PLAGIARISM FREE "A" WORK 18 HOURS or LESS

NeNe1994

Chapter10InstructionallySupportiveStandardsBasedTests.pdf

Home >Applied Sciences homework help >PLAGIARISM FREE "A" WORK 18 HOURS or LESS

1 3 9

IN THIS CHAPTER AND THE NEXT, I’LL BE CONSIDERING ONE OF THE MOST SIG-

nificant links between instruction and assessment: the role that test

results can play in the evaluation of a teacher’s instructional effec-

tiveness. In the previous chapter, you saw why the results of tradi-

tionally constructed standardized achievement tests should never be

used to judge teachers’ instructional effectiveness. Well, if those tests

aren’t the answer, what is?

Making the Best of the Situation . . . and Seeking the Best for the Students Some educators long for a return to those tranquil days of yesteryear

when there were no educational accountability programs and no

annual achievement tests with results employed to evaluate educa-

tors. In my view, that sort of yearning is a definite no-yield yearn.

Instead of resisting today’s test-based evaluative pressures, we ought

to be redirecting those pressures into processes that help children

receive a better education.

Let’s be realistic. Educational accountability programs aren’t going

to evaporate in the near or distant future. The public has serious

10 Instructionally

Supportive Standards-Based Tests

Popham-Teach Better ch10.qxd 8/6/2003 2:51 PM Page 139

Popham, W. James. Test Better, Teach Better : The Instructional Role of Assessment, Association for Supervision & Curriculum Development, 2003. ProQuest Ebook Central, http://ebookcentral.proquest.com/lib/amridge/detail.action?docID=5704436. Created from amridge on 2022-01-22 03:15:42.

C o p yr

ig h t ©

2 0 0 3 . A

ss o ci

a tio

n f o r

S u p e rv

is io

n &

C u rr

ic u lu

m D

e ve

lo p m

e n t. A

ll ri g h ts

r e se

rv e d .

T E S T B E T T E R , T E A C H B E T T E R1 4 0

doubts about whether educators are doing an adequate job of teach-

ing the nation’s children. What’s more, people want hard evidence,

not verbal reassurance, that classroom teachers are delivering high-

quality instruction. And the evidence forming the core of all educa-

tional accountability programs will continue to be students’ perform-

ance on tests that measure mastery of important skills and know-

ledge. Moreover, those tests will be standardized to prevent educators

from collecting evidence via a set of possibly suspect “home-grown”

tests. No, the public wants credible, test-based evidence of what stu-

dents have learned in school. Fine! Let’s supply them with that kind

of evidence. But let’s do it in a way that helps teachers to help

students.

Standards-Based Achievement Tests Many states now evaluate public school students through custom-

built statewide tests, administered each year. Typically, these tests are

constructed to be aligned with a state’s curricular goals (its set of offi-

cial content standards). Accordingly, these statewide standardized

achievement tests are usually referred to as standards-based tests.

As I’ve noted, many of these tests have (sadly) been constructed

in precisely the same way that traditional standardized achievement

tests are built, with the result that they are similarly score-spread

focused and thus unsuitable for evaluating educational quality.

What’s more, these customized, standards-based achievement tests

offer no more instructional benefits to teachers than traditional stan-

dardized tests do. What we need are untraditional standardized

achievement tests—ones that do not rely on norm-referenced com-

parisons but that still provide the evaluative evidence that the archi-

tects of educational accountability programs demand. The trick,

then, is to install standardized tests that do a decent accountability

job but also help teachers do a better job of instructing students. This

can be done.

Popham-Teach Better ch10.qxd 8/6/2003 2:51 PM Page 140

C o p yr

ig h t ©

2 0 0 3 . A

ss o ci

a tio

n f o r

S u p e rv

is io

n &

C u rr

ic u lu

m D

e ve

lo p m

e n t. A

ll ri g h ts

r e se

rv e d .

I n s t r u c t i o n a l l y S u p p o r t i v e S t a n d a r d s - B a s e d T e s t s 1 4 1

What Makes a Standards-Based Achievement Test Instructionally Supportive? Before we look at the three features that an accountability focused

standardized test must have in order to help teachers boost instruc-

tion, I need to explain how to distinguish between tests that yield

norm-referenced interpretations and tests that yield criterion-referenced

interpretations.

As you read in the previous chapter, when a test’s results allow a

student’s score to be “referenced” back to the scores of a norm group,

we call this a norm-referenced interpretation or a norm-referenced inference.

If a test’s results allow us to “reference” a student’s score back to a

clearly defined skill or body of knowledge, this is called a criterion-

referenced interpretation or a criterion-referenced inference. A norm-

referenced interpretation might be something like this: “Jamal scored

at the 93rd percentile on the standardized achievement test.” A

criterion-referenced interpretation might be “Hillary’s test score indi-

cated she had mastered 85 percent of the 500 words in the ‘spelling

demons’ word list.”

