Chapter 5 Constructing Tests and Performance Assessments

Chapter 5 Constructing Tests and Performance Assessments 123
Chapter 5 Constructing Tests and Performance Assessments
In a “standards based” approach to education and training, informed by Constructivist

theory, assessment informed instruction is the expectation as is continuous improvement.
One of the most widely used tools in assessment and evaluation is the traditional or
classic classroom achievement test, whether the classroom is on- or offline. These
measures are often fraught with reliability and validity problems as the process for
constructing such tests is often not followed or misunderstood, thereby introducing
significant measurement error into the measurement process. Poor measurement
frequently leads to inaccurate data-based inferences, which in turn leads to bad decision-
making.
A test is any device (written, observational, or oral) utilized to gather data for
assessment and evaluation purposes. The chief assessment device in education and
training is the test. Lyman (1998, pp. 21-26) offers one classification taxonomy of
various types of tests:
1. Maximum performance tests (MPT): with these, we assume that all examinees
will perform their best as all examinees are equally, highly motivated. Examinee
performance is influenced by the effects of heredity, schooling or training, and his
or her personal environment. A test may fall into more than one classification.
a. Intelligence tests: These are tests of general aptitude. IQ (Intelligence
Quotient) scores will most likely vary according to the intelligence test taken
as they do tend to be different. IQ is influenced by previous academic
achievement. IQ tests rely on a theory of intelligence, requiring construct
validity.
b. Aptitude (or ability) tests: These tests measure performance potential.
Aptitude tests imply prediction and are sometimes substituted for intelligence
tests, and used for classification (e.g., ability grouping). Ability test scores are
influenced by prior achievement (e.g., reading and math knowledge and
skills.). These rely on content validity.
c. Achievement tests are used to measure examinees’ current knowledge and
skill level. These tend rely on content validity.
d. Speeded tests are maximum performance tests where the speed at which the
test is completed is a vital element in scoring (e.g., typing test, rifle assembly,
foot race, etc.) If a test has a time limit such that virtually all students finish
the test, then it is not considered a speeded test and is called a power test.
Most achievement tests are also power tests.
e. Performance tests are designed to require examinees to demonstrate
competence by constructing a response. In “traditional” testing, this takes
form as a constructed response, i.e., brief or extended essays. A performance
assessment may also require a competence demonstration via the construction
of a work product; in this instance, detailed product specifications, scoring
criteria, and rating or scoring sheets are used.
Measuring Learning & Performance: A Primer | Retrieved from CharlesDennisHale.org

2. Typical performance test (TPT): These tests include personality, interest,

preference, and values tests. They are called instruments, scales, inventories, and
indexes. With these types of tests, we assume that examinees will perform
typically or “as usual.” There is less agreement with what is being measured and
what the score means. That is why it is essential that any theory upon which a
typical performance test is based be explicitly defined and explained. There is an
assumption that an examinee answers test items truthfully. The application of
typical performance tests should be cautious in educational applications. These
“tests” rely primarily on construct validity.
3. Standardized tests are aptitude or achievement tests which are administered under
standard conditions and whose scores are interpreted under standardized rules,
typically employing standard scores such as norms, except on criterion-referenced
tests. On most standardized tests, the breadth of content coverage is broad and test
items are written to maximize score variability. Thus, items which are too hard or
too easy are not used in constructing the test. Standardized achievement and
aptitude tests rely on content validity.
4. Informal tests are those typically intended for a localized purpose, such as the
simple classroom achievement test.
We focus on achievement testing which is used most to measure student or trainee

mastery or command of a subject area (e.g., math, biology, safety laws, customer service
policies, etc.). The achievement test construction process is seven (7) steps:
1. First, when planning the assessment or test, consider examinees’ age, stage of
development, ability level, culture, etc. These factors will influence construction
of learning targets or outcomes, the types of item formats selected, how items are
actually written, and test length.
2. Second, it is necessary that the content, intellectual or thinking skills (see

Appendix 5.1) psychomotor skills, and/or attitudes to be assessed are fully and
clearly identified and written. This is usually done through learning outcomes or
standards and benchmarks. Some write lesson plans at this step.
3. Third, the test item specifications are written. Item specifications intended for a
state-, provincial- or nation-wide tests will be more detailed than will a common
departmental or classroom examination, which will probably be just the learning
outcome/target benchmarks and possible test item formats. Detailed item specs
are written if different item writers construct tests for different examinee pools,
e.g., school districts permitting each school to write its own unit examinations.
4. Fourth, a table of specifications (or test blueprint) is developed which integrates

the content, psychomotor skills, and/or attitudes to be assessed with the
intellectual skills and selected test item formats, including the number of items
per learning target benchmark. See Table 5.1.

5. Fifth, once steps 1 to 4 are completed, the plan is initially reviewed by

knowledgeable colleagues to assess accuracy, clarity, continuity, and “fit.” The
reviewers also try to detect bias. This is an iterative process which should be
engaged until there is general agreement, there is no bias present. This step is
critical to establishing content validity.
6. Sixth, once a consensus has been reached, items are written, based on item writing
guidelines, and revised until there is agreement that each item (a) meets its item
specification, (b) conforms to item writing guidelines, (c) “fits” the test blueprint,
and (d) no bias is present; the determination is done by the same or similar subject
matter experts completing Step 4, above. Step 6 is again critical to establishing
content validity. The outcome of Step 6 is an initial version of the test.
7. Seventh, once the initial test is written, it is typically pilot tested with examinees
similar to the intended target audience and evaluated statistically (using item
analysis indices; see Part IV of the chapter). The statistical evaluation provides
guidance for the revision or rejection of poorly functioning items. Pilot more test
items than will be actually needed. The outcome of Step 7 is the final version of
the test.
In this chapter, we will first examine “mechanics of achievement test construction”

(e.g., test planning, building learning targets, test blueprinting, test item type and item
format selection), writing select response items (multiple-choice, true/false, matching,
completion and short-answer) and supply response items (brief and extended response) is
discussed. Next, statistical strategies for improving test items are presented, followed by
performance assessment. Ethical strategies for preparing examinees for a testing session
are presented in Appendix 5.7.
I. The Mechanics of Achievement Test Construction

A. Learning Standards, Targets, or Outcomes: The Assessment Drivers
1. In education and training, learning outcomes, also called standards or targets,
are framed to specify the content, skill, or attitude to be measured. These
standards drive curriculum writing, teaching and learning, test construction,
and testing (assessment).
a. Learning outcomes or standards have also been referred to as content
standards, performance standards, or behavioral objectives. Here, these
terms are used interchangeably.
b. There are three (3) types of learning standards, targets, or outcomes.
(1) Types of Learning Standards
(a) Attitude standards state explicitly what attitudes, based on defined
values, the faculty expect students to hold and articulate as a
function of program or course enrollment.
(b) Content standards express explicitly what content students are
expected to know.

(c) Skill standards state very clearly the specific skills students are
expected to have mastered at a specified performance level.
(2) It is often easy to confuse content and skill standards.
(a) More specifically, content standards specify declarative
knowledge, e.g., mathematical rules, statistical formulas, important
historical facts, grammar rules, or steps in conducting a biology
experiment, etc.
(b) Skill standards specify procedural or conditional knowledge, e.g.,
conducting statistical or mathematical operations based on
formulas or mathematical rules; interpreting or explaining
historical facts or analyzing historical data; correcting a passage of
text for poor grammar; or conducting a biology experiment.
(c) The key difference between content and skill standards is that with
content standards, students are required to possess specific
knowledge; skill standards require students to apply that
knowledge in some expected fashion at an expected level of
performance.
c. In crafting learning standards, targets, or outcomes, Oosterhof (1994, pp.
43-47) suggests the writer consider:
(1) Capability. Identify the intellectual capability being assessed.
Performance taxonomies such as Bloom, et al. or Gagne are helpful
here.
(2) Behavior. Indicate the specific behavior which is to be evidence that
the targeted learning has occurred. The behavior should be directly
observable, requiring no inference.
(3) Situation. Often, it is helpful to specify the conditions under which the
behavior is to be demonstrated. Describe the circumstances the
behavior is to be demonstrated.
(4) Special Conditions. Depending on the circumstances, one may need to
place conditions on the behavior, e.g., name a letter or word correctly
80% of the time in order to conclude that the targeted learning has
occurred.
2. There are several models for framing these intended outcomes or standards;
we briefly examine two and then integrate these two approaches into one,
guided by Oosterhof’s (1994, pp. 43-47) suggestions.
a. Mitchell (1996) offers the following Taxonomy and Definitions
(1) Content (or academic) standards. These standards identify what
knowledge and skills are expected of learners at specified phases in
their educational progression.
(2) Performance standards. Performance standards have levels, e.g., 4, 3,
2, 1; exceeds expectations, meets expectations, or does not meet
expectations; unsatisfactory, progressing, proficient, or exemplary,
which are intended to show degree of content mastery.
(3) Opportunity to learn standards. There are instances where enabling
learning and/or performance conditions are identified to ensure that

learners have a fair chance to meet Content and Performance

Standards.
b. The State of Florida (1996, pp. 28-30) developed the following taxonomy:
(1) A Strand is a label (word or phrase) for a category of knowledge;
consider it to be a general content (knowledge and/or skill) organizer.
(2) A standard is a general statement of expected learner achievement.
(3) A Benchmark describes in, in a series of statements more precisely
what a learner is expected to know and/or be able to do at the end of a
training or educational unit or program. When all associated
benchmarks are accomplished, the standard is met.
3. The Integrated Model

a. A standard is typically composed of five elements.
(1) The first element states “who is to do something”, usually a student.
(2) The second element is an action oriented verb (e.g., articulate,
describe, identify, explain, analyze, etc.); it is at the verb level that
intellectual skill taxonomies (e.g., Bloom) exert their influence. For
example, Quellmalz outlined five cognitive functioning levels
(Bloom’s equivalencies are noted), which are:
Quellmalz (1987) Bloom, et al. (1956)

Recall Knowledge & Comprehension
Analysis Analysis
Comparison Synthesis
Inference* Application & Synthesis
Evaluation Synthesis & Evaluation
*(deductive & inductive reasoning)
For instance, according to Stiggins, Griswold, and Wikelund (1989),

verbs associated with inference include generalize, hypothesize, or
predict. Within Quellmalz’s “comparison level,” one could use the
words compare and contrast. We can examine intellectual skills
taxonomy to find the most precise verb to describe what we want an
examinee to know, do, or feel. There are other intellectual skill
taxonomies such as Canelos (2000) and Gange’s (1985); others are
presented in Chapter 1. The key point is to identify a useful one and
use it to guide learning standard, target, or objective writing.
(3) The third element describes under what condition(s) the student is to
demonstrate something (e.g., fully, briefly, clearly, concisely,
correctly, accurately, etc.).
(4) The fourth element specifies what the student is to demonstrate (e.g.,
algebra calculation, leadership theories, decision-making models, etc.)
(5) The fifth element (optional) describes the medium in which the
demonstration is to take place, e.g., in written or oral form, via
examination, case report, or portfolio, etc. We don’t recommend using

the fifth element as assessment options may become too limited. Two
sample standards are:
(a) The student will accurately compute algebraic equations.
(b) The student will accurately and concisely describe modern
leadership theories.
b. Construct an Operational Definition, called a benchmark for each Attitude,

Content, and/or Skill Standard
(1) Standard operational definitions are constructed through benchmark
statements. A benchmark, in plain English, is a specific action
oriented statement which requires the examinee to do something.
When a student has achieved each relevant benchmarks, the standard
is met. Benchmarks further define the relevant attitude, content, and
skill domains. Illustrative benchmarks are:
(a) The student will accurately compute algebraic equations.
[1] The student will correctly compute linear equations.
[2] The student will correctly compute quadratic equations.
[3] The student will correctly compute logarithms.
(b) The student will accurately and concisely describe modern
leadership theories.
[1] Describe content and process motivation theories citing
education leadership examples.
[2] Describe leadership trait theories of leadership, respecting
assumptions, elements, research, and education application.
[3] Describe leadership styles and situational models in terms of
assumptions, characteristics, and education application.
(2) Benchmarks are very useful in framing examinations, assignments,
etc. Test items are written based on specific benchmarks. As with
standards, subject matter experts (SME’s) should agree on the
appropriateness of each benchmark and its association with its specific
standard. Further, it is not usually necessary to stipulate a
performance level as in percent correct; the school’s grading scale or
the organization’s performance assessment system should suffice.
c. Once a sufficient number of standards, in depth and breadth, have been

drafted, units, courses, degree programs, etc. are then constructed around
relevant “bundles” of standards. Such “bundles” of relevant attitude,
content, and skill standards define what is to be taught (i.e. the curriculum-
- also called a domain), learned, and tested.
(1) Attitude standards can be assessed by surveys; see Chapter 4.
(2) Content standard mastery can be studied through examinations which
require the student to demonstrate his or her knowledge.
(3) Skills can be tested by examinations using application oriented item
formats (e.g., brief and expended responses or solving problems),
projects, portfolios, or cases, etc.

d. The five part model presented here can also be used to evaluate standards
written by others. Regardless of the approach employed, each standard
should meet at least the first four components, be developmentally
appropriate, and ensure that students have had the opportunity to learn,
develop the desired attitude, and/or acquire the specified skill(s).
B. Constructing a Test Blueprint

1. A table of specifications sorts the performance standards, i.e., content and/or
skills, by intellectual skills to be performed. The number of test items is
inserted in the appropriate table cell. An example is presented Table 5.1,
where the number of test items per learning benchmark and intellectual skill is
presented. An alternative is to list the actual test item number instead of just
the number of test items.
2. The more important the content, the greater the number of test items. To
arrive at these numbers, the test developer will typically
a. Determine the total number of items to be included. This number is
influenced by the testing time available, the test environment,
developmental status of the examinees, and size of the content and/or skill
domain to be tested.
b. Test items are allocated to each learning target, objective, or outcome;
more critical standards are allotted more items.
3. Next, test items are sorted across each learning target or benchmark, by
intellectual skill. Test item classification guidelines are:
a. Determine exactly what the action verb in the Learning Target benchmark
asks or requires of the examinee.
b. Ensure the test item asks or requires the examinee to know or do what is
specified in the benchmark. The test item and benchmark must match.
c. Examine the benchmark’s action verb, to classify the item.
(1) If the examinee is to just recall or know data or information, classify
under “Knowledge,” which typically is declarative knowledge.
(2) If the examinee is to recall or identify data or information in a different
manner than he or she first learned, classify under “Comprehension,”
which typically is declarative knowledge.
(3) If the examinee is to apply (i.e., use) the knowledge or skill learned,
then classify under “Application,” which typically requires procedural
and/or conditional knowledge.
(4) If the examinee is to analyze (i.e., take apart to thoroughly examine)
an idea, recommendation, poem, report, argument, or an interpretation
of a dataset or what someone else produced, classify under “Analysis.”
(5) If the examinee is to write a poem, article, term paper, play or perform
an artistic routine, or paint a painting, or prepare a photography
exhibit, classify under “Synthesis.” Of all of Bloom’s levels, creativity
is most associated with “Synthesis,” followed by “Analysis.” An
example is the construction of a term paper where “Analysis” is

demonstrated by reading articles on a topic and then assembling the

learning from the articles into a coherent paper (Synthesis).
d. Carefully select the best action verb which requires the examinee to know
the content, perform the skill or display the attitude intended. There are
many sources from which to select the best action verb, e.g., the Internet,
colleagues, or the synonym list found in most word processing programs.
e. If there is a question about how to classify an item, consult an experienced
colleague. Sometimes there is item classification disagreement, due to
differing perspectives; try to reach consensus. The test designer usually
makes the final classification decision.
4. Application: Table 5.1

a. Presented in Table 5.1 is a test blue print for an end-of-term test in grades
3-5 language arts. The performance standards are benchmarks drawn from
Standard 2 of the 1996 Florida Sunshine State standards for Language
Arts. The benchmarks are presented in the far-left column and the
intellectual skills are presented in columns to the right. The numbers in
table cells are the number of items per intellectual skills.
b. Upon reviewing the table, the reader sees that this instructor selected
(LA.A.1.2.2) as more critical than the others.
c. For (LA.A.1.2.2), there are 8 knowledge items, which require simple
recall of information. Knowledge of phonetic sounds and word structure
is critical if a student is to be able to apply this knowledge to construct
meaning from various texts. The types of knowledge to be demonstrated
drive which item formats are selected. Possible item formats are simple
multiple choice, matching, fill-in-the-blank, or true/false.
d. Next, the examinee should be able to demonstrate a comprehension of
what is read in the form of retelling and self-questioning. Comprehension
involves translation, interpretation, and/or extrapolation. In the present
instance, we would focus on interpretation and extrapolation. Possible
item formats are multiple choice, short response, or extended response
items.
e. The test developer elected to test at the application level. To test
application, the test developer can craft one table and one graph. Several
multiple choice questions to assess an examinee’s ability to comprehend
and apply data from the table and graph can be constructed. An alternative
is to have the student write several sentences as to the meaning and
potential use of the data presented by table and graph.
f. Students will also need to selected texts or illustrations; he or she could be
asked to construct a statement explaining the texts and predicting a
potential outcome from that meaning. Answers could be selected from
multiple choice items and/or supplied as in a written paragraph or two.
Since there is only two hours for the examination and more time is
required to demonstrate higher order intellectual skills (i.e., analysis,
synthesis, and evaluation), there are fewer items included on the test.

