variable. If these can not be controlled,
these must be taken into consideration
when interpreting results.

7. Intervening processes that are not directly observable which like extraneous variables
can alter the results of research. In Language teaching for example, these are usually
inside the subjects heads including various language learning processes which the
researcher cannot observe. If teaching technique is the independent variable and mastery
of objectives is the dependent variable, then the language learning processes used by the
subjects are the intervening variables.

Intervening Variables include


motivation, fatigue, boredom,
and any other factor that arises
during the course of the research
either on the part of the
researcher or the respondents

8. Moderator affect the relationship between


the independent and dependent variables by
modifying the effect of the intervening variable/s .
Unlike extraneous variables, moderator variables
are measured and taken into consideration.
Typical moderator variables in language
acquisition research, when these are not the focus
of the study, include, age, gender, culture or
language proficiency of the subjects

Descriptive Statistics

Measures of Location/Center /Central Tendency


or Averages
A. When N is less than 30 (ungrouped data)
1. Median - called the counting median
- middlemost score of the distribution
of scores arranged from highest to
lowest when N is odd, the middle
most score is the counting median
- when N is even, the counting median
is the average of the two middlemost
scores.

Example:
Given: X
85
72

X
80 79
86
83 83
85
79 74 86
83
N=9
83
80 Md = 80
79
79
74
72

Given: X
59 53
40 54
48 44
40 40
N = 12

X
51
50
43
41

59
54
53
51

50
48
Md = 48 + 44
44
2
43
= 92
41
2
40
= 46
40

40

2. Mean - most reliable average


- called arithmetic average
- the average of the given
scores using the formula:
M = x
N
cases

where: x is the sum of scores


N is the number of
M is the mean

Example:
Given: X

X
85 80
72 83
89 74
N=9

79
83
85

85
83
80
89

79
74
85
83
72
-----------M = x = 730 = 81.11
N
9

3. Mode - used when one wants to see the trend of the


scores
- called rough mode
- the score with the greatest frequency in the
distribution
Example:
X
56
58
63
61
Mode = 58
62
60
55
58
--------N=8

- a distribution may have two or more modes


Example:
X
98
90
92
90 Modes: 90, 98 (bi-modal distribution)
85
89
98
97
80
90
89
98___
N = 12
*Therefore, when N is less than 30, the measures of central tendency
are the counting median, arithmetic mean and rough mode by the
inspection method.

B. When N is more than 30 (grouped data)


1. Median is the most stable measure of central tendency.
Formula: X = ll + i (N F) where: ll lower limit of class
2
interval where median
fm
falls
N half of the scores
2
F sum of all scores
below the lower limit
fm number of scores
within interval where
median falls
i class interval

Example:
Ungrouped scores in a 75-item test:
25 38 48 35 28 62 48 38 35
60 47 38 34 66 59 46 38 34
57 44 37 33 22 56 44 37 32
55 43 37 32 20 52 42 36 31
50 39 35 29 17 50 41 36 30
N = 50
a. Determine the Range (R)
R = Hs Ls
6617 = 49
b. Determine the class interval (i)
i=(R + 1) - 1
10
= (49 + 1) 1
10
= (50) - 1
10
= 51
= 4

27
23
21
19
18

Grouped scores:
Ci of scores

64 67
60 63
56 59
52 55
48 51
44 47
40- 43
(ll is 35.5) 36 39
32 35
28 31
24 27

1
2
3
2
4
4
3
10 (fm)
8 = 21(F)
4
2

20 23
16 19

4
3
______
N = 50

Given: N = 25
2
F = 21
fm = 10
ll = 35.5
i= 4
X = 35.5 + (25 21) 4
10
= 35.5 + .4 (4)
= 35.5 + 1.6
= 37.1 or 37 (Rating for this
score is 75%)

2. Mean is the most reliable


measure of central tendency.
There are several methods
used to find the mean when X
is greater than 30 but the
assumed mean (AM) method
is the most efficient and most
meaningful.