Please note that it is not a test that is norm-referenced or criterion-

referenced, but the score-based inference. Indeed, a test can be specif-

ically built to provide both norm- and criterion-referenced interpre-

tations. What people usually mean when they talk about a “norm-ref-

erenced test” is that the test is built so it typically yields norm-refer-

enced inferences about test-takers.

Well, the kind of instructionally supportive standardized achieve-

ment test educators need to both boost instructional quality and sup-

ply accountability evidence must be constructed to provide mean-

ingful criterion-referenced interpretations. This is essential, because it

means that test developers are off the hook when it comes to score-

spread. If no one is interested in comparing test-takers with the norm

group, then it’s no longer necessary to include the score-spreading

Popham-Teach Better ch10.qxd 8/6/2003 2:51 PM Page 141

C o p yr

ig h t ©

2 0 0 3 . A

ss o ci

a tio

n f o r

S u p e rv

is io

n &

C u rr

ic u lu

m D

e ve

lo p m

e n t. A

ll ri g h ts

r e se

rv e d .

T E S T B E T T E R , T E A C H B E T T E R1 4 2

items. What the test-makers must do, however, is create a standard-

ized achievement test that will provide reasonably accurate evidence

about children’s mastery of particular content standards. And that

brings us to the first characteristic of an instructionally supportive

standardized achievement test.

Standard-by-Standard Reporting An instructionally supportive standardized achievement test must

provide standard-by-standard results for each student who is tested.

If teachers have this kind of per-standard information for individual

students, they can correlate the achievement test’s results with what

they already know about particular students’ strengths and weak-

nesses. They can then use this combined information to reconsider

the merits of their instructional approaches. For instance: “James,

Sarita, Bisham, and Jack all struggle with mathematics. They all

scored well on Math Standards 1, 2, and 4, where I used Instructional

Approach A; they all scored poorly on Standard 3, where I did not.

Perhaps next year, I should try to employ Instructional Approach A

with Standard 3.” Unless teachers discover which content standards

have or haven’t been mastered by students, they cannot know which

elements of their instruction to retain and which to retool.

Currently, few states’ standards-based tests yield per-standard

results for students. And yet, these statewide tests supposedly serve as

the cornerstones of a “standards-based reform” strategy intended to

improve a state’s public education. What teachers in most states are

given, after each year’s tests have been scored, is some sort of global

report indicating how a group of students have done generally with

respect to the state’s total array of official content standards. A

statewide test might provide teachers with the following sort of

report: “Based on our state’s mathematics test, on the 6 content stan-

dards related to statistics, 37 percent of your students scored at the

advanced level, 23 percent scored at the proficient level, 24 percent

scored at the basic level, and 16 percent scored at the below basic

Popham-Teach Better ch10.qxd 8/6/2003 2:51 PM Page 142

C o p yr

ig h t ©

2 0 0 3 . A

ss o ci

a tio

n f o r

S u p e rv

is io

n &

C u rr

ic u lu

m D

e ve

lo p m

e n t. A

ll ri g h ts

r e se

rv e d .

I n s t r u c t i o n a l l y S u p p o r t i v e S t a n d a r d s - B a s e d T e s t s 1 4 3

level.” Is it any wonder that teachers do not regard such standards-

based tests as instructionally helpful?

Standard-by-standard reporting would not only benefit teachers,

but also parents, who would receive specific information about their

child’s mastery of key content standards. And, of course, there are the

students themselves. Wouldn’t your students benefit if they could

each discover which standards they had mastered? Standard-by-stan-

dard reporting is indispensable if a standards-based test is going to

make a genuine contribution to improved instruction.

Accurate Assessment of Only the Highest Priority Content Standards The second characteristic of an instructionally supportive standards-

based test proceeds necessarily from the first. If a standards-based test

is to supply standard-by-standard results, it must contain a sufficient

number of items per content standard to allow for accurate inferences

about students’ per-standard mastery. However, allowing for more

items per content standard will mean that tests will be able to assess

fewer standards overall.

Presumably, then, an instructionally supportive standards-based

test that must be administered within a two-hour period might well

end up assessing only a half-dozen or so content standards. It is there-

fore critical that all of the assessed standards be truly significant. This

will require the architects of standards-based tests to prioritize their

state’s official content standards so that every test form always assesses

the most important standards. From an instructional perspective, it is

better for tests to measure a handful of powerful skills accurately than

it is for tests to do an inaccurate job of measuring many skills.

Ideally, the standards at the top of the test-developers’ prioritized

list will be those that embody significant cognitive skills that sub-

sume lesser subskills and bodies of knowledge. A good example of

such a powerful skill is students’ ability to compose original essays of

various types—persuasive essays, for example. When educators assess

students’ mastery of this skill by having students create writing

Popham-Teach Better ch10.qxd 8/6/2003 2:51 PM Page 143

C o p yr

ig h t ©

2 0 0 3 . A

ss o ci

a tio

n f o r

S u p e rv

is io

n &

C u rr

ic u lu

m D

e ve

lo p m

e n t. A

ll ri g h ts

r e se

rv e d .

