Research paper writing
An Assessment & Accountability Brief: 2012-2013 EOCT Validity and Reliability
Georgia Department of Education Dr. John D. Barge, State School Superintendent
August 2013 Page 1 of 10 All Rights Reserved
The Georgia Department of Education (GaDOE) oversees the development of the Georgia End-of-Course Tests
(EOCT) and adheres to the Standards for Educational and Psychological Testing (1999) as established by the
American Educational Research Association (AERA), the American Psychological Association (APA), and the
National Council on Measurement in Education (NCME). The intent of these standards is “to promote the sound
and ethical use of tests and to provide a basis for evaluating the quality of testing practices” (AERA, APA, NCME, 2).
Critical to any assessment is the collection of evidence to establish validity and reliability, and the Standards
provides guidance on what test developers should do. While validity is the most important consideration in the
test development process, a test cannot be valid without a high degree of reliability.
Validity
According to the Standards, “validity refers to the degree to which evidence and theory support the
interpretations of test scores entailed by proposed uses of tests” (9). It is important to understand some key
elements of validity. First, validity exists in context. A test may have a high degree of validity for one proposed
use, but less validity for another. Second, validity is a matter of degree and is not an “all or nothing” condition.
Finally, validity is associated with a multi-faceted process and collection of evidence over time. Questions of
validity cannot be accurately summed up in a single statistic. Ultimately, the answer to a validity question for a
test rests in careful documentation of the test development process. The Standards include 24 separate standards
related to validity; however, these standards are not prescriptive for each test but should be considered as to
whether they are applicable for the particular test under consideration. This brief will succinctly describe the
major sources of validity evidence for the EOCT, including the following: content/curricular validity, construct
validity, and criterion-related validity.
Content/Curricular Validity
One of the first pieces of evidence for establishing a tests’ validity is a clear identification of the purpose of the
test. The purposes of the EOCT are to assess student achievement in the Georgia Performance Standards (GPS) in
the designated core courses and to provide diagnostic data to improve student achievement. These tests provide valuable information for students, educators, and parents about student strengths and areas for improvement.
The tests identify students who may need additional instruction in the core content and skills required for a
diploma. Georgia law (O.C.G.A. §20-2-281) states “the State Board of Education shall adopt end-of-course
assessments for students in grades 9 through 12 for all core subjects to be determined by the state board.” The
EOCT assess only a sample of the knowledge and skills that educators agree comprise a complete curriculum for
each subject. Georgia students have opportunities to learn – and are expected to master – much more than the
tests address. The Georgia Performance Standards are the source of the knowledge and skills assessed on the
Validity and Reliability
for the 2012-2013
Georgia End-of-Course Tests
An Assessment & Accountability Brief: 2012-2013 EOCT Validity and Reliability
Georgia Department of Education Dr. John D. Barge, State School Superintendent
August 2013 Page 2 of 10 All Rights Reserved
EOCT in the content areas of English/Language Arts, Mathematics, Science and Social Studies. The following
courses are assessed by the EOCT:
Ninth Grade Literature & Composition
American Literature & Composition
Biology
Physical Science
US History
Economics
Mathematics I
Mathematics II
Coordinate Algebra
Algebra
Geometry
These assessments yield information on academic achievement at the student, class, school, system, and state
levels. The evidence then for the validity of the EOCT relies primarily on how well the assessment instruments
match the intended curriculum and how the score reports inform the various stakeholders – students, parents, and
educators – about the students’ performance.
The test development cycle for the EOCT must begin with the state’s mandated curriculum. Because the GaDOE
believes that the curriculum resides both in the approved published curricular documents as well as the way it is
enacted in the classroom, the test development process also relies heavily on the inclusion of educators from
around the state. Once the purpose of the test is established, committees of educators are formed to review the
curriculum and establish which concepts, knowledge, and skills will be assessed and how they will be assessed.
The results of this curriculum review are a test blueprint and test specifications in which the committees of Georgia
educators indicate which standards can and will be measured and how they will be represented on the
assessment.