Table 5.1
Test Blue Print for End of Term Test on Language Arts Strand A: Reading
Standard: 2. The student writes uses the reading process effectively, Grades 3-5
Benchmark Knowledge Comprehension Application Analysis Synthesis Total
The student will… 5a,b 6d,e 1f 12
Uses a table of contents, index, headings, captions,
illustrations, and major words to anticipate or predict
content and purpose of a reading selection. (LA.A.1.2.1)
Selects from a variety of simple strategies, including the use

of phonetics, word structure, context clues, self-questioning, 8a,d 4c 2f 1g 15
confirming simple predictions, retelling, and using visual
cues to identify words and construct meaning from various
texts, illustrations, graphics, and charts.
(LA.A.1.2.2)
Uses simple strategies to determine meaning and increase

vocabulary for reading, including the use of prefixes,
suffixes, root words, multiple meanings, antonyms, 6a,c 3a 3e 12
synonyms, and word relationships. (LA.A.1.2.3)
Clarifies understanding by rereading, self-correction,

summarizing, checking other sources, and class or group 5f 5
discussions. (LA.A.1.2.4)
Item Totals 19 13 11 1 44
Note: The source of the example is: Florida Curriculum Framework: Language Arts PreK-12 Sunshine State Standards and Instructional
Practice (Florida Department of Education, 1996, pp. 36-37). It is increasingly common to insert the actual test item numbers once the test has
been finalized as a quality control strategy. Item Formats: amultiple choice, btrue/false, cmatching, dfill-in-the blank, ecompletion, fbrief
response, gextended response.

D. Directions and Arranging Test Items

1. Directions
a. Directions should be clear, concise, and to the point.
b. Directions should include: how to record answers; the time available; the
points associated with specific subtests or items; what to do when the test
is completed; and explain the permissible use of “scratch” paper to show
computations, if allowed.
c. Keep the directions and associated items on the same page, even if the
directions need to be repeated on subsequent page(s).
2. Arranging Items on the Test Instrument

a. Include an “Ice breaker” item which virtually all examinees will answer
correctly to build confidence. Since the purpose of test items is to identify
those who can answer the item correctly, not all examinees should be able
to answer every item on the test.
b. Group similar content together on the testing instrument. Use items
designed to measure important content, as testing time is almost always
limited.
c. Use items of appropriate known difficulty levels (see Part IV of this
chapter) when possible.
d. Don’t break items across a page.
e. Keep charts or figures, pertaining to an item, on the same or next page.
f. If using a computer printer, use a consistent font or font size.
g. If students are required to supply an answer, provide enough room for the
answer.
E. Timing and Testing

1. Testing time limits are useful because:
a. Examinees complete the test more rapidly.
b. Examinees learn to pace themselves.
c. Examinees may be more motivated.
d. A time limit can be selected so that most examinees will complete the test.
2. For speeded tests, an examinee’s score is materially influenced by the number

of items answered correctly within a specified time limit. Thus, item order
and format are important. Care must be taken to ensure that item order and
format are consistent with the purpose and cognitive (e.g., intellectual or
thinning) skill level(s) to be measured. Before time limits are set, there should
be pilot-testing to ensure that the time limit is sufficient to accomplish the
purpose of the test. Speeded testing is used in emergency skills testing, crisis
management training and other circumstances where responding quickly and
correctly in the shortest or least amount of time is essential.

3. For power tests (i.e. achievement testing), an examinee’s score is influenced

by item format and item difficulty as well as the cognitive skill(s) to be
assessed. If used, the set time limit should be sufficient so that virtually all
examinees will complete the test. Recommended time limits are:
a. Simple true/false or matching: 15-20 seconds for each item.
b. Complex true/false or matching and one or two word completion: 20-30
seconds for each item.
c. Simple multiple choice: 25-30 seconds for each item.
d. Complex multiple choice: 45-60 seconds for each item.
e. Brief or restricted response essay: 10 minutes for each item.
f. Remember, the time required for examinees to answer test items is
influenced by reading skill, item format, item length, and cognitive skill(s)
being tested. Items which require computations will take longer. The
above time limits are intended only to serve as a guide.
F. Interpreting Test Scores

1. Test items are proxy measures of an unobservable psychological construct or
trait; declarative, procedural or conditional knowledge; and/or psychomotor
skill. Test items also require examinees to use intellectual and/or thinking
skills (see Chapter 1). Measuring psychological attributes, such as ability or
achievement is often not directly possible; that is why test items are written,
so that inferences can be made from the examinee’s behavior (answering the
test item or performing the skill as requested).
a. Test items consist of a stimulus which is intended to prompt a prescribed
or expected answer. A correct answer or endorsement to a test item is seen
as an indicator that the examinee has the attribute, knows the knowledge,
or can perform skill taught.
b. Test item formats include: multiple choice, true/false, completion, fill-in-
the blank, short answer, check-list, and essay. Each item type has its
strengths and weaknesses and should be selected carefully. Guidelines for
writing each type of test item are provided in Part III of this chapter.
2. There are three types of test items; the key difference is the test item’s
purpose.
a. Mastery items measure essential minimums that all examinees should
know; measure memorization of facts or simple computations (e.g.,
Bloom, et al.’s knowledge or comprehension skill levels); and are
commonly used in licensing or certification tests.
b. Power items are designed to measure what typical examinees are expected
to know; these items may range in difficulty form very easy to very hard,
depending on the test purpose and the content, skills, or attitudes being
measured. Power items are commonly found in achievement tests.
c. Speed items are designed to assess higher level concepts and skills (e.g.,
Bloom, et al.’s application, analysis, synthesis, or evaluation) expected of
the most able examinees. Speed items are found on tests of mental abilities
(e.g., IQ tests). Don’t confuse these item types with speeded tests where

the strictly observed time limit is part of the measurement process (e.g.,
decision-making accuracy in an emergency situation by a paramedic,
firefighter, or police officer).
3. Test items must have two (2) very important characteristics: test items are
unidimensional and logically independent
a. Items are Unidimensional
(1) Each test Item must measure one attribute, e.g., knowledge, attitude, or
psychomotor skill, etc.
(2) If it were possible, in theory, to write every possible test item for an
attribute (e.g., knowledge, skill, or attitude), they would fully describe
every single aspect of that attribute. There would be too many test
items; so, we write test items to measure the most important
attribute(s) of the content, skill or attitude of interest.
(3) This unidimensional assumption is critical for two reasons: (a) it
permits a more accurate interpretation of a test item and (b) allows the
explanation that a single trait (e.g., knowledge, skill, or attitude)
accounts for an examinee responding correctly to the test item.
b. Items are logically independent
(1) An examinee’s response to any test item is independent of his or her
response to any other test item. Responses to all test items are
unrelated, i.e., statistically independent.
(2) Practical implications for writing test items are: (a) write items so that
one item doesn’t give clues to the correct answer to another item and
(b) if several items are related, e.g., to a graphic (picture, table or
graph), they should be written so as not to “betray” one another but to
test different aspects of the graphic. See Interpretative Exercise in the
discussion on writing multiple choice items and Appendix 5.1.
4. Test items are either dichotomously or polychotomously scored.

a. Dichotomously scored test items are scored as correct or incorrect, right or
wrong, etc. This scoring applies to multiple-choice; true/false; matching;
completion; and fill-in-the-blank item formats. These items are usually
worth at most a few points each.
b. In polychotomously scored items, there is more than one correct response
for a test item or partial credit, i.e., differing point awards are given (e.g.,
7 of 10 possible points). This scoring is typically applied to
brief/restricted or extended response items, which are usually weighted
with more points than select response items. A scoring rubric is used to
guide point awards.

5. Allocating Test Points

a. For a traditional classroom achievement test, usually given at the end of a
unit or module, once the test blue print and item formats are selected,
points are allocated. Points are necessary to compute, report, and interpret
test scores, as well as any subsequent statistical analysis.
b. One must know how scores are to be interpreted to build a scoring plan
that will permit the desired interpretation which is usually either pass/fail
or performance segmented as in a grading scale (e.g., “A,” “B,” “C,” etc.).
In short, the examiner must know how he or she wants to interpret the data
(test score) before the test is constructed.
c. The most critical knowledge or skills are weighted with the most points,
followed by less critical knowledge and skills, receiving fewer points.
(1) Let’s look at an example drawn from the test blueprint presented in
Table 5.1. In this example, we will consider each standard as a
separate subtest.
(2) Assigning Points to Items (Table 5.2)
(a) Brief (or restricted) and extended response items were weighted at
5 and 8 points respectively due to the higher difficulty level and
the higher order intellectual skills required to correctly answer the
item. Since partial credit is awarded for these items, a scoring
rubric to guide awarding points is required.
(b) Multiple choice and completion items were weighted at 3 points
each as more is the required of the examinee to answer the item
correctly than let’s say a true/false item.
(c) Fill-in-the blank and matching items were scored at 2 points each
as the context of the sentence or clue may trigger examinee
memory, thus making it easier for the examinee to recall the
correct answer.
(d) True/false items were weighted at one point each as the examinee
is expected to simply recall the correct information.
(3) There are 4 subtests for a total of 129 points. The examiners weighted
Subtest 3 (Standard L.A.A.1.2.2) at 46 points as they thought it
assessed more important skills than the other three.
(4) For interpretative purposes, we can divide earned points by possible
points to arrive at a percentage level of mastery (for tests not totaling
exactly 100 points) on each subtest and the entire unit for the examine
group. Let’s suppose the class scored as shown in Table 5.3, using the
class average scores on each subtest and the total test score.
(a) If 80% was the passing or “cut” score, the class as a whole passed
the unit examination. However, group/class remediation appears
necessary on the Subtest A (Learning Standard L.A.A.1.2.1)
content, since the pass rate was 72% (18/25).
(b) Individual examinee performance can be reviewed and interpreted
in the same manner. First, the total test score is examined. Second,
each subtest score is examined; if the total test score and individual
subtest scores are acceptable, then no further investigation is

warranted. However, if ether the total test or any subtest score is

below the passing or “cut” score, then individual items need to be
reviewed to determine what re-teaching or remediation is needed
for each particular examinee.
Table 5.2
Language Arts Unit 3 Test Scoring Plan
Standard Item Format Item Points Subtest Points
L.A.A.1.2.1 Multiple Choice 3 x 2 items = 6
(Subtest 1) True/False 1 x 3 items = 3
Fill-in-Blank 2 x 3 items = 2
Completion 3 x 3 items = 9
Brief Response 5 x 1 item = 5
Total = 25 points
(Subtest 2) Matching 2 x 4 items = 8
Fill-in-Blank 2 x 4 items = 8
Brief Response 5 x 2 items = 10
Extended Response 8 x 1 item = 8
Total = 46 points
(Subtest 3) Matching 2 x 3 items = 6
Completion 3 x 3 items = 9
Total = 33 points
L.A.A.1.2.4 Brief Response 5 x 5 items = 25
(Subtest 4) Total =25 points
Table 5.3
Language Arts Unit 3 Subtest and Total Test Scores
Subtest x Points Earned Points Possible Percent Mastery
A (L.A.A.1.2.1) 18 25 72%
B (L.A.A.1.2.2) 39 46 85%
C (L.A.A.1.2.3) 28 33 85%
B (L.A.A.1.2.4) 20 25 80%
Total Test Score 105 129 81%
d. For a direct performance assessment, each assignment has a task

description and scoring rubric so students clearly understand what is
expected of them.
(1) An accurate assignment score or grade assumes a clearly understood
task description and consistent application of the scoring rubric for
each assignment.
(2) An assignment score or grade should not be based on anything other
than the examinee’s performance using the task description and
scoring rubric. Don’t include participation, effort, or attendance.
(3) Points are allocated in the same manner as a traditional classroom test
with the most critical information receiving the highest point
allocation and less import receiving fewer points. See Appendices 5.2,
5.4 and 5.5, which contain task descriptions and scoring rubrics.
6. Assigning a Course or Training Program Grade

a. A course, consisting of several units or modules with multiple assignments
and/or tests, may present points possible in a table, such as Table 5.4.
(1) The 2 performance tasks are weighted the most points as they require
the demonstration of higher order intellectual skills, skillful
application of the content, and procedural knowledge.
(2) Discussion prompt answers are a maximum of 500 words each
requiring references; they must evidence at least one higher order
intellectual skill. Answers are graded with an analytical rubric, which
is discussed in Part V of this chapter.
(3) Students are required to comment on other student discussion prompt
answers or comments. Comments are scored holistically.
(4) The mid-term and final exam assesses knowledge, comprehension, and
limited application of critical course knowledge and skills.
b. Points possible are then usually “segmented,” based on point levels or
percentages, into performance bands, with qualitative labels; see Table
5.5. Be clear and consistent on the “rounding” rules, 85.8% becomes 86%.
Table 5.4
Course Assignment & Test Point Weights
Assignment Points
Traditional Classroom Test Construction Task 224
Direct Performance Assessment Construction Task 160
Discussion Prompt Answers (8 x 12 points x 2) 192
Discussion Answer Comments (16 x 4 points x 1.5) 96
Mid-term and Final (2 @ 75 points each) 150
Total Points 822
c. Combing the information from Tables 5.4 and 5.5, a student earning 739
points is assumed to have “mastered” 90% (actually 0.899%) of the
content (specified by the learning outcomes or targets) for a performance
indicator of an “A-” or an “Excellent” performance.
Table 5.5
Grading Scale
Grade Percentage Meaning
A 95-100% Exceptional
A- 90-94% Excellent
B+ 87-89% Very Good
B 83-86% Good
B- 80-82% Fair
C 75-79% Marginal
F ≤75% Failure
Note. Saint Leo University (2010). Graduate academic
catalog 2010-2011. St. Leo, FL: Author.