Formula:
M = AM + i (fd)
N
Ci of scores
64 67
60 63
56 59
52 55
48 51
44 47
40 - 43
(AM = 37.5) 36 39
32 35
28 31
24 27
20 23
16 19

fd

1
7
7
2
6
12
3
5
15
2
4
8
4
3
12
4
2
8
3
1
3 = 65
10
0
0
8
-1
-8
4
-2
-8
2
-3
-6
4
-4
-16
3
-5
-15 =-53
______
________
N = 50
fd = 12
M = AM + i (fd)
N
= 37.5 + 4 (12/50)
= 37.5 + 4(.24)
= 37.5 + .96
= 38.46

3. Mode (Mo) when N is greater than 30, the


Pearson mode is used.
Formula:
Mo = 3Md - 2M
Example:
Using the same distribution for the Md and M:
Mo = 3Md - 2M
= 3 (37.1) - 2 (38.46)
= 111.3 76.92
= 34.38

* Therefore:
Md = 37.1
M = 38.46
Mo = 34.38
* The mode is the highest of the three if
the Md is larger than the M. It is the
lowest if the M is greater than the MD.

Measures of Spread/Variability/Dispersion
- Show the tendency of the scores to scatter or
disperse above or below the measures of
central tendency
1. Range (R) the distance between the highest
score and the lowest score. Preferred than the
standard deviation to represent dispersion of
small data sets (e.g. number of samples is less
than 10)
Formula:
R = Hs Ls

2. Variance (S2) is simply the square of the


standard deviation. It represents the
average squared deviation from the mean.
It is seldom used.
Formula:
Example:

S2 = (d2/N) 2
Given:
SD= 3.67
S2 = (3.67)2
= 13.47

3. Standard Deviation (S or SD) is the


most accurate measure of variability. In
comparing two groups, the smaller is the
SD, the more homogonous is the group
while the higher is the SD, the more
heterogeneous is the group.
a. When N is less than 30

Formula:
SD = d2/N

Example:
X
d
23 (23-23)
0
22 (22-23)
-1
23
0
25
2
18
-5
30
7
18
-5
25
2
_______
M = 184/8
= 23

d2
0
1
0
4
25
49
25
4
___

SD = d2/N
= 108/8
= 13.5
= 3.67

______
d = 108
2

The Weighted Mean


It is an average in which each
quantity to be averaged is
designated a weight that
determines the relative
importance of each quantity on
the average

Formula: X = fixi
fi where:
x = weighted mean

xi = x1, x2, x3, xj


fi = f1, f2, f3, fj

= items/options given

frequencies corresponding

The Data-Gathering Questionnaire and


Validation
Research Instrument/Questionnaire a list of
standardized questions printed on a sheet of paper and
handed to the respondent who writes his/her responses
on the sheet itself. The questions may either be in the
Closed Form (structured/restricted) or in the Open
Form (unstructured/unrestricted). The questions in
the Closed Form permit only certain responses, while in
the Open Form allow the respondents to make any
responses as they wish, in their own words. Sometimes,
different response formats may be constructed in one
instrument. A mixture of two forms may be employed
depending on the objective of a particular question.

Guidelines in Questionnaire Construction


All aspects of the problem must be covered by the
questionnaire. Avoid questions which are not related
to the problem.
Make sure that the questions truly answer or measure
what is being investigated.
The questionnaire must be well-organized and within
the comprehension of those who will answer it.
The questions must be clearly and briefly worded.
The questionnaire should require a minimum amount
of writing only. Most of the questions should be briefly
answered with a checkmark or a fact or figure and the
number of questions requiring extensive subjective
replies be kept to a minimum.

Validation is the process of making


a research instrument reliable,
acceptable, contextual and applicable
to the participants of the study. A
researcher-designed instrument may
be validated by conducting a dry-run
to determine its reliability and validity.
A standard instrument should be
validated in the context of the study.

Questionnaire/Instrument
Validation
Reliability is the extent to which
the test/questionnaire is dependable,
self-consistent and stable. The test
agrees with itself. It is concerned with
the consistency of responses; even if
the person takes the test twice, the
same results are obtained.

