Assessment in Learning 2 – Chapter 10

Chapter 10: Statistics and Computer: Tools for Analyzing of Assessment

Data
Presented by Group 10
Members Topic Distribution
Fabilane, Ryan Jeffrey Descriptive and Made the content for his
Q. Inferential Statistics, assigned topic.
Standard Scores,
Provided test questions.
Application to MS Excel
Gayapa, Kristian Blair G. Statistics, Measures of Made the content for his
Variability, Indicators of assigned topic.
Relationships Provided test questions.
Macawili, Jhonrey T. Introduction, Measures Made the content for his
of Central Tendency, assigned topic.
Computer: Aid in Provided test questions.
Statistical Computing
and Data Presentation

Reporter: Jhonrey
Overview
For most educators, just hearing the term statistics conjure images causing anxiety.
Even to those educators who do not consider themselves numerically deficient,
statistics intimidates them. In descriptive statistics, which will be the focus of
discussion in this chapter, the emphasis is on describing a set of scores. Description
may take the form of tables, graphs or as a single number (e.g., an average).
In this chapter, the basic concepts of statistics, the relevant statistical treatment that
can facilitate the interpretation and analysis of data, as well as the use of computer as
aid in analyzing and data presentation, will be covered.
Chapter Intended Learning Outcome
At the end of this chapter, you should be able utilize processed data and results in
reporting and interpreting learners' performance
Engage
Many teachers think that statistics requires advance mathematical skills and tedious
calculations. The elementary statistical concepts necessary in organization,
interpretation, analysis, and presentation for evaluation of grades and scores of
students is not what majority thinks of. At present, not only calculators are at hand to
facilitate computations, but most especially computers and some software which are
available for statistical computation and treatment.
The statistics that is discussed in this chapter is descriptive in nature and brief
overview of inferential statistics is given.
 The three types of statistical measures covered in this chapter are:

(a) Measures of Central Tendency

(b) Measure of Variations; and
(c) Measures of Relationship.

The statistical treatment for individual data subject to interpretation is also covered in
this chapter

Reporter: Kristian
1. STATISTICS
Statistics is concerned with the organization, analysis, and interpretation of test scores
and other numerical data. Statistical techniques help teachers to (1) analyze and
describe results of measurement obtained in their own classrooms, (2) understand the
statistics used in test manuals and research reports, and (3) interpret the various
types of derived scores used in assessment.
Statistical methods enable us to look at information from a small collection of people
or items and make inferences about a larger collection of people or items. Procedures
for analyzing data, together with rules of inference, are central topics in the study of
statistics. Brase & Brase (2012) define statistics as the study of how to collect,
organize, analyze, and interpret numerical information from data.

Reporter: Ryan
2. DESCRIPTIVE AND INFERENTIAL STATISTICS
Descriptive Statistics
This is used to describe a group of individuals or describe the data that have been
collected; to describe variables that were grouped in order to determine the measure
of certain dependent variable that needs to be measured. Various data analysis
techniques provide meaningful description of scores with small number of numerical
indices. Such indices are calculated using samples drawn from a population and
are called statistics. On the other hand, when indices are calculated using the entire
population, it is called parameters. This type of statistics is used for summarizing
and describing data sets.
Inferential Statistics
When there is a need to make a decision, estimate prediction, or generalize about a
population based on a sample, then inferential statistics will be utilized. There are two
types of tests in inferential statistics. These types are: (a) Parametric; and (b)
Nonparametric.
In parametric test, a test of significance is used if the data represent an interval or
ratio scale of measurement and other assumptions have been met. On the other
hand, non- parametric test is used when data represent an ordinal or nominal scale,
when a parametric assumption has been greatly violated, or when the nature of the
distribution is not known.
3. STATISTICAL TOOLS
To carry out meaningful comparisons among sets of tests scores, a statistical tool is
needed. Measures of central tendency and measures of variability are the most
commonly used statistical tools.

Reporter: Jhonrey
3.1 Measures of Central Tendency

Measures of Characteristics
Central
Tendency
The Mean • Most frequently used index of distribution's central
tendency.
• It is simply an arithmetic average of all the
scores in the distribution.
• Since it is computed using every score in the
distribution, it is an extremely representative
measure.
• Its strength is also its weakness because several
atypically
high or atypically low scores (sometimes called liars)
can distort the value of the mean.
The Median • The second most frequently used index of central
tendency.
• It is the point that divides the scores into two equal
halves.
• One advantage of median as a measure of central
tendency is that it is not unduly affected by peculiarly
large or small scores.
• Unlike the mean, the median treats each of these wild
scores as merely another score, no more no less
than other scores.
• As the mean, the median's strength is also its
weakness, the median fails to reflect the magnitude
of the impact of every score in the distribution, even
when certain of those scores are very high or very
low.
The Mode • It is the most frequently occurring score in the
distribution.
• With most reasonably large set of scores, the mode
will occur somewhere near the middle of the
distribution, so it can also serve as an index of the
distribution's central tendency.
• In some cases, a distribution has two or even three
most frequently occurring scores. In such cases,
statistician refer to it as bimodal or trimodal
 distribution. Multimodal distribution refers to those
with three or more frequently occurring Scores.
• Since the mode takes into account even fewer data
than the
• median, and fewer than the mean, it is not used often
in describing a distribution's central tendency