II. Constructing Select Response Items to Assess Knowledge, Comprehension, &

Application
A. General Item Writing Guidelines
1. Make the difficulty and complexity of the item appropriate to the examinee’s
level. Ensure that the item reading level is appropriate for examinees.
2. Define the item as clearly as possible.
3. Write simple, straightforward items. Use the most precise words to
communicate the intent of the item.
4. Know the students’ mental processes and frame the item accordingly.
5. Vary item complexity and difficulty.
6. Make items as independent from each other, except when used in a series.
7. Avoid negatively phrased items; avoid double negatives.
8. Use correct grammar, syntax, spelling, etc.
9. When writing specific item formats, review the unique item construction
guideline for each format to ensure conformity.
10. Ensure that items do not contain language which is racially, ethnically,
religiously, or economically biased.
11. Select response items usually measure Bloom’s “Knowledge” and
“Comprehension” levels. Supply response items measure the higher
intellectual skills, “Application,” “Analysis,” “Synthesis,” and “Evaluation.”
12. Avoid clues that Airasian (1997, p. 177) calls specific determiners. These are
words which tend to “give away” the correct answer. For example, words such
as “always”, “all”, or “none” tend to be associated with false “true/false”
items.
B. Writing Multiple-Choice Items

1. Each multiple-choice item has a stem which presents the situation, problem,
definition, etc. Then there are answer choices (also called options or foils)
which include the correct answer and plausible wrong answers, called
distracters or foils. There are several multiple choice item variations:
a. Correct Answer: Only one answer is correct.
(1) Correct Answer
___In times of war in the early Roman Republic, the two consuls stepped aside in
favor of one person who would be responsible for making decisions quickly. That
person was a —
a. General c. Dictator
b. Czar d. Tyrant
(2) Correct Answer

___What is 68 x 22 = _____
a. 149 c. 1,496
b. 1,469 d. 4,196
(3) Answers: “c” & “c”

b. Best Answer: Examinees select the “best” response option that “fits” the
situation presented in the item stem.
(1) Best Answer
___This culture developed an accurate calendar. They built steep temple pyramids
and used advanced agricultural techniques. They developed a system of mathematics
that included the concept of zero. They were located mostly in the Yucatán
Peninsula. They ruled large cities based in southern and southeastern Mexico, as well
as in the Central American highlands. This passage best describes the ancient.
a. Aztecs c. Mayas
b. Incas d. Olmecs
(2) Answer: “c”
c. Rearrangement: This item is can be used to assess examinee procedural

knowledge or comprehension.
(1) Rearrangement
There are several preconditions which must be satisfied before a test is
constructed. Read each statement listed below and place each in the correct order
of operation, by using the codes presented below.
a. First Step c. Third Step
b. Second Step d. Fourth Step
____1. The intellectual skills, attitudes, and/or psychomotor are identified.

____2. The test item formats are considered and selected.
____3. A table of specifications (or test blueprint) is developed.
____4. Items are crafted.
(2) Answer: 1. a; 2. b; 3. c; 4. d
d. Substitution Item: In this variation, the item stem contains a blank or

blanks. The examinee then selects from the response options, either the
correct or best response for the stem; the purpose is usually intended to
make the stem a correct or true statement. The item is widely used to
assess examinee comprehension.
(1) Directions. Please the letter which represents the word which correctly completes
the statement in the corresponding blank.
Reliability and Validity Relationship

a. Concurrent c. Content e. Internal Consistency
b. Construct d. Test-retest f. Alternate Forms
At a minimum all tests must have (a) _________validity and (b) ____ reliability.
(2) Answers: 1c; 2e
e. Analogy
(1) A multiple choice item can be written in the form of an analogy.
(2) Hieroglyphics is to Egypt as cuneiform is to —
a. Phoenicia c. Persia
b. Sumer d. Crete
(3) Answer: b

2. Strengths and limitations of multiple choice items are:

a. Strengths
(1) Simple and/or complex learning outcomes are measured.
(2) Highly structured and clear tasks are provided.
(3) These items are highly efficient in assessing substantial content
domains quickly.
(4) All responses, even incorrect endorsements, provide useful item
revision and examinee performance information.
(5) Items are less susceptible to guessing than matching or true/false
items.
. (6) Scoring is simple, quick, and objective.
b. Limitations
(1) Constructing high quality multiple choice items is labor and time
intensive.
(2) Writing plausible foils is often difficult.
(3) Writing items to measure higher order intellectual skills is very
difficult and when successfully done provide only a proxy measure of
such skills.
(4) Like most test items, the multiple choice format can be influenced by
reading skills.
3. A review of the test item design literature (Oosterhof, 1994, pp. 33-147;
Gronlund, 1998, pp. 60-74; Popham, 2000, pp. 242-250) and the authors’
reveal these item writing guidelines.
a. Learning outcomes should drive all item construction. There should be at
least two items per key learning outcome.
b. In the item stem, present a clear stimulus or problem, using language that
an intended examinee would understand. Each item should address only
one central issue.
c. Generally, write the item stem in positive language, but ensure to the
extent possible, that the bulk of the item wording is in the stem.
d. Underline or bold negative words whenever used in an item stem.
e. Ensure that the intended correct answer is correct and that distractors or
foils are plausible to examinees who have not mastered the content.
f. Ensure that the items are grammatically correct and that answer options
are grammatically parallel with the stem and each other.
g. With respect to correct answers, vary length and position in the answer
option array. Ensure that there is only one correct answer per item.
h. Don’t use “all of the above,” and use “none of the above” unless there is
no other option.
i. Vary item difficulty by writing stems of varying levels of complexity or
adjusting the attractiveness of distractors.
j. Ensure that each item stands on its own, unless part of a scenario where
several items are grouped. Don’t use this item format to measure opinions.
k. Ensure that the arrangements of response options are not illogical or
confusing and that they are not interdependent and overlapping.

l. Avoid clues that enable examinees to eliminate incorrect alternatives or

select the correct answer without really knowing the content. Common
clues include:
(1) Restating the text or lecture note answer verbatim or nearly verbatim.
(2) Using implausible distractors.
(3) Writing the correct answer so that it is longer than other plausible
distractors.
(4) Writing distractors which are so inclusive that an examinee is able to
exclude other distractors or distractors which have the same meaning.
(5) Absolutes such as “never”, “always”, etc. are associated with false
statements. Their use is a clue to examinees that the alternative is
most likely incorrect.
4. Specific Applications
a. Best Answer Items: In this scenario, the examinee is required to select the
best response from the options presented. These items are intended to test
evaluation skills (i.e., make relative judgments given the scenario
presented in the stem).
b. To assess complex processes, use pictorials that the examinee must
explain.
c. Use analogies to measure relationships (analysis or synthesis).
d. Present a scenario and ask examinees to identify assumptions (analysis).
e. Present an evaluative situation and ask examinees to analyze applied
criteria (evaluation).
f. Construct a scenario where examinees select examples of principles or
concepts. These items may either measure analysis or synthesis,
depending on the scenario.
5. The Interpretive Exercise (IE)

a. A preferred item format for assessing application, analysis, synthesis, and
evaluation is the Interpretive Exercise. See Appendix 5.1.
(1) In IE, based on “introductory information”, several related items are
used to assess mastery of a particular intellectual skill expression.
(2) This assessment strategy is used to assess other complex skills such as
reading ability, comprehension, mathematical thinking, problem
solving, writing skills, graph and table interpretation, etc.
b. IE can be used to assess whether or not examinees can
(1) Recognize and state inferences;
(2) Distinguish between warranted/unwarranted generalizations;
(3) Identify and formulate tenable hypotheses;
(4) Identify and interpret relationships;
(5) Recognize assumptions underlying conclusions;
(6) Formulate valid conclusions and recognize invalid ones;
(7) Recognize relevance of information; and
(8) Apply and/or order principles.

c. Advantages
(1) Interpretive skills are critical in life.
(2) IE can measure more complex learning than any single item.
(3) As a related series of items, IE can tap greater skills depth and breadth.
(4) The introductory material, display or scenario can provide necessary
background information.
(5) IE measures specific mental processes and can be scored objectively.
d. Limitations
(1) IE is labor and time intensive as well as difficult to construct.
(2) The introductory material which forms the basis for the exercise is
difficult to locate and when it is, reworking for clarity, precision, and
brevity is often required.
(3) Solid reading skills are required. Examinees that lack sufficient
reading skills may perform poorly due to limited reading skills.
(4) IE is a widely used proxy measure of higher order intellectual skills.
For example, IE can be used to assess the elements of problem solving
skills, but the extent to which the discrete skills are integrated.
e. Construction Guidelines
(1) Begin with written, verbal, tabular, or graphic (e.g., charts, graphs,
maps, or pictures) introductory material which serves as the basis for
the exercise. When writing, selecting, or revising introductory
material, keep the material:
(a) Relevant to learning objective(s) and the intellectual skill being
assessed;
(b) Appropriate for the examinees’ development, knowledge and skill
level, and academic or professional experience;
(c) At a simple reading level, avoiding complex words or sentence
structures, etc;
(d) Brief, as brief introductory material minimizes the influence of
reading ability on testing;
(e) Complete, contains all the information needed to answer items; and
(f) Clear, concise, and focused on the IE’s purpose.
(2) Ensure that items, usually multiple choice:
(a) Require application of the relevant intellectual skill (e.g.,
application, analysis, synthesis, and evaluation);
(b) Don’t require answers readily available from the introductory
material or that can be answered correctly without the introductory
material;
(c) Are of sufficient number to be either proportional to or longer than
the introductory material; and
(d) Comply with item writing guidelines;
(e) Revise content, as is necessary, when developing items.
(3) When examinee correctly answers an item, it should be due to the fact
that he or she has mastered the intellectual skill needed to correctly
answer rather than failure to memorize background information. For

example, in statistics, give the examinee the formulas so that concepts

are tested and not formula memorization.
(4) Other item formats are often included in the interpretative exercise.
C. Writing True/False Items

1. True/false items are commonly used to assess whether or not an examinee can
identify a correct or incorrect statement of fact. Tests employing true/false
items should contain enough items so that the relevant domain is adequately
sampled and that the examiner can draw sound conclusions about examinee
knowledge. True and false statements should be equally represented.
2. Strengths and limitations of true/false items (Gronlund 1998, p. 79; Oosterhof,

1994, pp. 155-158; & Popham, 2000, pp. 232-233) are:
a. Strengths
(1) These items enable an adequate sampling from the content domain.
(2) They are relatively easy to construct.
(3) They can be efficiently and objectively scored.
(4) Reading ability exerts less of an influence than in multiple choice
items.
b. Limitations
(1) These items are susceptible to guessing. There is a 50% chance of an
examinee correctly guessing the correct answer given a true/false item.
(2) Such items are only able to measure higher order skills indirectly.
(3) These items can only be employed when dichotomous alternatives
sufficiently represent reasonable responses.
(4) Use of true/false items tends to encourage examinees to memorize
facts, etc. as opposed to learning the content. If the item is not
carefully constructed, student memorization “pays off.”
(5) True/false items provide no item diagnostic information as is found in
multiple choice items by selecting the wrong answer.
3. True/false item writing guidelines (Gronlund 1998, pp. 80-84; Oosterhof,

1994, pp. 156-164; & Popham, 2000, pp. 235-238) are:
a. For each statement, incorporate only one theme. When preparing an item,
write it as both correct and incorrect. Select the better of the two options.
b. Write crisp, clear, and grammatically correct statements. Brief is
recommended. However, use terms that would suggest an incorrect
response upon a quick reading by an examinee.
c. Select precise wording for the item so that it is not ambiguous. The item
should be either absolutely true or false; avoid statements that are partly
true and partly false. If there are adverb or adjectives which are key to
marking a statement correctly, underline or otherwise stress that word.
d. Avoid double negatives. Negatively worded statements tend to confuse
examinees so use them rarely.
e. Attribute opinion statements to their source. However, if testing
examinees’ ability to distinguish fact from opinion, don’t attribute.

f. If measuring cause and effect, use true statements.

g. Avoid absolute relative terms (e.g., always or never) usually adjective or
adverbs, which tend to clue the examinees to the correct response.
4. True/False Item Variations

a. True/false Items Requiring Corrections
(1) This variation requires the examinee to either correct a false item or to
identify the false portion of an item.
(2) This is actually a blend of true/false and short answer.
(3) Directions. Read the statement carefully. Determine whether or not each
underlined number is true or false. If you think the statement is true, circle “T” or
if you think the statement is false, circle “F.” If false, place the correct number in
the line provided.
Acceptable reliability standards for measures exist. For instruments where groups
are concerned, 0.80a or higher is adequate. For decisions about individuals, 0.85b is
the bare minimum 0.95c is the desired standard.
a. T F ______
b. T F ______
c. T F ______
Answer: a. T; b. F 0.90; c. T
b. Embedded Items
(1) In a paragraph are included underlined words or word groupings.
Examinees are asked to determine whether or not the underlined
content possesses a specified quality, e.g., being true, correct, etc.
(2) Embedded items are useful for assessing declarative or procedural
knowledge.
(3) Directions. Indicate whether the underlined word is correct within the context of
threats to a measure’s reliability. Circle “1” for correct or “2” for incorrect.
When a test is given to a very similar homogeneousa (1 or 2) group, the resulting

scores are closely clustered and the reliability coefficient will be highb (1 or 2).
Answer: a. 1 and b. 2.
c. Multiple True-False Items

(1) This variation is a blend of the true/false and multiple choice item.
Multiple statements, whose truth or falsity is being tested share a
common stem. Each statement is numbered as a unique test item.
(2) True/false item writing rules apply, but note that statements sharing a
common stem are usually narrower in focus than conventional
true/false items.

(3) Directions. Read each option and if you think the option is true, circle “T” or if you
think the statement is false, circle “F.”
The standard error of measurement is useful for

T F 1. Reporting an individual’s scores within a band of the score range
T F 2. Converting individual raw scores to percentile ranks
T F 3. Reporting a groups’ average score
T F 4. Comparing group differences
Answers: 1.T; 2.F; 3.F; 4.F
d. Focused Alternative-Choice Items

(1) While conceptually similar to true/false items, examinees are required
to select between two words or values which correctly complete a
statement. The words or values must have opposite or reciprocal
meanings.
(2) When compared to conventional true/false items, focused alternative-
choice items typically produce more reliable test scores.
(3) Directions. Read each statement carefully. Circle the letter that represents the word
or phrase which makes the statement correct.
A B 1. A measure’s reliability coefficient will likely (a. Increase or

b. Decrease) as the numbers of items are lengthened.
A B 2. An examinee’s answer to a restricted response essay should typically
not exceed (a. 150 or b. 250) words.
Answers: 1. a; 2.a
e. Standard Format
(1) Traditionally, a true/false item has been written as a simple declarative
sentence which was either correct or incorrect.
(2) Directions. Read the statement carefully. If you think the statement is true, circle
“T” or if you think the statement is false, circle “F.”
T F A measure can be reliable without being valid.
Answer: T
D. Writing Matching Items

1. The matching item is a variation of the multiple choice item format.
Examinees are directed to match (i.e., associate) items from two lists.
Traditionally, the premises (the portion of the item for which a match is
sought) is listed on the left hand side of the page. On the right hand side of
the page are listed the responses (the portion of the item which supplies the
associated elements).
a. The key to using this item format is the logical relationship between the
elements to be associated.
b. Elements in each column are homogeneous, i.e., related to the same topic
or content domain.

c. There are usually more responses (right hand column) than premises. This
reduces the biasing impact of guessing on the test score.
d. Example
Directions. Match the terms presented in the right column to their definitions which are
presented in the left column. Write the letter which represents your choice of definition
in the blank provided.
Definition Term
_____ 1. Measures essential minimums that all a. Mastery Items
examinees should know
_____ 2. Designed to assess higher level concepts & skills b. Power Items
_____ 3. Designed to measure what typical examinees c. Speed Items

are expected to know
_____4. Designed to measure achievement where the d. Independent

content is evolving Variable Items
e. Dependent Items
Answers: 1.a; 2.b; 3.b; 4.d
2. Strengths and limitations of matching items (Gronlund, 1998, p. 85-86;

Kubiszyn & Borich, (1996, p. 99; Oosterhof, 1994, p. 148; & Popham, 2000,
pp. 238-240) are:
a. Strengths
(1) An efficient and effective format for testing examinee’s knowledge of
basic factual associations within a content domain.
(2) Fairly easy to construct. Scoring is fast, objective, and reliable.
b. Limitations
(1) This format should not be used to assess other cogitative skills beyond
mere factual associations which emphasize memorization.
(2) Sets of either premises or responses are susceptible to clues which help
under-prepared examinees guess correctly which gives rise to false
inferences respecting examinee content mastery.
3. Item writing guidelines (Gronlund, 1998, p. 86-87 & Popham, 2000, pp. 240-
241) are:
(a) Employ only homogeneous content for a set of premises and responses.
(b) Keep each list reasonably short, but ensure that each is complete, given the
table of specifications. Seven premises and 10-12 options should be the
maximum for each set of matching items.
(c) To reduce the impact of guessing, list more responses than premises and
let a response be used more than once. This helps prevent examinees from
gaining points through the process of elimination.
(d) Put responses in alphabetical or numerical order.
(e) Don’t break matching items across pages.
(f) Give full directions which include the logical basis for matching and
indicate how often a response may be used.