From these two documents, an additional document is developed that provides more details for the item writing
phase. Through careful review of the state curriculum, content domain specifications are produced. These
indicate how specific standards or elements of the curriculum will be grouped into domains or strands. Moreover,
test item specifications are produced which give additional detail about what kinds of items will be written. This
document typically identifies the item format, content scope and limits, and cognitive complexity. For example,
item specifications for a Ninth Grade Literature assessment may address the genre, complexity, and/or length of
literary passages to be produced. All of these activities are conducted by the GaDOE with the assessment
contractor, and with substantial involvement by curricular specialists and Georgia educators. The content domain
specifications along with the test blueprints and test specifications are then combined into a document known as
the EOCT Content Descriptions. The EOCT Content Descriptions are posted on the department’s website so that all
stakeholders are informed of the test’s content and method of assessment. These documents organize the
assessed content under the test domains, the structure in which the test results will be reported. In addition, a
Content Weight document is produced that shows the relative proportion of items by domain that are included on
each content area test. These documents and the inclusion of Georgia educators serve as one piece of evidence of
the EOCT’s validity as a measure of the state’s curriculum. These documents are posted to the GaDOE’s website
on the EOCT testing page:
http://www.gadoe.org/Curriculum-Instruction-and-Assessment/Assessment/Pages/EOCT-Resources.aspx
Once these aforementioned documents are written and approved, item development can begin. Items are written
by qualified, professional assessment specialists specifically for Georgia tests. Committees of Georgia educators
review the items for alignment with the curriculum, suitability, and potential bias or sensitivity issues. The
educator review committees have the authority to accept the item as is, revise it, or reject the item. Items that are
An Assessment & Accountability Brief: 2012-2013 EOCT Validity and Reliability
Georgia Department of Education Dr. John D. Barge, State School Superintendent
August 2013 Page 3 of 10 All Rights Reserved
accepted are placed on field tests. Field tests, which are trial runs of the test items, are designed to help ensure
that the items function appropriately and are not confusing for students. This is typically accomplished by
embedding field test items in the operational test. Embedding items is a commonly used and well-regarded
practice that ensures the field test items are taken by a representative group of motivated students under
standard conditions.
After the items have been field tested, another committee of Georgia educators examines the items again, along
with the data from the field test. The committee reviews how items performed in terms of how many students
selected the correct answers and how many students selected each incorrect answer. The review also includes an
analysis of how different groups of students performed to detect potential bias (i.e., did the item appear to favor
one group of students over another?). Once again the review committees have the authority to accept items as is,
revise items for re-field testing, or reject items. Accepted items are then banked for future inclusion on an
operational test form. Only after items have been field tested and approved by Georgia educators do they appear
on an operational test form.
The next stage of test development consists of developing the actual test form that students will take. Items are
carefully selected for a test form based on the blueprint developed by Georgia educators. Putting together a test
form requires consideration of both content and statistical data. Each form of a test must assess the same range
of content as well as carry the same statistical attributes.
Tests must be equated when multiple test forms are used in the same administration or when a test is given in
subsequent administrations (e.g., year-to-year tests). Equating refers to the use of a statistical procedure to make
sure that the tests are of equal difficulty. This process is critical because it ensures that students are always held to
the same standard. Additionally, it permits one to interpret differences in test performance as the result of
changes in student achievement as opposed to fluctuations in the properties of the test form.
When a test is administered for the first time, performance level standards must be established for the test. The
standard setting process is the means by which educators decide what number of correct items a student must
obtain in order to meet or exceed expectations. These decisions, known as cut scores, become recommendations
to the State Board of Education that has authority to make the final decision on test standards.