Types/Methods in Determining the


Reliability of a Questionnaire:
Test-Retest
Reliability
the
same
measuring instrument is administered twice to
the same group of subjects. The reliability of
scores in the first and second administrations
of the test is determined by the Spearman
rank
correlation
coefficient
or
Spearman rho.

2. Internal consistency Method

determines whether the items or


questions in a test are consistent with
one another or not. An examinee either
passes or fails an item. One (1) is
assigned for a pass and zero (0) for a
failure.
The Kuder Richardson
Formula 20 is used.

Steps in Applying the Kuder Richardson


Formula:
Compute the variance (S2) of the test scores of the
whole group.
Find the proportion passing each item (pi) and the
proportion failing each item (qi). For instance, 9 of the
10 students passed or got the correct answer in Item 1,
(p = 9/10 = 0.9); and only 1 student failed in Item 1, (q i
= 1/10 = 0.1)
Multiply pi and qi for each item i.e 0.9 x 0.1 = 0.09;
then take the sum of all the items. This gives piqi value
of equivalent.
Substitute the computed values using the formula

Student
1
2
3
4
5
6
7
8
9
10
Total

X
4
5
5
7
12
14
17
18
19
20
121

(X-x)
-8.1
-7.1
-7.1
-5.1
0.1
1.9
4.9
5.9
6.9
7.9

S2 (X x)2
65.61
50.41
50.41
26.01
0.01
3.61
24.01
34.81
47.61
62.41
364.90

Validity is the extent to which a test


measures what it claims to measure. It is
vital for a test to be valid in order for the
results to be accurately applied and
interpreted. There are three types of
validity: content, criterion-related,
and construct. For criterion-related
validity and construct validity, the Pearson
r may be used while for content validity,
expert validation is usually sought.

Content validity the items in the


test represent the entire range of
possible items the test should cover.
Individual test questions may be
drawn from a large pool of items that
cover a broad range of topics. Experts
rate each items relevance. Items that
are rated as strongly relevant by all the
experts are included in the final draft.

Correlational Procedures
Pearson Product Moment Correlation the
commonly used type of correlation which determines the
relationship between two or more variables.
The
correlation (r) ranges from -1.00 to +1.00. A correlation of
1.00, whether it is positive or negative, is a perfect
correlation. This means that as scores on one of two
variables increase or decrease, the scores on the other
variable also increase or decrease by the same magnitude.
A correlation of zero (0) means there is no relationship
between the two variables, i.e., when scores on one of the
variables go up, scores on the other variable may go up,
down or whatever. The Pearson r is also used to establish
the reliability and validity of a data-gathering instrument.

Formula:

where: r is the correlation coefficient between X and Y


N is the sample size
X is the individuals score on the X variable
Y is the individuals score on the Y variable
XY is the product of each X score times its corresponding Y
score
X2 is the individual X score squared
Y2 is the individual Y score squared

Magnitude of Relationship by Henry Garrett:


r from .00 to + .20 low, negligible relationship
r from +.20 to + .40
present but slight relationship
r from +.40 to +.70 marked substantial relationship
r from +.70 to + 1.00 high to very high relationship

Example:
N
1
2
3
4
5
6
7
8
9
10

X
(Math
Scores)
12
10
14
11
8
7
10
4
11
12
99

Y
(English
Scores)
16
12
8
13
10
6
19
6
13
19
122

X2
144
100
196
121
64
49
100
16
121
144
1055

Y2
256
144
64
169
100
36
361
36
169
361
1696

XY
192
120
112
143
80
42
190
24
143
228
1274

Therefore, there is a marked substantial


positive relationship between the students
performance in the tests of Mathematics and
English. Their scores in Mathematics
increase as their scores in English also
increase, or the other way around.
*How significant is the obtained r ?

Formula:

Therefore, the relationship of the


students performance in their
Mathematics and English tests is not
significant. One cannot determine
whether they perform better in
Mathematics or in English. Meaning,
they have more or less the same level of
performance in Mathematics and
English.