(g) Follow applicable grammatical and syntax rules.

(h) Keep responses as short as possible by using key precise words.
E. Writing Completion and Short Answer Items

1. Completion and short answer items require the examinee to supply a word or
short phrase response. For example:
Directions. Read each statement carefully. Write in the word or words which completes
and makes the statement correct.
1. The two most essential attributes of a measure are: (a) ________ and (b)______.
2. If a test is to be re-administered to the same examinees, the researcher is concerned

about _________ reliability.
3. If the reliability of a test is zero, its predictive validity will be _______.
Answers: 1a. validity; 1b. reliability; 2. test-retest; 3. zero
2. Strengths and limitations (Gronlund, 1998, pp. 96-97; Kubiszyn & Borich,
1996, p. 99-100; Oosterhof, 1994, pp. 98-100; Popham, 2000, pp. 264-266) of
completion and short answer items identified are:
a. Strengths
(1) These item formats are efficient in that due to ease of writing and
answering, more items can be constructed and used. Hence, much
content can be assessed, improving content validity.
(2) The effect of guessing is reduced as the answer must be supplied.
(3) These item formats are ideal for assessing mastery of content where
computations are required as well as other knowledge outcomes.
b. Limitations
(1) Phasing statements or questions which are sufficiently focused to elicit
a single correct response is difficult. There are often more than one
“correct” answer depending on the degree of item specificity.
(2) Scoring maybe influenced by an examinee’s writing and spelling
ability. Scoring can be time consuming and repetitious, thereby
introducing scoring errors.
(3) This item format is best used with lower level cogitative skills, e.g.,
knowledge or comprehension.
(4) The level of technological support for scoring completion and short
answer items is very limited as compared to selected response items
(i.e., matching, multiple choice, and true/false).
3. Item writing guidelines (Gronlund, 1998, pp. 87-100; Oosterhof, 1994, pp.
101-104; & Popham, 2000, pp. 264-266) are:
a. Focus the statement or question so that there is only one concise answer of
one or two words or phrases. Use precise statement or question wording
to avoid ambiguous items.
b. A complete question is recommended over an incomplete statement. You
should use one, unbroken blank of sufficient length for the answer.

c. If an incomplete statement is used, some authors recommend a separate

blank of equal size for each missing word, others consider such an action
to be a clue to examinees as to elements of the correct response. Avoid
using broken lines which correlate to the number of letters in each word of
the desired answer.
d. Place the blank at the end of the statement or question.
e. For incomplete statements, select a key word or words as the missing
element(s) of the statement. Use this item variation sparingly.
f. Ensure that extraneous clues are avoided due to grammatical structure,
e.g., articles such as “a” or “an”.
g. Ensure that each item format allows for efficient and reliable scoring.
III. Constructing Restricted and Extended Response Items to Assess Analysis,

Synthesis, and Evaluation
A. Introduction
1. A restricted response essay poses a stimulus in the form of a problem,
question, scenario, etc. where the examinee is required to recall, organize, and
present specific information and usually construct a defensible answer or
conclusion in a prescribed manner.
a. Restricted responses are best suited for achievement tests and the
assessment of lower order intellectual (knowledge, comprehension, and
application) skills.
b. The restricted response essay provides the examinee with guidance for
constructing a response.
c. It also provides information on how the item is scored.
d. Oosterhof (1994, p. 113) suggests that it should take examinees 10 or
fewer minutes to answer a restricted response essay item. Limit the
number of words in a response to a maximum of 100 or 150.
e. Examples
(1) Compare and contrast each of the four types of reliability. (Restricted Response)
(2) Define each type of validity. (Restricted Response)
2. An extended response essay is one where the examinee determines the

complexity and response length.
a. An extended response essay is an effective device for assessing higher
order cognitive (e.g., Bloom, et al.’s analysis, synthesis or evaluation)
skills.
b. This item format is significantly influenced by writing ability which may
mask achievement deficits.
c. Oosterhof (1994, p. 113) recommends that extended response essays not
be used on tests given the limitations cited below. He does suggest that
extended response items are useful as assignments.
d. Examples
(1) Illustrate how test validity and reliability using an original example involving at least
one type of validity and two types of reliability would be established. (Extended
Response)

(2) Using an original example, explain the process for conducting an item analysis
study, including (a) a brief scenario description, (b) at least three sample items in
different formats and (c) the computation, interpretation, and application of relevant
statistical indices. (Extended Response)
3. Strengths and limitations of essay items (Gronlund, 1998, p. 103; Kubiszyn &
Borich, 1996, pp. 109-110; Oosterhof, 1994, pp. 110-112; Popham, 2000, pp.
266-268) are:
a. Strengths
(1) These formats are effective devices for assessing lower and higher
order cognitive skills. The intellectual skill(s) must be identified and
require the construction of a product that is an expected outcome of the
exercise of the specified intellectual skill or skills.
(2) Item writing time is reduced as compared to select response items, but
care must be taken to construct highly focused items which assess the
examinee’s mastery of relevant performance standards.
(3) In addition to assessing achievement, essay formats also assess an
examinee’s writing, grammatical, and vocabulary skills.
(4) It is almost impossible for an examinee to guess a correct response.
b. Limitations
(1) Since answering essay format items consumes a significant amount of
examinee time, other relevant performance objectives or standards
may not be assessed.
(2) Scores may be influenced by writing skills, bluffing, poor penmanship,
inadequate grammatical skills, misspelling, etc.
(3) Scoring takes a great amount of time and tends to be unreliable given
the subjective nature of scoring.
B. Writing and Scoring Restricted and Extended Response Essays

1. Item writing guidelines (Gronlund, 1998, pp. 87-100; Kubiszyn & Borich,
1996, pp. 107-109; Oosterhof, 1994, pp. 101-104; Popham, 2000, pp. 264-
266) are:
a. Use extended essay responses to assess mastery of higher order cognitive
skills. Use restricted responses for lower order cognitive skills. Use
Bloom, et al.’s Taxonomy or similar cognitive intellectual skills
classification system to identify what cognitive skills are to be assessed.
b. Use the appropriate verbs in preparing the directions, statement, or
question to which examinees are to respond.
c. Ensure that examinees have the necessary enabling content and skills to
construct a correct response to the prompt.
d. Frame the statement or question so that the examinee’s task(s) is explicitly
explained.
e. Frame the statement or question so that it requires a response to a
performance objective or standard.
f. Avoid giving examinee’s a choice of essays to answer. They will most
likely pick the ones they know the answer to, thus adversely affecting
content validity.

g. Have sufficient time to answer each essay item, if used on a test, and
identify the point value. It’s best to use restricted response essays for
examinations.
h. Verify an item’s quality by writing a model answer. This should identify
any ambiguity.
2. Item scoring guidelines (Gronlund, 1998, pp. 107-110; Kubiszyn & Borich,
1996, pp.110-115; Oosterhof, 1994, pp. 113-118; Popham, 2000, pp. 270-271)
are:
a. Regardless of essay type, evaluate an examinee’s response in relation to
the performance objective or standard.
(1) Ensure that the item’s reading level is appropriate to examinees.
(2) Grade all examinees’ responses to the same item before moving on to
the next. If possible, evaluate essay responses anonymously, i.e., try
not to know whose essay response you are evaluating.
(3) For high stakes examinations (i.e., degree sanction tests), have at least
two trained readers evaluate the essay response.
b. For restricted/brief response essays, ensure that the model answer is
reasonable to knowledgeable students and content experts, and use a point
method based on a model response. Restricted essays are not
recommended for assessing higher order intellectual skills.
c. For an extended response essay, use a holistic or analytical scoring rubric
with previously defined criteria and point weights. When evaluating
higher order cognitive skills, generally evaluate extended essay responses
in terms of
(1) Content: Is the content of the response logically related to the
performance objective or standard being assessed? Knowledge,
comprehension, or application skills can be assessed depending on
how the essay item was constructed.
(2) Organization: Does the response have an introduction, body and
conclusion?
(3) Process: Is the recommendation, solution, or decision in the essay
response adequately supported by reasoned argument and/or
supporting evidence? This is the portion of the response where
analytical, synthesis, or evaluation skills are demonstrated. Again, be
very specific as to which skill is being assessed and that the essay
item, itself, requires the examinee to construct a response consistent
with the intellectual skill being assessed.
d. When evaluating essays where a solution or conclusion is required,
consider the following in constructing your scoring plan:
(1) Accuracy or reasonableness: Will posited solution or conclusion
work?
(2) Completeness or internal consistency: To what extent does the solution
or conclusion relate to the problem or stimulus?

(3) Originality or creativity: Is an original or creative solution advanced?

While an original or creative solution might not be expected, when one
is posited, credit should be given.
IV. Statistical Strategies for Improving Select Response Test Items (Item Analysis)
A. Item Analysis: Purpose
1. The purpose of analyzing item behavior is to select items which are best
suited for the purpose of the test.
2. The purpose of testing is to differentiate between those who know the content
and those who don’t. This can be done by:
a. Identifying those items answered correctly by knowledgeable examinees.
b. Identifying those items answered incorrectly by less knowledgeable
examinees.
3. There are two indices: item difficulty or p-value, Index of Discrimination (D).
B. Item Analysis: Indices

1. Item Difficulty (p-value)
a. Item difficulty is determined by the number of examinee’s correctly
endorsing (i.e., answering) a dichotomously scored item. The p-vale has
also been called the “item mean.” The p-value is expressed as a
proportion, with a range from “0.00 to 1.00.”
b. Item format affects p-values, particularly where the item format is able to
guess a response. For example:
(1) 37 x 3 = _____ (This format presents little opportunity to correctly guess--no choices.)
(2) 37 x 3 = ______ (This format presents more opportunity to correctly guess.)
(a) 11.1 (c) 1111
(b) 111 (d) 11.11
c. Target p-values
From the test designer’s perspective, to maximize total test variance
(needed for high reliability coefficients), each item p-value should be at or
close to 0.50. Some recommend a range between 0.40 and 0.60 (Crocker
& Algina, 1986, pp. 311-313).
d. Oosterhof (1994, p. 182) recommend differing p-value targets based on

item format. These are:
(1) True-false and 2 option multiple choice, 0.85.
(2) Three option multiple choice, 0.77.
(3) Four option multiple choice, 0.74.
(4) Five option multiple choice, 0.69.
(5) Short-answer and completion, 0.50.
e. p-values are rarely the primary criterion for item selection into a test and
for tests designed to differentiate between those who know the content and
those who don’t; p-values should be of consistent, moderate difficulty.

2. Item Discrimination Index (D)

a. Items with high discrimination ability are prized as they contribute to
sorting examinee performance. Determining an item’s discrimination
power requires computing “D,” the item discrimination index. “D” is an
important index when selecting items for a test whose purpose is to sort
examinees based on knowledge. “D” works best with multiple choice and
matching items.
b. Formula: D = Pu - Pl
(1) Term Meanings
(a) D = discrimination index
(b) Pu = upper proportion of examinees who answered item correctly
(c) Pl = lower proportion of examinees who answered item correctly
(2) To determine Pu and/or Pl, select the upper 25-27% of examinees and
the lower 25-27% of examinees. Determine the proportion in each
group answering the item correctly.
(3) For example: Item 1, Pu = 0.80 & Pl = 0.30, so 0.80 - 0.30 = 0.50.
Thus, D = 0.50. Eighty percent answered Item 1 correctly in the high
scoring group; whereas, 30% did so in the lower scoring group. See
Interpretive Guidelines below.
c. Properties of “D”
(1) Ranges from -1.0 to 1.0
(2) Positive values indicate the item favors the upper scoring group.
(3) Negative values indicate the item favors the lower scoring group.
(4) The closer “D” is to 0.00 or 1.0 the less likely the item is going to have
a positive value.
d. Interpretative Guidelines (Crocker & Algina, 1986, pp. 314-315)

(1) D > .40, well-functioning item—keep as is.
(2) 0.30 <D > 0.39, little or no item revision needed.
(3) 0.20 < D > 0.29, marginal item, needs revision.
(4) D < 0.19, eliminate or completely revise item.
3. Distractor Analysis (For multiple choice or matching items)

a. Distractor analysis assesses how well each multiple choice item response
option (also called foils or distractors) attracts examinee endorsement (i.e.,
selection as correct answer); all foils should attract an endorsement (i.e.,
be selected). Don’t use with True/False items.
b. To determine an acceptable level of distractor functioning, one considers:

(1) A distractor not attracting any endorsement is poorly functioning.
(2) A distractor attracting an incorrect endorsement from a high scoring
student is fine (most likely an error), provided there are only a few.

(3) The most likely incorrect response should have the 2’nd highest
endorsement level. If all incorrect foils are equally likely, then lower
scoring examinee selection should be fairly consistent across the
incorrect foils.
(4) Revise foils which high- and low-scoring examinees together select.
c. Many commercial item analysis software programs will only provide an

“all examinees” distractor analysis. However, understanding how
distractors function is best achieved by reviewing high and low scoring
student foil endorsements along with the “all examinees” option. Use the
“high” and “low” scoring examinee groups used in an item discrimination
analysis.
4. Interpreting an Item Analysis Report

a. Item 6 A B* C D E Omit
High .07 .86 .00 .07 .00 .00 (Must = 1.0 or 100%)
Lower .13 .60 .07 .13 .07 .00 (Must = 1.0 or 100%)
All .12 .70 .05 .10 .03 .00 (Must = 1.0 or 100%)
p-value = 0.70 D = 0.26 * = Correct foil Rows = 1.0 or 100%

(1) Distractors function adequately. All foils attracted some endorsement.
(2) For a five distractor item, the p-value is okay, as the target is 0.70.
(3) The “D” value suggests some revision as it falls into the 0.20 to 0.29
range. Make foils “C,” “D,” and/or “E” more attractive to higher
scoring students.
b. Item 13 A B C* D E Omit
High .00 .00 100 .00 .00 .00
Lower .00 .00 100 .00 .00 .00
All .00 .00 100 .00 .00 .00

(1) Unless the item is an “ice breaker,” it is a very poorly functioning
item. It doesn’t meet any targeted p-value and has no discrimination
ability. The item will not contribute to test score variance (i.e.,
reliability.).
(2) Item should be discarded or completely revised, unless it is an “ice
breaker.”
c. Item 29 A B* C D Omit
High .36 .36 .00 .28 .00
Lower .41 .24 .10 .25 .00
All .24 .60 .10 .06 .00

(1) The target p-value (0.74) for a 4-foil multiple choice item suggests that
the item has promise, but requires revision. However, the 0.11 “D”
value indicates that the item has very little discriminating power.
(2) Distractors “A”, “B”, and “D”, attracted similar endorsements from
both the higher and lower scoring groups. This suggests item
ambiguity.
(3) A revision strategy would be to make foil “C” more attractive and
revise foils “A”, “B”, and “D” so that they are more attractive to lower
scoring or higher scoring examinees, but not both.
5. It should be clear that improving test items, regardless of format, requires a

balance between targeted p-values and acceptable “D” values. Revision
requires a thorough understanding of the tested content and the cognitive (i.e.,
intellectual and thinking) skills required to correctly answer an item. The item
analysis study model advanced by Crocker and Algina (1986, pp. 321-328) is
instructive in guiding test developers to efficiently and effectively revise test
items. The elements of the model are:
a. Decide whether you want a norm-referenced or criterion-referenced score
interpretation (see Chapter 1).
(1) Do you want a norm-referenced (NRT) or criterion-referenced test
(CRT) score interpretation? In either case, the IA indices discussed
above are appropriate.
(2) Given how content validity is established, it can be argued that test
results (e.g., scores) can be used to assess instructional efficacy and as
a basis for initiating remedial instruction, as NRT and CRT tests are
constructed in much the same manner.
b. Select relevant item performance indices.
(1) For most IA studies, the parameters of interest are p-value, “D” and
distractor analysis (for multiple choice items).
(2) Comparing higher and lower scoring examinees is standard practice.
c. Pilot the test items with a representative examinee sample for whom the
test is intended.
(1) Pilot test group size:
(a) 200 examinees for local studies or 5-10 times the number of
examinees compared to the number of items; for example, if there
are 30 test items, a minimum of 150 examinees is recommend.
(b) Several 1000 examinees are needed for large scale studies.
(2) The above recommendations are for school district-wide, regional, or
state-wide testing programs. It is not expected that classroom
teachers/trainers developing examinations would need a pilot test with
such large samples, but having a knowledgeable colleague(s) review
your test is recommended.
d. Compute each item performance indices (See Part IV of this Chapter.)

e. Select well performing items and/or revise poorly functioning ones.