The final stage in test development is to produce scores and distribute results. Scores are typically reported as
scale scores and performance levels. A scale score is based on the raw score (i.e., number of items correct) on a
test. The changing of raw score to scale scores is analogous to converting from the Celsius scale to the Fahrenheit
scale to report temperature. Scale scores are commonly used in large assessment programs. As an example,
scores on the SAT, the widely used college entrance exam, are reported on a scale ranging from 200 – 800. Each
time a new version of the SAT is administered, the raw scores are converted to this same scale, in order to take
into account any differences between various forms of the tests. Likewise, the EOCT results are presented in scale
scores, meaning that results can be consistently and meaningfully interpreted by students, parents, and educators.
To assist the stakeholders in interpreting results, an interpretive guide is produced. This document sets forth how
test scores should be interpreted and used and clearly indicates how scores for students with accommodations
should be interpreted with caution.
An Assessment & Accountability Brief: 2012-2013 EOCT Validity and Reliability
Georgia Department of Education Dr. John D. Barge, State School Superintendent
August 2013 Page 4 of 10 All Rights Reserved
The following table shows the scale scores, performance levels, and grade conversion information for the EOCT.
For more information on interpreting scores, find the EOCT Interpretive Guide 2012-2013 on the GaDOE website:
http://www.gadoe.org/Curriculum-Instruction-and-Assessment/Assessment/Pages/EOCT-Resources.aspx
Georgia End-of-Course Tests – Scale Scores and Performance Levels
Does Not Meet Expectations
Meets Expectations Exceeds Expectations
Scale Score
Grade Conversion
Scale Score Grade
Conversion Scale Score
Grade Conversion
Ninth Grade Literature & Composition
Below 400 Below 70 400 to 449 70 to 89 450 or Above 90 or Above
American Literature & Composition
Below 400 Below 70 400 to 449 70 to 89 450 or Above 90 or Above
Biology Below 400 Below 70 400 to 449 70 to 89 450 or Above 90 or Above
Physical Science Below 400 Below 70 400 to 449 70 to 89 450 or Above 90 or Above
US History Below 400 Below 70 400 to 449 70 to 89 450 or Above 90 or Above
Economics Below 400 Below 70 400 to 449 70 to 89 450 or Above 90 or Above
Mathematics I Below 400 Below 70 400 to 449 70 to 89 450 or Above 90 or Above
Mathematics II Below 400 Below 70 400 to 449 70 to 89 450 or Above 90 or Above
Algebra Below 400 Below 70 400 to 449 70 to 89 450 or Above 90 or Above
Geometry Below 400 Below 70 400 to 449 70 to 89 450 or Above 90 or Above
Coordinate Algebra Below 400 Below 70 400 to 449 70 to 89 450 or Above 90 or Above
Another piece of validity evidence relates to how the test is used. Federal law (P.L.107-110; NCLB) called for
rigorous examinations based on rigorous academic content to be in place for reading and mathematics in grades
3–8 and high school and for these tests to yield scores that identified a minimum of three specific categories of
student achievement, such as Basic, Proficient, and Advanced. Although the law did not require states to have
end-of-course assessments, some form of test for all students at the high school level was required. Since 2003,
the Georgia High School Graduation Test (GHSGT) served as the accountability indicator for Adequate Yearly
Progress. However, with the acceptance of Georgia’s waiver for NCLB, the EOCT becomes the high school indicator
for Georgia.
A significant change to Georgia’s assessment program was approved by the State Board of Education in April 2011.
This plan begins the process of phasing out the GHSGT and heightens the role of the EOCT by increasing the weight
the EOCT carries in the overall course grade of students. The Secondary Assessment Transition Plan resulted in the
following changes:
For students enrolled in grade nine for the first time on or after July 1, 2011:
The GHSGT is no longer required to earn a high school diploma for students. These students will not take
and are not required to pass the GHSGT.
An Assessment & Accountability Brief: 2012-2013 EOCT Validity and Reliability
Georgia Department of Education Dr. John D. Barge, State School Superintendent
August 2013 Page 5 of 10 All Rights Reserved
The EOCT will contribute 20% to the course grade (for those core courses in which an EOCT is mandated).
This cohort is required to pass the course with the increased weight of the EOCT applied to the course
grade.