In a normally distributed set of scores, the mean, the median, and the mode are
identical. Imagine a bell-shaped symmetrical curve, the value of all these indicators
will coincide.
Oftentimes, if this is the case, only one of these indicators is used to describe how
the distribution's scores tend to center. In such cases, that indicator is usually the
mean. When there is a great difference between the numerical values of the mean
and the median, it is a good idea to describe the distribution's central tendency by
providing both the media and the mean, better yet the mode as well (Popham,
2000).

Reporter: Kristian
3.2 Measures of Variability
Aside from describing how a set of scores tends to the center of distribution, the
information about how variables the score are, is also important. It is also relevant to
know how spread out the scores are.
Measures of Characteristics
Variability
The Range • Most readily calculated index of distribution's
variability.
• It is calculated by subtracting the lowest score from
the highest score.
• The simplicity of the range's computation is just the
only redeeming virtue, because there are only two
scores involved in its computation. If there is an
abnormally highest and/or lowest score, the resulting
range will yield
a misleading indication of the distribution's overall
variability.
The Quartile • It is based on the range of the middle 50% of the
Deviation scores, instead of the range of the entire set.
• Itis also called semi-interquartile range.
Quartiles are points on the scale (like averages and
percentiles), the quartile deviation represents a
distance on the scale. It indicates the distance that is
necessary to go above and below the median to
include approximately the middle 50% of the scores.
The • It offers a way of thinking about the average
Standard variability of a set of scores when they compute the
Deviation mean.
• It tells the average distance from the mean for
 each of the scores in the distribution.
• The more spread out the scores are, the larger the
value of the standard deviation. The less spread out
the scores, the smaller the standard deviation is.
• The most useful measure of variability
• Because it takes into account the amount that each
score deviates from the mean, it is a more stable
measure of variability than the others.

In summary the quartile deviation is used with median and is reasonable for
analyzing small number of scores. These statistics are obtained by counting and
thus are not affected by the value of each score; they are especially useful when one
or more scores deviate markedly from the others in the set of scores. On the hand,
the standard deviation is used with the mean. It is a most reliable measure of
variability, and is especially useful in testing. Also, it is useful in describing the set of
scores in a group, it serves as basis for computing standard scores, the standard
error of measurement, and other statistics used in analyzing and interpreting
scores.

Reporter: Ryan
3.3 Standard Scores
Aside from the Measures of Central Tendency and Measures of Variability, another
group of standard scores that may be used in assessment are discussed in the
following table.
Standard Characteristics
Score
Z-Score • The simplest of the standard scores, and the one on which
others are based.
• It expresses test performance simply and directly as the
number of standard deviation units a raw score is above or
below the mean.
• A Z-Score is always negative when the raw score is smaller
than the mean. Thus z-score is seldom used directly in test
norms, but are transformed into a standard score system
that uses only positive numbers.
T-Score • It was originally given to a type of normalized score based
on a group of unselected twelve-year-old children.
• It refers to any set of normally distributed standard
scores that has a mean of 50 and a standard deviation
of 10.
• T-score is obtained by multiplying the z-score by 10 and
adding the product to 50. This formula is true only when
the original distribution of raw scores is normal.
• The reason that T-scores is preferred reporting test results
is that only positive integers are O to z-scores for produced.
 The Stanines • These are simple type of normalized standard score that
illustrates the process of normalization.
• They are a single digit scores ranging from 1 to 9.
• It is named as such because the distribution of raw score
is divided into nine parts wherein stanine 5 is precisely
the center of the distribution and includes all cases within
one-fourth of a standard deviation on either side of the
mean.
• Major strengths of stanines are the following:
• The Stanine system uses a nine-point scale in which 9 is
high, 1 is low and 5 is average
• Stanines are normalized standard scores that make it
possible to compare student's performance on various
assessments.
• It makes easy to combine different types of data because
it is computed as percentile ranks but are expressed in
standard score form.
• It uses a single-digit score, it is easily recorded and takes
up less space than others scores.
The Percentile • Indicates student's relative position in a group in terms of
Rank the percentage of students' scoring lower (Linn &
Gronlund, 2000). One of the most widely used and easily
understood methods of describing assessment
performance.

In summary, standard scores provide a student's relative position in a group in terms

of standard deviation units above or below the mean. In a normally distributed set
of scores, the different standard score scales and the percentile scale are interrelated,
thus making it possible to convert from one to another. One of the shortcomings of
standard scores is that students and parents find it difficult to understand. Another
limitation is that, interpretation is difficult unless the scores are evenly distributed.