(1) Use target p-values and the item discrimination index criteria.
(2) Crocker and Algina (p. 323) have described and resolved two common
situations:
(a) When there are more items than can be administered, the task is to
select those items which make the most significant contribution to
the desired level of test reliability and validity. Items with p-
values between 0.40 and 0.60 are recommended as such items have
been shown to reliably discriminate across virtually all ability
groups.
(b) If the item pool is not overly large and the test developer wants to
keep every possible item, revise those items with low “D” values.
(3) If necessary, repeat the pilot test with revised or newly written items.
f. Select the final set of items.
(1) Use the “D” value and p-value criteria to guide decisions for test item
selection and revising flawed items. Eliminate only as a last resort.
(2) Look for unexpected or abnormal response patterns, particularly in
distractor analysis.
g. Conduct a cross-validation study to determine whether or not the desired
results have been attained.
(1) This not usually done in classroom achievement testing. However, for
reasonably high stakes tests such as a common departmental course
final, one should consider conducting a validation study. Such a study
might involve using the same or similar items several times and then
taking the best functioning items for a final version of the examination.
(3) At this stage, the test developer’s interest is to see how well items
function a second time. Some items will likely be removed. However,
the examination is for the most part, constructed.
V. Constructing Direct Performance Assessments

A. Introduction
1. Direct performance assessments (DPA’s) are used when specific processes
(e.g., company procedures), skills (i.e., behaviors), outcomes (e.g., products),
affective dispositions (e.g., attitudes, opinions, intentions), or social skills
(e.g., self-direction, capability to work with others, manners, etc.) are to be
directly assessed.
2. In direct performance assessment (DPA), examinees are presented a simulated

or “real world” task, problem, or process where they are to demonstrate a
procedure or construct a verbal, written, product, or blended product usually
in the form of a demonstration, presentation, project, report, or term paper.
The constructed response is then rated by a judge(s) (i.e., trainer, judge,
teacher, professor, etc.), using checklists, rating scales or scoring rubrics.

a. Direct performance assessment inferences are highly contextually

dependent. Any performance inference is based only on the demonstration
or product produced and the context within which it was constructed or
demonstrated.
b. Common assessment devices include checklists, rating scales, portfolios,
and task descriptions with scoring rubrics.
3. Direct performance assessment can also be used to supplement traditional

forms of assessment, assess higher order intellectual skills, and/or assess
psychomotor skills.
a. Traditional forms of testing (e.g., paper and pencil or computer based
testing) will not allow inferences or judgments, based on direct
performance observation.
(1) Recall that traditional testing is most efficient, valid, and reliable for
testing lower order intellectual skills (i.e., knowledge, comprehension,
and simple applications).
(2) Disposition (i.e., attitudes) assessment is most often indirect and
employs rating scales (also called an index).
(a) An examinee can individually respond to a rating scale to self-
report his or her motivation, attitude, opinion, or intention.
(b) Direct performance assessments, using behaviorally oriented rating
scales, made by observation, can be used to assess or confirm
indirect measurement of examinee motivation, intention,
cooperation, social interaction skills, etc.
b. Higher order intellectual skills (i.e., complex application, analysis,
synthesis, and evaluation) can be assessed using DPA’s.
(1) Assessing higher order intellectual skills would include tasks which
require the acquisition, organization, and use or application of
information to real or simulated problems, scenarios, or opportunities.
(2) Assessment devices or tools for these products include charts, tables,
map reading, drawings, essay composition, experiments, projects,
completion of tasks, etc.
c. The assessment of psychomotor skills or the integration of intellectual and
psychomotor skills includes tasks which require the examinee to provide
an observable performance, which might take form as a speech, written
communication, typing, dance, gymnastic routine, sales presentation,
mock interview, or correctly cooking a recipe.
B. Direct Performance Assessment: Advantages and Limitations

1. DPA Advantages
a. DPA usage allows for performance inferences or judgments which are
impossible or at least very difficult to make with traditional testing
strategies. Skill performance diagnosis and correction is also possible.

b. Higher order intellectual skills, built on lower order intellectual skills, can
be directly assessed or tested through performance assessment. Well
designed and implemented tasks can be “correctly” solved in a variety of
ways thus facilitating examinee critical thinking skills and for the making
of evaluative inferences by examiners.
c. Very clear connections to teaching or training quality can be made. Both
the process of constructing the response (i.e., product) and the response
itself can be assessed and evaluated; thereby improving teaching and
learning, and assessment.
d. Examinees’ commitment is increased given their sense of control over the
learning, assessment and evaluation processes. Motivation tends to be high
when examinees are required to produce an original work or product.
2. DPA Limitations
a. Performance assessments are very labor and time intensive to design,
prepare, organize, and evaluate. Records must also be maintained.
b. Performance has to be scored immediately if a process is being assessed.
c. Scoring is susceptible to rater error. Raters or examiners must be highly
trained on the task description, performance criteria, and rating form to
assess performance similarly and consistently. A plan for “breaking ties”
is needed; two examiners may rate a performance quite differently. A third
more senior or experienced rater can serve as the “tie breaker.”
d. Complex intellectual or psychomotor skills are composed of several
different but complimentary enabling skills (e.g., reading level, drawing
talent, physical coordination, stress tolerance, etc.) which might not be
recognized or assessed. Examinees will likely perform some enabling
skills more proficiently than others. Critical enabling skills should be
identified and specifically observed and rated.
e. Other potential limitations include time, cost, and the availability of
equipment and judges. Due to these issues, performance assessment must
be used to assess very highly relevant skills, which are teachable.
C. Constructing a DPA: The 3 Phase Process

1. Plan the Direct Performance Assessment (DPA).
a. Senior examiners are responsible to determine what process, skill, product,
or disposition is/are to be assessed; salient indicators which define realistic
performance levels; the task specification; the scoring mechanics; and
rater qualification and proficiency training.
(1) Only key indicators or “primary performance traits” should be
assessed so that the raters don’t become overwhelmed, ineffectual, and
inaccurate by being mired in excessive detail.
(2) Ensure and document that examinees have mastered prerequisite
knowledge and skills prior to launching a performance assessment.
(3) Performance raters’ or examiners’ training is planned
b. The outcome of phase 1 is a comprehensive DPA construction plan.

2. Construct the Individual DPA Elements (Stiggins, 1997)

a. First, specify what decision (mastery, rank order, or a combination) is to
be made. For example (Appendix 5.2), graduate students in an assessment
course must successfully construct a direct performance assessment task
according to performance criteria prepared by the professor at least to the
“proficient level;” this is a mastery decision.
b. Second, write the task description.
(1) Is the exercise structured and/or natural? Natural might include a
school or work environment which is familiar to the examinee.
(2) What actual examinee performance is to be assessed in terms of
content and skill foci, processes to be employed, and the work product
to be produced?
(3) What are the performance dimensions/criteria? These should have
been documented in the DPA planning phase. For example (Appendix
5.2), graduate students’ skill (working in pairs) in constructing a direct
performance assessment task is measured across the dimensions of
“task description clarity,” “performance criteria,” “scoring system,”
and “authenticity;” see the scoring rubric (labeled PTQAI).
(4) Ensure the task description is understandable to intended examinees.
c. Third, write the performance scoring mechanism.
(1) Explicitly set and define performance levels. For example (Appendix
5.2), a project is assessed by the professor using a rating scale with
four performance criteria (“Task Description,” Performance Criteria,”
Scoring System,” and “Authenticity”) and three performance gradients
or levels (“Meets Standard,” “Mostly Meets Standard, etc.). The
performance categories must represent realistic levels of examinee
performance.
(2) The scoring mechanism is constructed with points allocated to ensure
consistent and valid performance segmentation. For example, see
Appendix 5.2. Each work team’s DPA will be reworked until
performance criteria are minimally met at the “Proficient” level
according to the professor’s judgment.
(a) Develop a draft of the checklist or rating scale, if the decision has
been made to use either, see Part V.D below. Be sure to closely
edit and proof.
(b) Decide whether a holistic or analytic scoring strategy will be used
as well as the type of scoring rubric (i.e., task-specific, skill-
focused, or generic). See the more detailed discussion in Part V.E,
below. A holistic rubric produces a single score across multiple
performance criteria; whereas an analytical rubric produces scores
for specific criteria, but the individual criterion scores are
combined into one composite (i.e., total) score. An analytical
rubric holds much instructional value.
d. Fourth, the rater’s training is designed; identify raters (i.e., examiners).
e. The outcomes of phase 2 are the specific decision to be made, a task
description, a performance scoring strategy, and a rater training schema.

3. Assemble and Field-test the Task Description and Scoring Plan

a. Have qualified colleagues review, comment, edit and proof the task
description and checklist, rating scale or scoring rubric. Ensure the two
align.
b. Recruit desired raters and thoroughly train through role-playing.
c. Repeat pilot testing until satisfied with checklist, rating scale, or rubric
performance.
(1) Look carefully for administration difficulties, unclear items (i.e.,
indicators), incomplete decision-making information, accuracy of
performance descriptions, etc.
(2) Compute and interpret appropriate reliability indices, if the checklist,
rating scale, or rubric format permits such computations; it should.
(3) Evaluate the results; change evaluative criteria or standards, if needed.
d. Prepare documents to inform examinees. Examinees should be told what
is expected of them and be given copies of the task description, rating tool
(checklist, rating scale, or rubric), and descriptions or definitions of the
performance levels prior to instruction/training and “grading.”
D. Constructing DPA Checklists & Rating Scales

1. Checklists
a. A checklist is a list of specific, discrete actions to be performed by
examinees which are to be observed by a rater. Often, the actions are
listed in the expected order of performance.
(1) Checklists typically employ a binary choice (e.g., “Met Expectation”
or “Not Meet Expectations”) with a “Did Not Observe” option.
(2) Checklists are well suited to score procedures. They are time
consuming to construct; efficient to score; and highly reliable and
defensible, provided they are content valid. A scoring mechanism is
need to classify examinees into performance levels and compute
reliability coefficients.
(3) Checklists provide examinees with quick quality feedback on
performance. See Appendix A.9, section A.6 for a brief example.
b. If you have ever taken a first aid or CPR class, at least a portion of your
performance was most likely assessed with a checklist. If you’ve not
taken a first aid or CPR class, you should, if you can.
2. Rating Scales
a. Rating scales are efficient in assessing dispositions (attitudes), work
products, and social skills. Rating scales are more difficult to construct
than checklists, but tend to be both reliable and defensible, provided they
are content valid. Rating scales are efficient to score and can provide
quality feedback to examinees.
b. It is essential that each response option be fully defined and that the
definition be logically related to the purpose of the rating scale and be
progressive (i.e., represent plausible examinee performance levels). Like
checklists, rating scales tend to be unidimensional in that each assesses

one characteristic of performance. However, related unidimensional rating

scales are often combined into multidimensional rating scales.
(1) See Appendix 5.6, Part A (presentation, Execution) for an example of
a unidimensional rating scale. When Parts A, B (presentation
Presence), and C (presentation, Technology) are combined, a
multidimensional rating scale is produced. Performance levels are
defined by key words, such as “Poor” or “Good.”
(2) Presented in Appendix 3.1 is a unidimensional rating scale with five
point response options. Appendix 5.4 is a dichotomously scored rating
scale.
c. Rating scales tend to be analytical in that they are more descriptive and
diagnostic than checklists. For example, problem solving is thought of as
a series of sequential, but related steps and sub-steps. A rating scale will
assess examinee performance across each key sub-step and step in the
process. When using a checklist or rating scale, examinee performance
should be compared only to the established performance scale, not other
examinees.
3. Constructing and Scoring DPA Checklists & Rating Scales

a. General Construction Guidelines for DPA Checklists and Rating Scales
(1) Ensure that the process, skill, product, disposition, or combination
thereof, along with key indicators have been clearly and fully
specified. Use only the number of indicators necessary to allow for
import inferences to be accurately made. Rate only the key indicators.
(2) Make sure that there are sufficient resources (e.g., time, money,
equipment, and supportive colleagues) to successfully implement the
performance assessment. Decide a priori whether or not (a) examinees
are permitted reference materials, (b) examinees have sufficient
enabling knowledge to successfully perform, and (c) what are the
scoring criteria.
(3) Estimate the amount of time for virtually all examinees to complete
the task. If sufficient time is not available, then reframe the task.
(4) Use the fewest, most accurate words possible to define the points on
the checklist or rating scale performance scoring continuum. Avoid
using redundant phrases. Telegraphic phrases, centered on key nouns
and verbs, which are understood by each rater are recommended.
(5) Ensure that the form, itself, is efficient to score. Rater responses
should be recorded by circling a number or using an “X” or “” or
“N” (see Appendix 5.3 or 5.5 are examples). For either checklists or
rating scales, provide a convenient space to mark or circle. Provide
space for brief comments.
(6) When rating a process, individual items must be grouped in the correct
sequence. For product assessments, items of similar content or stage of
production should be grouped together.
(7) Keep item polarity (i.e., response options consistent).

(a) For rating scales, place the least desired characteristic or lowest
rating (e.g., “Strongly Disagree”) on the left end of the continuum
and the most desired quality or highest rating on the right end of
the continuum (e.g., “Strongly Agree”).
(b) Arrange similarly for checklists (e.g., “Poor Quality” “Acceptable
Quality,” or “High Quality”).
(8) Provide a space for rater comments beside each item or at the end of
the checklist or rating scale.
(9) Always train examiners or raters so they proficient in using the
checklist or rating scale.
(a) After each practice session, debrief the raters so that all can
understand each other’s logic in rating an examinee’s performance,
and where necessary arrive at a rating consensus.
(b) Stress to raters the importance of using the full range of rating
options on rating scales; many raters use only the upper tier of the
rating continuum, which is fine provided all examinees actually
perform at those levels.
b. Scoring DPA Checklists and Rating Scales

(1) The binary choice (e.g., “No” or “Yes”) can be weighted “1” or “2,”
respectively (Appendix 5. 3) and then summed to arrive at a total score
and if items are grouped, subtest scores as well. Avoid weighting the
“Not Observed” Option. For most checklist applications, the “Not
Observed” should never be endorsed.
(2) The multi-point rating options (see Appendix 4.6 for other examples),
used in rating scales, can be weighted where “1” (Strongly Disagree)
represents the lowest rating and “5” (Strongly Agree) the highest, if a
five point scale is used. After point weighting, total and/or subtest
scores can be computed. See Appendix 4.9 on writing scorable items.
(3) A common strategy is to place examinees into performance categories
based on their ratings; see Appendices 3.1 and/or 4.9, Table 4.9.2.
E. Constructing DPA Scoring Rubrics

1. A scoring rubric, similar to a rating scale, is (a) composed of multiple rating
subscales or subtests, which when assembled together, represent the complete
process, skill, or product to be rated (e.g., Appendix 5.5) or (b) a series of
performance descriptions (e.g., Doesn’t, Meets, or Exceeds Expectations)
which address multiple qualities simultaneously using a common scoring
scale (Appendix 5.4).
2. Definition of Terms
a. A rubric is an instrument for rating an examinee response to a stimulus or
task description and is scored either holistically or analytically.
b. Evaluative criteria are performance standards which are rated, typically
using points. See the 33 standards identified in Appendix 5.2. Each
performance standard should be logically related but functionally

independent of other standards. Group related evaluative criteria together;

but in the proper ordered sequence.
c. Quality Definitions are various performance levels (e.g., “Exceeds
Expectations,” “Meets Expectations” or “Doesn’t Meet Expectations”)
which are specified and defined so that differences in examinee
performance are determined. Each performance level needs its own
definition.
3. Scoring Strategy is either holistic or analytic.

a. Holistic: All evaluative criteria are considered but summarized into a
single, overall quality judgment, usually presented in a single score.
(1) While fairly easy to adapt or write, holistic rubric reliability tends to
be weak and generalizability very limited.
(2) The holistic rubric is subject to scoring bias and rater inconsistencies.
(3) Unless holistic rubrics are well defined and detailed, raters are likely to
substitute their personal opinions or definitions. Each performance
description must represent plausible examinee performance levels.
These definitions should be written by senior examiners who are very
familiar with the process, skill, or product being rated.
(4) The holistic rubric is satisfactory (a) if an overall appraisal of a
specific skill or ability can be achieved and makes sense and (b) if
used on less critical process, skills, or products and not used when high
stakes decisions are to be made about an examinee’s performance.
(5) Holistic Scoring Rubric for rating a business memo
6 = The memo is exceptional in conforming to the model format
provided and meeting writing mechanics and punctuation
expectations.
5 = The memo is very good in conforming to the model format
provided and meeting writing mechanics and punctuation
expectations.
4 = The memo is a good in conforming to the model format provided
and meeting writing mechanics and punctuation expectations.
3 = The memo is adequate in conforming to the model format provided
2 = The memo is poor in conforming to the model format provided and
meeting writing mechanics and punctuation expectations.
1 = The memo is very poor in conforming to the model format provided
b. Analytic: Each evaluative criteria is scored and contributes directly to an

examinee’s subtest and/or total score.
(1) If the focus of the rating exercise is to assist individual examinees to
improve process or skill performance, then analytic rating is
recommended, despite its time intensive nature and high cost.