The requirement to achieve a passing score on the Georgia High School Writing Test (GHSWT) remains for
all students, including those in this cohort.
For students who enrolled in grade nine for the first time between July 1, 2008, and June 30, 2011:
The graduation assessment requirement for diploma eligibility may be achieved through the
demonstration of proficiency on one of the two EOCTs in each content area (ELA, Mathematics, Science,
and Social Studies) or the corresponding content area GHSGT.
Under this flexibility, a passing score on a content area EOCT can serve as an alternate demonstration of
proficiency and stand in lieu of the requirement to pass the corresponding content area GHSGT.
For this cohort, the EOCT will continue to contribute 15% to the student’s course grade.
The GHSGT will remain available for any student unable to meet the testing requirement through the
EOCT in any content area.
The requirement to achieve a passing score on the GHSWT will remain for all students, including those in
this cohort.
For all students, EOCT scores account for either 15% or 20% of a student’s overall grade in EOCT-associated
courses: 9 th
Grade Literature; American Literature; Biology; Physical Science; US History; Economics; Mathematics
I; Mathematics II; GPS Algebra; and, GPS Geometry. The student must have a final course grade of 70 or above to
pass the course and earn credit toward graduation. Thus, a situation could exist in which a student does not pass a
particular EOCT, yet passes the associated course. Conversely, a similar situation could exist in which a student
passes a particular EOCT, yet does not pass the associated course. For certain cohorts of students, a passing score
on a content area EOCT can stand in lieu of the requirement to pass the corresponding content area GHSGT.
However, for both GHSGT and EOCT, it is incumbent upon all those involved in the education process to ensure
students have had the opportunity to learn the curricula. Improved teaching and learning are the main focus of
Georgia’s education system. The EOCT align with the Georgia curriculum standards and include assessment of
specific content knowledge and skills. The assessments provide diagnostic information to help students identify
strengths and areas of need in learning, therefore improving performance in all high school courses and on other
assessments. The EOCT also provide data to evaluate the effectiveness of classroom instruction at the school and
system levels. Due notification is provided to high school students as they enter the relevant EOCT course.
Students are made aware of the requirements for the course, as well as the impact of the EOCT on the course
grade. Student guides are available for each EOCT to help students prepare to take the assessments. The study
guides provide information about the EOCT, tips on how to prepare for it, and some suggested strategies students
can use to perform their best. The study guide can be found on the GaDOE website:
http://www.gadoe.org/Curriculum-Instruction-and-Assessment/Assessment/Pages/EOCT-Guides.aspx.
By attending carefully to each phase of the test development process, the GaDOE can ensure that the EOCT is a
valid instrument. The EOCT contractors produce documentation of each phase of the test development process
and produce various pieces of evidence. The alignment of the EOCT with the state’s curriculum and the reliance of
input from Georgia educators at every phase of test development are critical to the test’s validity. In addition, the
department has collected evidence through separate independent alignment studies to ensure that the test
measures the state’s curriculum. The department has also conducted analyses as evidence of external validity by
comparing how the constructs the EOCT measures compare with other assessments.