Reporter: Kristian
3.4 Indicators of Relationship
There are techniques of describing the relationships of the scores in different subject
areas. These are: (1) Graphic Presentation; (2) The Product-Moment Correlation
Coefficient; and (3) The Rank-Order Correlation Coefficient.
3.4.1 Graphic Presentation
• By using graphic display techniques, the relationships between two variables
can be described according to its magnitude (strong or weak) and its direction
(positive negative). This technique though shows that it is difficult to get a firm
grasp on graphically displayed relationship because of the imprecision of
visually derived estimates.
3.4.2 The Product-Moment Correlation
• This is the most widely used indicator or the magnitude and direction of the
relationship between two variables, the product-moment correlation coefficient
 (popularly known as Pearson correlation coefficient named after the English
statistician, Karl Pearson.
• This product-moment correlation coefficient ranges from +1.00 to -1.00. An r of
+1.00 represents a perfect positive relationship while a value of -1.00
represents a perfect negative relationship. An r of zero indicates that there is
no linear relationship that exist between two variables. This correlation
approach is used with linearly related data, meaning, data whose scatterplots
shows a more or less straight-line relationship.
34.3 The Rank-Order Correlation Coefficient
• The Rank-Order Correlation Coefficient is another technique for estimating the
magnitude and direction of the relationship between two variables. This
technique was introduced by Sir Charles Spearman. Spearman's correlation
coefficient (rs) is interpreted in essentially the same way as Pearson product
moment r, but simpler to compute. Aside from its computational simplicity, it
may be used with data that represent only an ordinal scale and not necessarily
an interval.
• Recall that there are different measurement scales that are normally used in
education, most especially by educators involved in assessment activities.
These scales are interval scales, ratio scale, and ordinal scale. An interval
scale is one that allows you to believe that there are equal intervals that are
equidistant points on the scale. A ratio scale, is an interval scale for which a
zero point exists. An ordinal scale is based on rank.

Reporter: Jhonrey
4. COMPUTER: AID IN STATISTICAL COMPUTING AND DATA
PRESENTATION
In the previous paragraphs of this chapter, fundamental concepts of statistics and
statistical tools used for different statistical methods were discussed. In the
succeeding paragraph, simple computations of statistical treatment will be discussed
with emphasis on the use of computer and the software MS Excel, a software
developed by Microsoft, to make computation more efficient and less prone to error.

Basic Steps in Opening the MS Excel Program

1. Click the icon of MS Excel or if not available, click start button to choose all
programs and double click on MS Excel to open the program. This is the initial
window that will appear and if you're using the latest version of MS Excel, simply
click on blank document option. If you have a previously created file to work on, then
simply double click Open Other Worksheets. (figure on the left)
2. Once a worksheet has been opened, it is now ready for data entry. You may
now encode the scores (data) that will be processed. (figure on the right)
Now we may start encoding our data. We start our discussion with Charts and
Frequency Distribution.
Charts and Frequency Distribution
The table below contains the grades of 10 students in their Mathematics and
English subjects. The first column is the student class number, second column is the
Math grade and the third column is the English grade.
Using the encoded data, it may be converted into charts and/or diagrams.

A scatter diagram was created by using the insert chart function of MS Excel.
Detailed steps on how to have scatter diagrams are as follows:
1. Click the Insert Tab, then among the Chart options, choose insert Scatter.
• Note that other types of charts may be chosen based on preference of the
user on how thev would want to present their data. Bar graphs, pie charts, line
graphs, etc. are just some of the charts available and can be utilized.
Computing for the Measures of Central Tendency, Variability and
Correlation
Statistical functions are available in MS Excel. Click the FORMULAS tab and the
different available formulae logically organized will be displayed. Choose the More
Functions icon, then choose Statistical. Several statistical formulae are provided on
the list. For the Mean, the function average will be used.
• For example: For the Mean, choose AVERAGE from the list, once you click
the function average, the following screen will be displayed.
Fill-in the required fields to indicate which among the encoded data will be
considered in the computation of the mean, press ok, and automatically the mean
will be computed and placed on the space where you indicated to place the result, for
this case, the ones enclosed with rounded edge box.
Similar steps shall be followed for other statistical formulae, such as Standard
Deviation (STDEV), Median (MEDIAN), Mode (MODE), Range (RANGE), Pearson
(PEARSON), z-test (ZTEST) same user interface is displayed so it is easier for the
user to encode the necessary information such as the raw data that will be included in
processing/computation. Not all may be available in MS Excel, nevertheless majority
of those statistical computations are.
In the older version of MS Excel, the statistical formulae are not available. Instead,
most of the formulae are packaged all together and displayed into a single table
containing the computed values. This is packaged under Descriptive Statistics
function, Statistics Summary. The information contained in the "packaged function"
consist of the following: Mean, Median, Mode, Standard Deviation, Sample Variance,
Kurtosis, Skewness, Range, Minimum, Maximum, Sum, and Count.
Another software that was really designed to support statistical computations is the
SPSS (Statistical Package for Social Sciences), which has evolved into Statistical
Product and Service Solution. All the statistical formulae are available in this software.