(2) Analytic rubrics are richer in detail and description, thereby offering
greater diagnostic value. This provides examinees with a detailed
assessment of strengths and weaknesses.
(3) Analytic rubrics should be based on a thorough analysis of the process,
skill, or task to be completed. Performance definitions must clearly
differentiate between performance levels. Detailed checklists and
rating scales are often used in analytic ratings.
(4) Analytic rubrics are presented in Appendices 5.2, 5.4, and 5.5.
4. Types of Analytical Scoring Rubrics

a. Task-specific rubrics (Appendix 5.2) focus on a particular task rather than
the skill associated with that task. Task specific rubrics contribute to
improving inter-rater reliability. Popham (2000, p. 290) recommends
against use. However, Haladyna (1997, pp. 134) has observed:
(1) Task-specific scoring rubrics are expensive to create; but, if well-
constructed, it can be used repeatedly.
(2) Task-specific rubrics are both valid and reliable and provide
examinees with accurate performance information.
b. Skill-focused rubrics (Appendix 5.4) center on the skill being assessed.

Haladyna’s comments on task-specific rubrics apply to skill-focused
rubrics. Popham (2000, p. 290) recommends the use of skill-focused
rubrics. He argues that skill mastery is more instructionally relevant than
task mastery. Popham provides guidelines for constructing a skill focused
rubric:
(1) The rubric should consist of 3 –5 evaluative criteria.
(2) A teachable attribute or sub-skill should form the basis of the
evaluative criterion
(3) The evaluative criteria should apply to any task designed to assess
student skill mastery within the same or similar context.
(4) The format of the rubric should be well organized.
c. Generic rubrics are applicable across a variety of contexts, such as scoring

a written assignment regardless of content, topic, or author. See Figure
5.1. The numbers in parenthesis are point values associated with each
performance level; 20 points are possible (4 points x 5 criteria).
(1) While generic rubrics improve the generalizability of results, Haladyna
(1997, p. 134) has identified two problems with generic rubrics:
(a) “[A] single rating scale seldom leads to a reliable test result”
(b) “[N]o ability [skill or task] is so simple that a single rating scale
can cover it.”
(2) Where possible, convert a generic rubric to either a task or skill
focused rubric.

Criterion Novice (1) Apprentice (2) Proficient (3) Exemplary (4)

Content Little or no Content was Content was Content was
substantive partially substantive, substantive,
content was substantive, but meeting exceeding
provided. not meeting expectation. expectation.
expectation.
Organization Organization Organization Organization Organization

was lacking. partially met met expectation. exceeded
expectation. expectation.
Originality The writer The writer’s The writer’s The writer’s

provided little or insightfulness insightfulness insightfulness
no insight into partially met met expectation. exceeded
the topic. expectation. expectation.
Critical The critical The critical The critical The critical

Thinking thinking shown thinking thinking thinking
was limited or demonstrated demonstrated demonstrated
not in evidence. partially met met expectation. exceeded
expectation. expectation.
Writing The writing The writing The writing The writing

Mechanics mechanics were mechanics mechanics mechanics
(Spelling, poorly executed demonstrated demonstrated demonstrated
punctuation, or not in partially met met expectation. exceeded
etc.)
evidence. expectation. expectations.
Figure 5.1 Generic Rubric Example

References
Airasian, P. W., (1997). Classroom assessment (3rd ed.). New York, NY: McGraw-Hill.
Bloom. B. S., Engelhart, M. D., Frost, E. J., & Krathwohl, D. (1956). Taxonomy of
educational objectives. Book 1 cogitative domain. New York, NY: Longman.
Crocker, L. & Algina, J. (1986). Introduction to classical and modern test theory.
New York, NY: Holt, Rinehart, & Winston.
Gagne, R. M. (1985). The conditions of learning and theory of instruction (4th ed.).
Chicago, IL: Holt, Rinehart, & Winston, Inc.
Gronlund, N. E. (1998). Assessment of student achievement (6th ed.). Needham Heights,

MA: Allyn & Bacon.
Haladyna, T. M. (1997). Writing test items to evaluate higher order thinking. Needham
Heights, MA: Allyn & Bacon.
Kubiszyn, T. & Borich, G. (1996). Educational testing and measurement. New York,
NY: Harper Collins College Publishers.
Lyman, H. B. (1998). Test scores and what they mean (6th ed.). Needham Heights, MA:
Allyn & Bacon.
Mitchell, R. (1996). Front-end alignment: Using standards to steer educational

practice. Washington, DC: The Education Trust.
Oosterhof, A. (1994). Classroom applications of educational measurement (2nd ed.).

New York, NY: Merrill.
Popham, W. J. (2000). Modern educational measurement (3rd ed.). Needham Heights,

MA: Allyn & Bacon.
Quellmalz, E. S. (1987). Developing reasoning skills. In Joan Baron & Robert Sternberg
(eds.) Teaching thinking skills: Theory and practice (pp. 86-105). New York, NY:
Freeman.
State of Florida (1996). Florida curriculum framework: Language arts PreK-12

Sunshine State standards and instructional practice. Tallahassee, FL: Florida
Department of Education.
Stiggins, R. J. (1997). Student-centered classroom instruction. Columbus, OH: Merrill.
Stiggins, R. J., Griswold, M. M., & Wikelund, K. R. (1989). Measuring thinking skills
through classroom assessment. Journal of Educational Measurement, 26, (3), 233-
246.

Appendices
Appendix 5.1 is an interpretive exercise example. Multiple choice items can be

constructed such that they assess higher-order thinking skills. An interpretive exercise has
a core prompt or scenario around which several multiple-choice items are constructed.
Appendix 5.2 is a task specific, analytical rubric which includes a detailed description of
the task to be accomplished and the performance scoring rubric.
Appendix 5.3 is a simple rating scale that was used by team members to rate the
contribution of each other's contribution in the construction of a class project.
Appendix 5.4 is a skill focused scoring rubric used to rate student critiques of an article
from a juried journal. It has a very brief task description but most of the performance
criteria are presented in the rubric. The score for each criterion is summed to reach a final
performance rating on the assignment.
Appendix 5.5 is a three-dimensional (three subtest) skill focused, analytical scoring

rubric measuring presentation performance.
Appendix 5.6 describes ethical test preparation strategies for examinees as well as
guidance on test taking skills.

Appendix 5.1 Interpretative Exercise Example

The following essay was written by a student and is in rough-draft form. Read “Mt.
Washington Climb.” Then answer multiple-choice questions 15 through 18.
Mt. Washington Climb

1 I had gotten up around four-thirty that cool summer morning, anxious to get on the road.
2 My parents were still asleep, however, my mother had my lunch packed and waiting on the
3 kitchen table. I double-checked my gear, grabbed my boots, backpack, and lunch, and left
4 the house. Walking to my car, the first pale hint of red was entering the sky as the sun rose
5 higher and higher. I was right on schedule and greatly anticipating a good day.
6 It was a hot, dry summer morning when I reached the parking lot designated for hikers.
7 The cars in the parking lot were mostly from other states. I opened the trunk of my car and
8 grabbed my hiking boots. I had definitely chosen the right day for my first climb of Mt.
9 Washington. I tied my boots, put on my pack, made sure the car was locked up, and walked
10 over to the map displayed at the head of the trail. I studied the map for a few minutes and
11 then started my six-mile journey to the summit.
12 For the first two miles I walked slowly, enjoying the scenery. It was very beautiful out
13 there. The birds were out, the air was crisp and clean, and a soft breeze tickled my ears. After
14 an hour and a half had passed, I gradually picked up the pace.
15 I reached an intersection in the trail at about eleven o’clock. The sun was almost
16 directly overhead, and I judged the temperature as about ninety degrees. I had three miles
17 to go and I felt great. I drank a bottle of water before continuing. As I was about to.
18 get up, a huge deer walked right out in front of me; I never even heard it. It was by
19 far the most magnificent-looking animal I had ever seen. The deer’s fur was light brown,
20 almost tan, and its antlers had strips of golden velvet that were torn and dirty. Just as
21 soon as the deer was there, he was gone, and I was back on my way to the summit
22 I walked cautiously among the trees for another hour. As I was walking, I noticed the
23 sky got brighter and brighter. Soon I broke through the tree line, and I could see the summit.
24 The sun glistened off the tower. Hundreds of people climbing toward the summit. I hesitated
25 for a moment, awed by the view, and then scrambled over the rocks toward the summit.
26 Beads of sweat ran down my face as I pushed toward the top. The summit was half a
27 mile away, yet it seemed like two feet. My legs burned but nothing could have stopped
28 me—not a rockslide, an earthquake, an avalanche—nothing. Determination filled my body
29 and gave me phenomenal energy. What seemed like two minutes was forty-five and before I
30 knew it I was on my knees at the summit. I had made it.
1. What is the correct punctuation edit for the sentence in lines 2 and 3 (My . . . table.)?
A. My parents were still asleep; however, my mother had my lunch packed and
waiting on the kitchen table.
B. My parents were still asleep however; my mother had my lunch packed and waiting the
kitchen table.
C. My parents were still asleep, however, my mother had my lunch; packed and waiting on
the kitchen table.
D. My parents were still asleep, however, my mother had my lunch packed and waiting; on
the kitchen table.
ID:27480 Mt. Washington A

2. Which is the correct revision of the sentence in lines 4 and 5 (Walking . . . higher.)?
A. The first pale hint of red was entering the sky as the sun rose higher and higher,
walking to my car.
B. As the sun rose higher and higher, walking to my car, the first pale hint of red was
entering the sky.
C. Walking to my car, I noticed the first pale hint of red entering the sky as the sun rose
higher and higher.
D. Walking to my car, the sun rose higher and higher as the first pale hint of red was
entering the sky.
ID:88949 Mt. Washington C
3. Which is a sentence fragment?

A. I opened the trunk of my car and grabbed my hiking boots. (lines 7–8)
B. It was very beautiful out there. (lines 12–13)
C. Hundreds of people climbing toward the summit. (line 24)
D. I had made it. (line 30)
ID:27490 Mt. Washington C
4. Which sentence does not add to the development of the essay’s main idea?
A. The cars in the parking lot were mostly from other states. (line 7)
B. I had definitely chosen the right day for my first climb of Mt. Washington. (lines 8–9)
C. After an hour and a half had passed, I gradually picked up the pace. (lines 13–14)
D. Beads of sweat ran down my face as I pushed toward the top. (line 26)
ID:27487 Mt. Washington A
Source: New Hampshire Educational Assessment and Improvement Program, New Hampshire Department
of Education. End-of-Grade 10 COMMON ITEMS, Released Items 2000-2001.
Retrieved June 1, 2002 from: http://www.ed.state.nh.us/Assessment/2000-2001/2001Common10.pdf

Appendix 5.2 Task Specific, Analytical Scoring Rubric

Direct Performance Assessment Task Description
Student work teams will construct a direct performance task which meets the standards in the attached
Performance Task Quality Assessment Index (PTQAI).
Each work team will develop a unit or mid-term performance task with a scoring rubric suitable for the 6 th -
12th grade classroom. For this task, your work team will rely on the text, Internet research, team member
experience, external consultants (e.g., senior teachers, etc.) and research and/or professional journals. The
final work product is to be submitted electronically using Microsoft Word 2007 or a later edition.
The professor will give one “free read,” using the Performance Task Quality Assessment Index (PTQAI)
before “grading.” The work product will be assessed on four dimensions or traits: quality of the task
description, clarity and relevance of the performance criteria, scoring rubric functionality, and task
authenticity, i.e., how realistic the intended task is for examinees. The rubric is analytical during the
formative phase of the task and holistic at “grading,” as a total score is awarded. Maximum points are 132.
See the PTQAI.
The task will be divided into two parts: task description and rubric. Students will submit a clear set of
directions (PTQAI items 1-3), appropriate task description (PTQAI items 4-10), relevant performance
criteria (PTQAI items 11-15), a suitable scoring rubric (PTQAI items 16-23), and be authentic (PTQAI
items 24-33). The performance criteria may be embedded within the scoring rubric.
The team must describe the classroom performance assessment context, using the attached Direct
Performance Assessment Task Scenario.
Direct Performance Assessment Task Scenario
Before constructing the direct performance task, complete the scenario description to set the context within
which it is set. Provide the following information about examinees and their school setting by either filling
in the blank with the requested information or checking a blank as requested.
Ensure that intended learning target content or skills are fully described; do not refer the professor to a web
site or other document containing this information. Otherwise, the task will be returned until the learning
target information is supplied as required.
Examinee Characteristics
1) Ages: _______ (Average Age) 2) Grade Level: _________ (Specify)
3) Ethnic Mix (name ethnicity, percent of class):

______________________________________________________________________________________
__________________________________________________________
4) Free/reduced Lunch: ________% of class 5) LEP: ______% of class
6) Exceptionalities present in class: ___________________________________________
7) Classroom: Regular Ed: ________ Inclusive: _______ (Check one)
8) School Setting: Urban____ Suburban_____ Rural_____ (Check one)
9) School Ownership: Public: _____ Private: ______ Charter: _____ (Check one)
10) School Type: Elementary: ____ Middle: ____ High: ___ (Check one)

11) School Grade Range: ________ 12) School Size: _______
13) Other Relevant Context Setting Information:

______________________________________________________________________________________
______________________________________________________________________________________
Learning Target(s): Clearly state the intended learning target content and/or skills which are to be
assessed in your extended performance task.
Classroom Assessment Plan: Describe in a few paragraphs your general classroom assessment plan. Next,
describe how your team’s extended performance task would fit into that assessment plan and how you
would use the information to improve classroom teaching and learning.
Performance Task Quality Assessment Index (PTQAI)
The PTQAI is the scoring rubric used to assess the product to be produced in response to the task
description. The PTQAI is composed of 33 evaluative criteria (called standards) distributed across these
four subtests or dimensions: Task Description, Performance Criteria, Scoring System, and Authenticity.
Possible scores are “Meets Standard,” 4 points: “Mostly Meets Standard,” 3 points; “Marginally Meets
Standard,” 2 points; “Does Not Meet Standards,” 1 point; and “Missing or Not in Evidence,” 0 points.
Performance level quality definitions are:
“Meets Standard” means the response fully and completely satisfies the standard.
“Mostly Meets Standard” means the response nearly fully and completely satisfies the standard.
“Marginally Meets Standard” means the response partly satisfies the standard.
“Does Not Meet Standard” means the response does not fully and completely satisfy the standard.
“Missing or Not in Evidence” means the response presents no evidence of meeting the standard.
In order to compute a total score for summative evaluation purposes, the following criteria are used:
Exceptional corresponds to an A (95-100%). Performance is outstanding; significantly above the usual

expectations.
Proficient corresponds to a grade of B to A- (83-94%). Performance is at the level of expectation.
Basic corresponds to a C to B- (75-82%). Performance is acceptable but improvements are needed to meet
expectations well.
Novice corresponds to an F (< 74%). Performance is weak; the skills or standards are not sufficiently
demonstrated at this time.
To compute a summative performance level, total earned points are divided by points possible (132).