An Assessment & Accountability Brief: 2012-2013 EOCT Validity and Reliability
Georgia Department of Education Dr. John D. Barge, State School Superintendent
August 2013 Page 6 of 10 All Rights Reserved
Construct Validity
The term construct validity refers to the degree to which the test score is a measure of the psychological
characteristic (i.e., construct) of interest. A construct is an individual characteristic assumed to exist in order to
explain some aspect of behavior (Linn & Gronlund, 1995). When a particular characteristic from the assessment
results is inferred, a generalization or interpretation of some construct is made. For example, problem solving is a
construct. An inference that students who master the mathematical reasoning portion of an assessment are good
problem-solvers implies an interpretation of the results of the assessment in terms of a construct. To make such an
inference, it is important to demonstrate that this is a reasonable and valid use of the results. Construct-related
validity evidence can come from many sources. The American Psychological Association provides the following list
of possible sources (APA, 1999):
high inter-correlations among assessment items or tasks attest that the items are measuring the same
trait, such as a content objective, sub-domain or construct
substantial relationships between the assessment results and other measures of the same defined
construct
little or no relationship between the assessment results and other measures that are clearly not of the
defined construct
substantial relationships between different methods of measurement regarding the same defined
construct
relationships to non-assessment measures of the same defined construct
The collection of construct-related evidence is a continuous process. Two current metrics of construct validity for
the EOCT are item point-biserial correlations (i.e., item total correlation) and Rasch fit statistics. A point-biserial
correlation is the correlation between an item and the total test score. Conceptually, if an item has a high point-
biserial correlation (i.e., 0.30 or above), it indicates that students who performed well on the test overall answered
the item correctly and students who performed poorly on the test overall answered the item incorrectly. That is,
the item did a good job in discriminating between high-ability and low-ability students. Presuming that the total
test score represents the extent to which a student possesses the construct being measured by the test, high
point-biserial correlations indicate that the items on the test require knowledge of this construct in order to be
answered correctly.
In addition to item point-biserial correlations, Rasch fit statistics also provide good evidence of construct validity.
The Rasch model is a unidimensional Item Response Theory (IRT) model. Therefore, statistics showing that the
items fit the measurement model also provide evidence of construct validity. Rasch fit statistics are observed
closely during the test construction process to ensure evidence on construct validity.
Another piece of supporting evidence of the test unidimensionality is the single dominating factor extracted from
the exploratory factor analysis. There was a dominant factor observed in each of the scree plots for the tests,
suggesting that the IRT unidimensionality assumption held reasonably well. This is further evidence on construct
validity.
An Assessment & Accountability Brief: 2012-2013 EOCT Validity and Reliability
Georgia Department of Education Dr. John D. Barge, State School Superintendent
August 2013 Page 7 of 10 All Rights Reserved
Criterion-Related Validity
Although the primary evidence for the validity of the EOCT lies in the content and construct validity of the test, it is
also informative to collect criterion-related validity evidence. The term criterion validity refers to the degree to
which a test correlates with one or more outcome criteria. Criterion validity addresses how accurately criterion
performance can be predicted from test scores. The key to criterion-related evidence is the degree of relationship
between the assessment items or tasks and the outcome criterion. To ensure a good relationship between the
assessment and the criterion, the criterion should be relevant to the assessment and it should also be reliable.
Independent of a clearly defined criterion, the results of a performance assessment should agree, for the most
part, with the results of a standardized assessment. For example, a system can correlate scores of a standardized
mathematics assessment with other measures of mathematics content to provide criterion-related evidence. In
other words, if the inferences about student performance based on the mathematics performance assessment are
valid, one would expect them to be in agreement with the results from other measures of mathematics content.
Reliability
For a test to be valid, it must also have reliability. However, the inverse of this is not true – a reliable measure is
not necessarily valid. It is imperative that a test’s validity must be established first and foremost, but aspects of
the test’s reliability must also be addressed during test development. So just what is reliability? Reliability is the
consistency (and hence, precision) of the results obtained from a measurement. When a score is reported for a
student, there is an expectation that if the student had instead taken a different but equivalent version of the test,
a similar score would have been achieved. A test that does not meet this expectation (i.e., a test that does not
measure student ability and knowledge consistently) has little or no value. Further, the ability to measure
consistently is a prerequisite to making appropriate interpretations of scores on the measure (i.e., showing
evidence of valid use of the results). According to the Standards, reliability is “the degree to which test scores for a
group of test takers are consistent over repeated applications of a measurement procedure and hence are inferred
to be dependable, and repeatable for an individual test taker; the degree to which scores are free of errors of
measurement for a given group” (180). In other words, a reliable assessment is one that would produce stable
scores if the same group of students were to take the same test repeatedly without any fatigue or memory effects.