Performance Task Quality Assessment Index (PTQAI)

Each direct performance task should meet the 33 standards presented in this index. Four dimensions of
performance assessment are examined: the task description, performance criteria, the scoring system, and
authenticity. For each standard, one of five performance levels or scores is possible: 0= Missing, 1= Does
not meet standard, 2=Marginally standard, 3= Mostly meets standard or 4= Meets standard.
Task Description Score
1. Directions are explicitly stated.
2. Directions are likely to be understood by examines.
3. Directions are likely to produce intended examinee behavior.
4. Critical learning target content/skills are integrated into the task.
5. The task allows for multiple plausible solutions.
6. Expected examinee activities (e.g., individual or group effort, internet research, interviews, etc.)
required to complete the task is explicitly stated.
7. Resources needed to complete the task are explicitly stated.
8 Expected performance or work products are so described to enable examinee understanding.
9. The teacher’s role in relation to the task is fully described.
10. Scoring procedures are so described to enable examinee understanding.
Performance Criteria Score
11. Performance criteria are explicitly stated (i.e., unambiguous).
12. Performance criteria are likely to be understood by examinees.
13. Performance criteria focus on important task dimensions.
14. Performance criteria are logically related to the task.
15. Performance criteria are directly observable, little or no inference making.
Scoring System (Rubric) Score
16. The type of rating scale (holistic or analytical) is appropriate.
17. The scoring procedure is feasible, i.e., workable.
18. The scoring procedure minimizes scoring error.
19. The scoring procedure is likely to be understood by examinees.
20. Rating scale increments (e.g., numerical or qualitative) are suitable.
21. Performance level descriptions are logically related to the task.
22. Performance level descriptions are suitable given the task.
23. Performance level descriptions are likely to be understood by examinees.
Authenticity Score
24. The task is feasible, but reasonably challenging for examinees.
25. The task requires examinees to employ suitable intellectual skills (e.g., analysis, synthesis).
26. The task requires examinees to use apt thinking skills (e.g., creativity, problem solving, etc.).
27. Examinees must apply learning target content and/or skills.
28. The task replicates or simulates an academic, personal, civic, or vocational event, experience,
etc. examinees are likely to encounter.
29. Examinees are required to efficiently and effectively apply learning target content and/or skills
to a suitably complex task.
30. Examinees must rehearse or practice required learning target content and/or skills to
successfully complete the task.
31. Examinees must consult resources (e.g., texts, internet, libraries, people, etc.) to complete the
task.
32. Examinees are provided feedback before “graded” final performance or work product
submission.
33. Examinees must complete the task under suitable constraints (e.g., time, prior knowledge,
resources, etc.).
Total Score (Maximum = 132)
Comments:

Appendix 5.3 Group Contribution Rating Scale
Read each statement carefully. Next, circle either 2 for “Yes” or 1 for “No” if the
indicator (behavior) was demonstrated by the group member or not. The “?” indicates
that the indicator was not observed.
Standard (Evaluative Criterion) No Yes ?

1. The group member’s participation was focused on the task at hand. 1 2 0
2. The group member usually exhibited a respectful demeanor. 1 2 0
3. The group member contributed an acceptable quantity of data, e.g., research 1 2 0

articles, URLs, books, etc., given the team’s task.
4. The quality of the group member’s data (e.g., research articles, URLs, books, 1 2 0
etc.) contribution was high, given the task.
5. The group member’s contribution of data (e.g., research articles, URLs, 1 2 0

books, etc.) was relevant to the team’s task.
6. The group member acceptably met the team’s deadlines. 1 2 0
7. When required, the member exhibited appropriate mediating skills. 1 2 0
8. The member followed team directions in an acceptable manner. 1 2 0
9. The group member exhibited appropriate listening skills which assisted the 1 2 0
team in accomplishing its task.
10. The team member was sufficiently flexible so as to enable the work group to 1 2 0
complete the task at hand.
11. The team member demonstrated writing skills, which helped the work group 1 2 0
meet its objective.
12. By providing constructive feedback to team mates, the member contributed 1 2 0

towards accomplishing the team’s task.
Total Earned Points:_________
Interpretation: The higher the number of points, the higher the perceived contribution level.

Appendix 5.4 Skill Focused Scoring Rubric

Article Review Scoring Rubric
Names ______________________ Date Completed_____________ Total Score_____ of 100
Students will work in groups to review and critique a short article from a peer-reviewed electronic journal which
reports on a single quantitative, qualitative, or mixed-method study. Do not select a literature review or a meta-analysis
article.
The group will critique the work based on evidence of reliability, validity, design suitability, and practical usefulness of
the information. When asked to discuss the suitability and sufficiency of any supporting data, briefly summarize the
data and then critique it using prevailing best practices, professional standards, your text, and other research. Cite
references in an APA style reference list. The task is a three page maximum, excluding title and reference pages. Left –
margin headings: Intent, Type, Reliability/Authority, Validity/Verisimilitude, and Conclusion. The article’s URL
must be provided.
Rating:
Exceptional corresponds to an A (95-100%). Performance is outstanding; significantly above the usual
expectations.
Proficient corresponds to a grade of B to A- (83-94%). Skills and standards are at the level of expectation.
Basic corresponds to a C to B- (75-82%). Skills and standards are acceptable but improvements are needed to
meet expectations well.
Novice corresponds to an F (< 74%). Performance is weak; the skills or standards are not sufficiently
demonstrated at this time.
0 This criterion is missing or not in evidence.
Ratings
Criteria 0 Novice Basic Proficient Exceptional
Intent. The intent of the research is summarized succinctly and 1.0 - 7.4 7.5 – 8.2 8.25 – 9.4 9.5 - 10
thoroughly in a style appropriate to the research design. State the
purpose for which the research was conducted.
Type. State whether the study was primarily or exclusively 1.0 - 7.4 7.5 – 8.2 8.25 – 9.4 9.5 - 10
quantitative, qualitative, or mixed method. Briefly describe the study's
methodology to prove your designation.
Reliability/Authority. For an article which reports on a quantitative 1.0 – 18.4 18.5 – 19.9 20 - 23.4 23.5 - 25
study, describe the data collection device’s internal consistency,
stability, equivalence, or inter-rater reliability; give coefficients if
available.
For an article reporting on a qualitative study, describe the author’s
qualifications and experience, the sponsoring organization, number of
times the article has been cited in other research, cite other researcher’s
opinions of the article, how consistent the reported findings are with
other studies, etc.)
Discuss the suitability and sufficiency of any supporting data.
Validity/Verisimilitude. For an article which reports on a quantitative 1.0 – 18.4 18.5 – 19.9 20 - 23.4 23.5 - 25
study, describe the data collection device’s content, criterion-related,
or construct validity; also comment on the control or prevention of
internal research design threats to internal design validity.
For an article reporting on a qualitative study, show the article’s
verisimilitude (i.e., appearance of truth). Do this by commenting on the
logical analysis used by the authors; describe how consistent their
findings and recommendations are with other researchers; comment on
the consistency of the study's design and execution with other research
on the topic, etc. When assessing logical analysis, consider logic leaps,
the internal consistency of arguments, deductive and inductive
reasoning, etc.
Discuss the suitability and sufficiency of any supporting data.
Conclusion. Provide an overall assessment of research reported in the 1.0 – 11.4 11.5 – 12.4 12.5 – 14.5 - 15
article. Did the study meet prevailing best practices for its research 14.4
design and data collection strategies? Why or why not? Describe the
practical application of findings to professional practice.
Writing and grammar skills are appropriate to the graduate level 1 – 11.1 11.2-12.3 12.4-14.1 14.2-15
(including APA citations and references).
Total Earned Points: ____________

Appendix 5.5 Presentation Rating Rubric
The presentation scoring rubric is presented below and is composed of three subtests or sections: Organization,
Presence, and Technology. Odd number ratings reflect the “mid-point” between two even numbered scores.
A. Execution: Score ____ (out of 48)
1. Introduction: Score ____ (out of 6)

0 – Poor to nonexistent introduction, no “attention grabber”
2 – Vague objectives or simply reads the problem statement
4 -- Good “attention-grabber”, weak objective foundation
4 -- Weak “attention-grabber”, clear objective foundation
6 – Good “attention-grabber”, lays clear foundation of objectives
2. Content: Score ____ (out of 8)

0 – Wrong data, wrong problem, “great” leaps in logic, poor question anticipation
4 – Some understanding of presentation content demonstrated, some incorrect terminology,
a few small leaps in logic, fair question anticipation
6 -- Proficient understanding of presentation content demonstrated, little incorrect terminology,
no leaps in logic, good question anticipation
8-- Exemplary understanding of presentation content demonstrated, no incorrect terminology,
no leaps in logic, excellent question anticipation
3. Questioning: Score ____ (out of 6)

0 – Failed to answer questions, engaged in no discussion
2 -- Answered questions awkwardly, partial explanations, ineffective discussion
4 -- Answered questions somewhat effectively, with accurate explanations, poorly lead discussion(s)
6 -- Answered questions effectively with accurate explanations & effectively lead discussion(s)
4. Communication Strategy: Score ____ (out of 6)

0 -- Inappropriate strategy, ineffectually executed
2 -- Appropriate strategy, but ineffectually executed
4 -- Appropriate strategy, proficiently executed
6 -- Appropriate strategy, expertly executed
5. Use of Notes: Score ____ (out of 6)

0 -- Relies heavily on notes or prepared text, lost place several times
2 -- Relies moderately on notes or prepared text, lost place some times
4 -- Relies little on notes or prepared text, lost place once or twice
6 -- Doesn’t rely on notes or prepared text, lost place only once
6. Conclusion and Wrap-Up: Score ____ (out of 8)

0 – Poor to nonexistent conclusion
2 – Simple re-hash of main point(s), and “not tied together”
4 – Main point(s) clarified, implications weakly presented, & “tied together”
6 --Main point(s) clarified, implications reasonably well presented & “tied together”
8 --Main point(s) clarified, implications well presented & “tied together”
7. Professional Impression: Score ____ (out of 8)

0 -- “Slip-shod” appearance
2-- Amateurish in appearance
4 -- Less amateurish in appearance
6 -- Proficient in appearance
8 -- Expert in appearance
B. Presence: Score ____ (out of 24)
1. Eye Contact: Score ____ (out of 4)

0 – Makes little eye contact,
2 – Makes moderate eye contact, focuses on one group or side of the room
4 – Makes and holds eye contact with people all over the room,

2. Use of Hands and Body Movement: Score ____ (out of 4)

0 – Distracts or annoys audience or gives perception of being nervous
2 – Somewhat comfortable but movements interrupt flow of presentation
4 –Completely comfortable, appropriate hand-gestures, and non-awkward movements
3. Voice & Inflection: Score ____ (out of 6)

0 – Too hard to hear, sounds disinterested, did not project voice
2 – Speaks in a monotone, sounds disinterested, or many “ums” and “likes”, inconsistent voice projection
4 -- Varies voice and inflection appropriately, conveys some enthusiasm, fairly consistent voice projection
6 – Varies voice and inflection expertly, conveys enthusiasm, appropriate voice projection
4. Articulation & Pace: Score ____ (out of 4)

0 -- Poorly articulated words and/or sentences, very distracting speaking pace, many awkward pauses
2 -- Mispronounces some words and/or “mangles” sentences, inconsistent speaking pace, few awkward
pauses
4 -- Articulates words and sentences clearly, speaking pace appropriate, no awkward pauses
5. Professional Appearance: Score ____ (out of 6)

0 -- Very inappropriately attired for presentation subject, audience, and/or environment
2 -- Somewhat inappropriately attired for presentation subject, audience, and/or environment
4 -- Appropriately attired for presentation subject, audience, and/or environment
6 -- Very appropriately attired for presentation subject, audience, and/or environment
C. Technology: Score ____ (out of 28)
1. Slides, Graphics, Figures, etc. Layout: Score ____ (out of 4)

0 -- Visual aids are poorly designed, cluttered, & many have missing labels
2 -- Visual aids are sometimes difficult to read; some slides are cluttered, and labeling was inconsistent.
4 -- Visual aids easy to read, uncluttered, and fully labeled
2. Slides, Graphics, Figures Color: Score ____ (out of 4)

0 – Visual aid colors distracting and confusing for presentation
2 -- Visual aid color appropriateness inconsistent for presentation
4 -- Visual aid colors appropriate for presentation
3. Slide, etc. /Presenter Alignment: Score ____ (out of 4)

0—Visual aids and presenter were frequently out-of-alignment
2—Visual aids and presenter were occasionally out-of-alignment
4—Visual aids and presenter were rarely out-of-alignment
4. Slide, etc. Reading: Score ____ (out of 4)

0—Presenter frequently read slides to audience
2—Presenter occasionally read slides to audience
4—Presenter rarely read slides to audience
5. Slide Presentation Support: Score: _____ (out of 4)

0—Visual aids ineffective & hard to follow
2— Visual aids moderately effective & easy to follow
4— Visual aids effective & easy to follow
6. General Usage: Score ____ (out of 4)

0 -- Visual aids lacking or poorly utilized, very distracting
2 -- Visual aid effectiveness inconsistent and distracting, at points, during presentation
4 -- Visual aids utilized effectively and not distracting
7. Usage Impact: Score ____ (out of 4) Total Score: ________

0 -- Learning not enhanced
2 -- Learning enhancement inconsistent
4 -- Learning enhanced

Appendix 5.6: Preparing Examinees for a Testing Session
A. Preparing the Testing Environment

1. Each time an examination is administered, the testing environment should be
similar to that of a nationally standardized test, unless a performance
assessment is administered; modify these guidelines as appropriate.
a. The room should be clean, uncluttered, and ordered.
b. The temperature should be comfortable, adequately lighted, and well
ventilated.
c. Test materials should be ready and in sufficient number.
d. Distractions such as noise or unnecessary movement should be avoided.
e. Discipline should be maintained.
f. Answer individual questions about items carefully. This can be
substantially reduced if language and item construction are
developmentally appropriate.
2. Examinees are going to be anxious and some will be frustrated; reassure as

best as is possible. Breathing exercises may be helpful as will the progressive
relaxation of muscles. If an examinee’s text anxiety appears to be chronically
acute, refer the examinee for learning disability evaluation.
3. Special Learning Assessment Applications

a. Open book tests are helpful if the testing emphasis is on application, not
memorization.
b. Unannounced tests are not recommended as examinees tend to under-
perform due to anxiety and inadequate preparation time. Study time + test
preparation = high scores.
c. For summative evaluation, selection, and placement, single test
administrations are sufficient. For diagnostic and formative evaluation,
frequent testing is recommended. Frequent testing motivates examinees
and improves scoring.
B. Ethical Test Preparation

1. Test-preparation training tends to produce higher test scores. To improve test-
taking skills,
a. Examinees need to understand the mechanics of test-taking, such as the
need to carefully follow instructions, checking their work, and so forth.
b. Examinees should use appropriate test-taking strategies, including ways in
which test items should be addressed and how to make educated guesses.
c. Examinees need to practice test-taking skills.
2. Successful examinees tend to understand the purpose for testing, know how
the results will be used, comprehend the test’s importance and its relevance to
teaching and learning, expect to score well, and have confidence in their
abilities.