For the EOCT program, two reliability indices are reported. The first index is Cronbach’s alpha reliability coefficient
(1951). A reliability coefficient expresses the consistency of test scores as the ratio of true score variance to
observed total score variance (i.e., true score variance plus error variance). Cronbach’s alpha measures the
internal consistency over the responses to a set of items measuring an underlying unidimensional trait. Cronbach’s
alpha is computed using Crocker and Algina’s formula (1986):
2
1
2
1 1 X
N
i
Y
s
s
N
N i
An Assessment & Accountability Brief: 2012-2013 EOCT Validity and Reliability
Georgia Department of Education Dr. John D. Barge, State School Superintendent
August 2013 Page 8 of 10 All Rights Reserved
where N is the number of items on the test, 2
iY S is the sample variance of the
thi item (or component), and 2XS is
the observed score sample variance for the test. As shown in the table below, summer 2012 coefficient alpha values are slightly lower than winter 2012 or spring
2013 administrations in all courses except American Literature, Mathematics I, and Algebra. It is important to note
that, in general, student populations vary across administrations. Winter participation is usually comprised of
students on a block schedule. Spring administrations are generally comprised of both full year and block
scheduled students. However, participants in summer administrations are generally repeater students, and are
thus less representative of the general student population than participants in the other two administrations.
The internal consistency index alpha values are in the range of 0.75 to 0.92 for summer 2012 administration, 0.76
to 0.93 for winter 2012 administration, and 0.74 to 0.94 for spring 2013 administration. These values are in the
range of industry standards for a criterion-referenced test like the EOCT, suggesting adequate reliability of the test
scores. One thing to note is that the alpha index, being a correlation coefficient, is dependent on variability of the
scores.
SUMMARY OF COEFFICIENT ALPHA ACROSS ADMINISTRATIONS
Summer 2012 Winter 2012
Form 1/Form 2 Spring 2013
Form 1/Form 2
Ninth Grade Literature & Composition
0.89 0.92/0.92 .092/0.92
American Literature & Composition
0.89 0.89/0.90 0.90/0.90
Biology 0.86 0.92/0.92 0.94/0.93
Physical Science 0.87 0.90/0.91 0.91/0.89
US History 0.88 0.93/0.92 0.94/0.94
Economics 0.92 0.92/0.92 0.92/0.93
Mathematics I 0.79 0.76/0.79 0.83/0.83
Mathematics II 0.75 0.89/0.89 0.85/0.87
Algebra 0.82 0.77/0.76 0.74/0.78
Geometry 0.82 0.86/0.84 0.88/0.90
Coordinate Algebra 0.89/NA 0.87/0.86
Use of SEM
The Standard Error of Measurement (SEM) is used to quantify the precision of a test in the metric on which scores
will be reported. The SEM can be helpful for quantifying the extent of errors occurring on a test. An SEM band
placed around the observed score would result in a range of values most likely to contain the student’s true score.
This information is useful for determining the range within which a student’s true score is likely to fall. The true
score may be expected to fall within one SEM of the observed score 68% of the time.
An Assessment & Accountability Brief: 2012-2013 EOCT Validity and Reliability
Georgia Department of Education Dr. John D. Barge, State School Superintendent
August 2013 Page 9 of 10 All Rights Reserved
The SEM of a student who received a raw score of 48 on a test with reliability of 0.88 and a standard deviation of
12.1 is shown below.
19.488.011.12SEM
Placing one SEM band around this student’s raw score would result in a score range of 43.81 to 52.19 (i.e., 48 ±
4.19). Furthermore, if it is assumed the errors are normally distributed and if this procedure were replicated
across repeated testing opportunities, this student’s true score would be expected to fall within the ±1 SEM band
68% of the time (assuming no learning or memory effects). Since this error band is made up of ±1 standard error
of measurement, the chances are better than 2 out of 3 that a student with an observed raw score of 48 would
have a true score within this interval. This interval is called a confidence interval or confidence band. Increasing
the range of the confidence interval increases the likelihood that the confidence interval includes the true score;
an interval of ± 1.96 SEMs around the observed score covers the true score with 95% probability and is referred to
as a 95% confidence interval.