3. Test preparation is effective. A critical question is, “When does test

preparation become teaching to the test?”
a. Mehrens and Kaminski (1989) suggest that providing general instruction
on content and performance objectives without specific reference to those
tested by an intended achievement (often, a standardized) test, and the
teaching of test taking skills, is ethical.
b. They argue that:
(1) Teaching content and performance standards which are common to
many achievement tests (standardized or locally developed);
(2) Teaching content and performance standards which specifically match
those on the test to be administered; and
(3) Teaching to specifically matched content and performance standards
and where practice follows the same item formats as found on the test
to be administered is a matter of professional opinion.
c. However, provision of practice or instruction on a parallel form of the test
to be administered or on the actual test itself is unethical (Mehrens &
Kaminski, 1989).
d. Clearly, the “cut” lies between b(1), b(2), or b(3). Two guiding principles
should be employed.
(1) The justification for teaching test taking skills is that such instruction
reduces the presence of random error in an examinee’s test score so
that a more accurate measure of learning is taken. It seems that the
teaching of specific content known to be included on a test to be
administered and which is not part of the student’s regular
instructional program is teaching to the test and thus artificially
inflating the examinee’s test score. There should be very tight
alignment between any standardized test and the regular instructional
program or the standardized test should not be used to measure
achievement.
(2) The second guideline is the type of test score inference to be drawn.
Mehrens & Kaminski (1989) write, “the only reasonable, direct
inference you can make from a test score is the degree to which a
student knows the content that the test samples. Any inference about
why the student knows that content...is clearly a weaker inference.”
(a) Teaching to the test involves teaching specific content which
weakens the direct inference about what the examinee knows and
can do. Testing is done to generalize to a fairly broad domain of
knowledge and skills, not the specific content and item format
presented on a specific test.
(b) When one seeks to infer from a test score why an examinee knows
what he or she knows, an indirect inference is being made. Indirect
inferences are dangerous and often lead to interpretation errors.
e. Applying these two guidelines leads one to conclude that
(1) Providing general instruction on content and performance objectives
without reference to those tested by an intended achievement (often, a
standardized) test;

(2) Teaching test taking skills; and/or

(3) Teaching content and performance standards which are common to
many achievement tests (standardized or locally developed) are
ethical.
C. Test Preparation Principles Performed by Teachers for Students

1. Specific Classroom Recommendations are:
a. If practice tests are used in the examinee’s classroom, make it a learning
experience. Explain why the correct answer is correct and why the
incorrect answers are incorrect. Use brainstorming and other strategies
that promote a diversity of responses. Examinees should develop
questions for class discussions and practice tests. Have the class (as in
small groups) identify content from the text, notes, supplemental learning
materials etc. which points towards the correct answer.
b. Incorporate all intellectual skills into daily activities, assignments, class
discussion, homework, and tests. Teach examinees the different
intellectual skills; help them learn to classify course or class activities and
questions by intellectual skill so that the various categories of thinking
(recall, analysis, comparison, inference, and evaluation) are learned and
practiced.
c. Encourage examinees to explain their thinking, i.e., how they arrived at
their answer, conclusion, or opinion. Practice discerning fact from opinion
and the relevant from the irrelevant. Practice looking for relationships
among ideas by identifying common threads. Have examinees solve
verbal analogies, logic puzzles, and other classification problems.
d. Apply learned information to new and different situations or issues.
Encourage application of information by asking examinees to relate what
has been learned to their own experiences.
e. Ask open-ended questions which ensure that examinees do not assume
that there is one correct answer.
f. Assign time limits to classroom work and structure assignments, quizzes,
or tests in formats similar to those found on standardized tests.
2. Improving Examinee Motivation Recommendations

a. Expect good results and model a positive attitude.
b. Use appropriate motivational activities and provide appropriate incentives.
c. Discuss the test’s purpose and relevance and how scores are to be used.
d. Discuss with examinees the testing dates, times, content, time available to
complete the test(s), test item format, and length of reading passages if
any.
e. Know and use correct test administration procedures while ensuring a
quiet, orderly testing environment.
f. Ensure that all skills to be tested have been taught and practiced to
proficiency. Ensure that examinees know they have adequately prepared
for the testing experience. Test-preparation is not a substitute for thorough,
diligent preparation.

D. Test Preparation Principles Performed by Students

1. List the major topics for the entire course (if a comprehensive midterm or
final) or for chapter(s) which contribute test content. Review all of relevant
source material: textbook(s), notes, handouts, outside readings, returned
quizzes, etc.
2. For each topic, in an outline, summarize key or critical information. It will

also be helpful to construct a map of the major concepts so that you will note
connections between key concepts.
3. Plan study time. Study material with which you have the most difficulty.
Move on to more familiar or easier material after mastering the more difficult
content. However, balance time allocations so that there is sufficient time to
review all critical or key material before the test.
4. Study material in a sequenced fashion over time. Allocate three or four hours
each day up to the night before the test. Spaced review is more effective
because you repeatedly review content which increases retention and builds
content connections. “Don’t cram.” Develop a study schedule and stick to it.
Rereading difficult content will help a student learn. Tutoring can be a
valuable study asset.
5. Frame questions which are likely to be on the test in the item format likely to
be found. If an instructor has almost always used multiple choice items on
prior tests or quizzes, it is likely that he or she will continue to use that item
format. Anticipate questions an examiner might ask. For example:
a. For introductory courses, there will likely to be many terms to memorize.
These terms form the disciplinary vocabulary. In addition to being able to
define such terms, the examinee might need to apply or identify the terms.
b. It is likely the examinee will need to apply, compare and contrast terms
concepts or ideas, apply formulae, solve problems, construct procedures,
evaluate a position or opinion, etc. Most likely, the examinee will need to
show analytical and synthesis skills.
6. Form Study Groups

a. Study groups tend to be effective in preparing examinees provided the
group is properly managed. Time spent studying does not necessarily
translate into leaning, and hence improved test scores.
b. Study group benefits include shared resources, multiple perspectives,
mutual support and encouragement, and a sharing of class learning aides
(e.g., notes, chapter outlines, summaries, etc.).
c. To form a study group, seek out dedicated students (i.e., those who ask
questions in class and take notes), meet a few times to ascertain whether or
not the group will work well together, and keep the group limited to five
or six members.

d. Once started, meet regularly and focus on one subject or project at a time,
have an agenda for each study session, ensure that logistics remain
“worked out” and fair, follow an agreed upon meeting format, include a
time-limited open discussion of the topic, and brainstorm possible test
questions.
7. Other Recommendations
a. For machine-graded multiple-choice tests, ensure the selected answer
corresponds to the question the examinee intends to answer.
b. If using an answer sheet and test booklet, check the test booklet against the
answer sheet whenever starting a new page, column, or section.
c. Read test directions very carefully.
d. If efficient use of time is critical, quickly review the test and organize a
mental test completion schedule. Check to ensure that when one-quarter
of the available time is used, the examinee is one-quarter through the test.
e. Don't waste time reflecting on difficult-to-answer questions. Guess if there
is no correction for guessing; but it is better to mark the item and return to
it later if time allows, as other test items might cue you to the correct
answer.
f. Don't read more complexity into test items than is presented. Simple test
items almost always require simple answers.
g. Ask the examiner to clarify an item if needed, unless explicitly forbidden.
h. If the test is completed and time remains, review answers, especially those
which were guessed at or not known.
i. Changing answers may produce higher test scores, if the examinee is
reasonably sure that the revised answer is correct.
j. Write down formula equations, critical facts, etc. in the margin of the test
before answering items.
E. Strategies for Taking Multiple Choice Tests

1. Read the question. Advise examinees to think of the answer first as he or she
reads the stem. By thinking of the answer first, he or she is less likely to be
fooled by an incorrect answer. However, read all answer options before
selecting an answer.
2. Do not spend too much time on any one question, as the item may actually
have two correct answers, instead of one or no correct answer.
a. In these cases select an answer at random, provided there is no penalty for
guessing and wrong answers are not counted.
b. Advise the examinee to circle the question number so he or she can go
later if there is time. Go on to the next item.
3. If an examinee doesn’t know the answer, then he or she should mark out
options which are known to be incorrect. This increases the chances of a
correct guess. Many multiple choice items, have only two plausible answer
options, regardless of the number presented.
4. Do not keep changing answers. If the item seems to have two correct answers,
select the best option and move along.

a. Based on testing research, the first answer is probably correct. An

examinee is most likely to change a correct answer to an incorrect one.
b. Only change an answer if absolutely sure a mistake was made.
5. After finishing the test, go back to circled items.
a. Don’t leave a testing session early, unless absolutely sure each item is
correctly answered. Invest what time is available, to answer unanswered
items.
b. If an item is still not able to be answered, guess. You have a 25% chance
of selecting the correct answer, if four-options are presented. The chances
of a correct guess are higher if other answer options can be eliminated.
6. If the test is not properly constructed, the following “tricks” might raise
scores:
a. Where the examinee must complete a sentence, select the one option that
fits better grammatically.
b. Answer options which repeat key words from the stem are likely correct.
c. Longer answer options tend to be correct more often than incorrect.
d. If there is only one correct answer to an item, that answer is likely to be
different. Thus if two or three answer options mean the same, they must
be incorrect.
e. If guessing and a typing error is noted in an answer option, select another
option.
7. If two answer options have the same meaning, neither are likely to be correct.
8. If two answer options have opposite meanings, one is usually is correct.
9. The one answer which is more general than the others is usually the right
answer.
10. For an occasional test item whose answer options end with “all the above”;
select “all of the above”. If one answer option is incorrect, then “all of the
above” cannot be a correct option.
11. The answer option which is more inclusive, (i.e., contains information which
is also presented in other answer options), is likely to be the correct.
12. When you have studied and don’t know the answer, select “C” if there is no
guessing penalty.
13. If you do not lose points for incorrect answers, consider these guidelines for
making an educated guess:
a. If two answers are similar, save for a couple of words, select one of them.
b. If a sentence completion stem, eliminate any possible answer which would
not form a grammatically correct sentence.
c. If numerical answer options cover a wide range, choose a number in the
middle.
14. Answer options containing always, never, necessarily, only, must, completely,
totally, etc., tend to be incorrect.
15. Answer options which present carefully crafted statements incorporating such
qualifiers as often, sometimes, perhaps, may and generally, tend to be correct.

F. Strategies for Answering Other Test Item Formats

1. True-False Tests
a. If any part of a statement is false, the answer is false.
b. Items containing absolute qualifiers, e.g., always or never, often are false.
2. Open Book Tests
a. Write down any formulas you will need on a separate sheet.
b. Place tabs on critical book pages.
c. If using notes, number each page and make a table of contents.
d. Prepare thoroughly; these types of examinations are often very difficult.
3. Short Answer/Fill-in-the-Blank
a. These test items, ask examinees to provide definitions (e.g., a few words)
or short descriptions in a sentence or two.
b. Use flashcards with important terms and phrases, missing or highlighted
when studying. Key words and facts will be familiar and easy to recall.
4. Essay Tests
a. Decide precisely what the question is asking. If a question asks you to
contrast, do not interpret.
b. If an examinee doesn’t know an answer or if testing time is short, usually
it's a good idea to answer those items that the examinee knows the answers
to first. Then sort based on a combination of “best guess” and available
points. Attempt to answer those items worth the most points and for which
the examinee has the greatest amount of knowledge.
c. Verbs used in essays include: analyze, compare, contrast, criticize, define,
describe, discuss, evaluate, explain, interpret, list, outline, prove,
summarize. Look up any unfamiliar words in a dictionary.
d. Before writing, make a brief, but quick outline.
(1) Thoughts will be more organized and the examinee is less likely to
omit key facts and/or thoughts.
(2) The examinee will write faster and may earn some points with the
outline, if he or she runs out of time.
(3) Points are often lost as the examiner has little understanding of an
examinee’s response due to its poor organization. Use headings or
numbers to guide the reader.
e. Leave plenty of space between answers. The extra space is needed to add
information if time is available.
f. When you write, get to the point. Start off by including part of the
question in your answer to help focus your response.
(1) Build upon your answer with supporting ideas and facts.
(2) Review your answers for proper grammar, clarity and legibility.
g. Don’t pad answers; this tends to irritate examiners. Be clear, concise, and
to the point. Use technical language appropriately.
References
Mehrens, W. A. & Kaminski, J. (1989). Methods for improving standardized test scores:
Fruitful, fruitless, or fraudulent? Educational Measurement: Issues and Practice, 8,
(1), 14-22.

Chapter 5 Constructing Tests and Performance Assessments

Uploaded by

Copyright:

Available Formats

Chapter 5 Constructing Tests and Performance Assessments

Uploaded by

Document Information

Original Description:

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Chapter 5 Constructing Tests and Performance Assessments

Uploaded by

Copyright:

Available Formats

Chapter 5 Constructing Tests and Performance Assessments 123

Chapter 5 Constructing Tests and Performance Assessments

In a “standards based” approach to education and training, informed by Constructivist

Measuring Learning & Performance: A Primer | Retrieved from CharlesDennisHale.org

2. Typical performance test (TPT): These tests include personality, interest,

We focus on achievement testing which is used most to measure student or trainee

2. Second, it is necessary that the content, intellectual or thinking skills (see

4. Fourth, a table of specifications (or test blueprint) is developed which integrates

Measuring Learning & Performance: A Primer | Retrieved from CharlesDennisHale.org

5. Fifth, once steps 1 to 4 are completed, the plan is initially reviewed by

In this chapter, we will first examine “mechanics of achievement test construction”

I. The Mechanics of Achievement Test Construction

Measuring Learning & Performance: A Primer | Retrieved from CharlesDennisHale.org

Measuring Learning & Performance: A Primer | Retrieved from CharlesDennisHale.org

learners have a fair chance to meet Content and Performance

3. The Integrated Model

Quellmalz (1987) Bloom, et al. (1956)

For instance, according to Stiggins, Griswold, and Wikelund (1989),

Measuring Learning & Performance: A Primer | Retrieved from CharlesDennisHale.org

b. Construct an Operational Definition, called a benchmark for each Attitude,

c. Once a sufficient number of standards, in depth and breadth, have been

Measuring Learning & Performance: A Primer | Retrieved from CharlesDennisHale.org

B. Constructing a Test Blueprint

Measuring Learning & Performance: A Primer | Retrieved from CharlesDennisHale.org

demonstrated by reading articles on a topic and then assembling the

4. Application: Table 5.1

Measuring Learning & Performance: A Primer | Retrieved from CharlesDennisHale.org

Selects from a variety of simple strategies, including the use

Uses simple strategies to determine meaning and increase

Clarifies understanding by rereading, self-correction,

Measuring Learning & Performance: A Primer | Retrieved from CharlesDennisHale.org

D. Directions and Arranging Test Items

2. Arranging Items on the Test Instrument

E. Timing and Testing

2. For speeded tests, an examinee’s score is materially influenced by the number

Measuring Learning & Performance: A Primer | Retrieved from CharlesDennisHale.org

3. For power tests (i.e. achievement testing), an examinee’s score is influenced

F. Interpreting Test Scores

Measuring Learning & Performance: A Primer | Retrieved from CharlesDennisHale.org

4. Test items are either dichotomously or polychotomously scored.

Measuring Learning & Performance: A Primer | Retrieved from CharlesDennisHale.org

5. Allocating Test Points

Measuring Learning & Performance: A Primer | Retrieved from CharlesDennisHale.org

warranted. However, if ether the total test or any subtest score is

d. For a direct performance assessment, each assignment has a task

6. Assigning a Course or Training Program Grade

Measuring Learning & Performance: A Primer | Retrieved from CharlesDennisHale.org

II. Constructing Select Response Items to Assess Knowledge, Comprehension, &

B. Writing Multiple-Choice Items

(2) Correct Answer

Measuring Learning & Performance: A Primer | Retrieved from CharlesDennisHale.org

c. Rearrangement: This item is can be used to assess examinee procedural

____1. The intellectual skills, attitudes, and/or psychomotor are identified.

d. Substitution Item: In this variation, the item stem contains a blank or

Reliability and Validity Relationship

(2) Answers: 1c; 2e

Measuring Learning & Performance: A Primer | Retrieved from CharlesDennisHale.org

2. Strengths and limitations of multiple choice items are:

Measuring Learning & Performance: A Primer | Retrieved from CharlesDennisHale.org