As shown in the table below, the SEM value ranges from 3.34 to 3.82 for summer 2012 administration, 3.28 to 3.71
for winter 2012 administration, and 3.26 to 3.63 for spring 2013 administration. For tests with total possible raw
scores ranging from 54 to 75, the error bands are reasonably small. This indicates reliability is generally high across
various EOCT.
SUMMARY OF SEMS ACROSS ADMINISTRATIONS
Summer 2012 Winter 2012
Form 1/Form 2 Spring 2013
Form 1/Form 2
Ninth Grade Literature & Composition
3.65 3.35/3.35 3.30/3.31
American Literature & Composition
3.61 3.52/3.46 3.43/3.30
Biology 3.82 3.65/3.64 3.52/3.52
Physical Science 3.71 3.71/3.63 3.61/3.63
US History 3.68 3.49/3.54 3.43/3.39
Economics 3.61 3.53/3.50 3.52/3.44
Mathematics I 3.40 3.42/3.38 3.36/3.35
Mathematics II 3.34 3.30/3.28 3.35/3.34
Algebra 3.38 3.38/3.38 3.37/3.37
Geometry 3.35 3.30/3.30 3.28/3.26
Coordinate Algebra 3.28/NA 3.29/3.29
Summary
Foremost, the EOCT have a high degree of validity because they serve the purpose for which they are intended – to
measure student mastery of the state’s curriculum. Validity is established via the process of test development.
The careful development from inception of the EOCT program and all steps in-between such as alignment with
An Assessment & Accountability Brief: 2012-2013 EOCT Validity and Reliability
Georgia Department of Education Dr. John D. Barge, State School Superintendent
August 2013 Page 10 of 10 All Rights Reserved
curriculum, creation of test and item specifications, clear and timely notification to students and their teachers of
test content, multiple reviews by educators, and finally careful form construction by content experts and
psychometricians provide evidence that the EOCT are valid instruments for the uses for which the department has
developed the test. The various reliability indices for the EOCT indicate that the test provides consistent results
and that the various generalizations of test results are justifiable. These strong indicators of reliability also support
the tests’ claim for validity.
References To enhance one’s understanding of the concepts presented in this brief, the following references are provided.
The ones with an asterisk (*) are cited in the brief.
*American Educational Research Association (AERA), American Psychological Association (APA), & National Council on Measurement in Education (NCME). (1999). Standards for Educational and Psychological Testing. Washington, DC: AERA.
Anastasi, A. (1988). Psychological Testing. New York: MacMillan Publishing Company.
*Crocker, L., & Algina, J. (1986). Introduction to Classical and Modern Test Theory. Belmont, CA: Wadsworth.
*Cronbach, L. J. (1951). Coefficient alpha and the internal structure of tests. Psychometrika, 16 (3), 297-334.
Hambleton, R. K. & Swaminathan, H. (1985). Item Response Theory: Principles and Applications. Hingham, MA: Kluwer Nijhoff.
Hambleton, R. K. & Zaal, J. N. (Eds.). (1991). Advances in Educational and Psychological Testing. Boston: Kluwer Academic.
Lyman, H. B. (1993). Test Scores and What They Mean. Boston: Allyn and Bacon.
McMillan, J. H. (2001). Essential Assessment Concepts for Teachers and Administrators. Thousand Oaks, CA: Corwin Publishing Company.
Nunnally, J. C. & Bernstein, I. H. (1994). Psychometric Theory, 3rd ed. New York: McGraw Hill.
Popham, W. J. (1998). Classroom Assessment, What Teachers Need to Know. Boston: Allyn and Bacon.
Thorndike, R. M. (1996). Measurement and Evaluation in Psychology and Education, 6th ed. Upper Saddle River, NJ: Prentice Hall.
This brief is produced by the Assessment Research and Development staff of the Georgia Department of Education.
Questions should be directed to the Assessment Research and Development staff at 404-656-2668.