T E S T B E T T E R , T E A C H B E T T E R1 4 4

samples, those samples can be evaluated overall (using holistic

rubrics) and with respect to a variety of subskills such as organization

and mechanics (using analytic rubrics).

And students’ mastery of these high-import skills can definitely

be used in educational accountability programs to help evaluate

schools or districts. For example, suppose that students’ performances

on a writing-skills test had been categorized as advanced, proficient,

basic, and below basic. It would be quite straightforward to calculate

the percentages of a district’s students who scored in each of those

four performance levels, then compare districts on that basis.

(Indeed, the No Child Left Behind Act requires schools, districts, and

states to set at least basic, proficient, and advanced performance stan-

dards and demonstrate yearly increases in the proportion of students

deemed proficient or above. It’s easy to imagine comparisons based

on these data.) In short, the kinds of tests I’m describing here would

help teachers teach better and still produce the evidence necessary to

make accountability programs function in the long-term best inter-

ests of children.

The major drawback to the per-standard assessment of high-

priority content standards is the risk that pervasive pressure to

improve students’ test performances will lead schools to give too lit-

tle instructional attention to lower-priority content standards not

assessed on the statewide standards-based test. This kind of narrow-

ing of what’s taught will inevitably short-change our students.

To counter such tendencies toward curricular reductionism,

teachers should demand that district or state authorities provide an

array of optional classroom assessment instruments. Such instru-

ments could be used to measure students’ mastery of those state-

approved content standards that, although worthwhile for children

to master, are not assessed by a statewide standards-based test. In

addition, educators at the school, district, and state levels should

come up with ways of monitoring curricular breadth so that those

content standards (and those subject areas) not assessed on a

Popham-Teach Better ch10.qxd 8/6/2003 2:51 PM Page 144

C o p yr

ig h t ©

2 0 0 3 . A

ss o ci

a tio

n f o r

S u p e rv

is io

n &

C u rr

ic u lu

m D

e ve

lo p m

e n t. A

ll ri g h ts

r e se

rv e d .

I n s t r u c t i o n a l l y S u p p o r t i v e S t a n d a r d s - B a s e d T e s t s 1 4 5

statewide test will still receive instructional and assessment attention

in the state’s classrooms.

Clear Assessment Descriptions for Each Measured Content Standard Finally, the skills and knowledge being assessed by an instructionally

supportive standards-based test must be described well enough so

that teachers truly understand what they must teach to promote mas-

tery. Therefore, each of the high-priority content standards assessed

must be accompanied by an assessment description that sets forth, in

two or three teacher-friendly paragraphs, just what a student needs to

be able to do cognitively to perform well on the items measuring a

particular content standard. These brief assessment descriptions

would clarify for teachers what sorts of cognitive demands a given

content standard’s items will impose on students. Several illustrative,

but nonexhaustive sample items should accompany each assessment

description. Necessarily, all samples for a particular content standard

would impose the same cognitive demand. Ideally, they should be

different item-types, to promote generalizable mastery.

The availability of assessment descriptions for each assessed con-

tent standard would make it possible for teachers to aim their instruc-

tion toward students’ mastery of the skills and knowledge represented

by a standards-based test, not toward the test itself. And that, we

know, is just how the instruction game is supposed to be played. If

teachers aim their instruction only at a set of particular test items,

then test-scores may improve, but students’ mastery of what the test

represents surely won’t.

To review, then, the three key features of instructionally support-

ive accountability tests* are (1) standard-by-standard reporting, (2) accu-

rate assessment of only the highest priority content standards, and (3) clear

assessment description for each measured content standard.

*For a more detailed consideration of how instructionally supportive accountability tests can be constructed, see the October 2001 report of the Commission on Instructionally Supportive Assessment, Building Tests That Support Instruction and Accountability: A Guide for Policymakers. Available online at www.aasa.org, www.naesp.org, www.principals.org, www.nea.org, and www.nmsa.org.

Popham-Teach Better ch10.qxd 8/6/2003 2:51 PM Page 145

C o p yr

ig h t ©

2 0 0 3 . A

ss o ci

a tio

n f o r

S u p e rv

is io

n &

C u rr

ic u lu

m D

e ve

lo p m

e n t. A

ll ri g h ts

r e se

rv e d .

http://www.aasa.org

http://www.naesp.org

http://www.principals.org

http://www.nea.org

http://www.nmsa.org

T E S T B E T T E R , T E A C H B E T T E R1 4 6

What Could Be To give you an idea of what a statewide accountability test might look

like if it were constructed along the instructionally useful lines I rec-

ommend, I’ve provided a brief illustration in Figure 10.1.

The accountability-focused testing program described in the fig-

ure is fictional, but there’s no reason that it couldn’t be a reality. I

hope I’ve convinced you that it is possible to create high-stakes,

statewide tests that not only provide credible accountability evi-

dence, but also help teachers do a better job of instructing their

students.

So what can you, a teacher, do to make sure your schools use

instructionally supportive tests for their accountability programs?

Well, you can exert whatever influence you have, perhaps through

your professional association, to see that such tests are made avail-

able. Educational policymakers at every level must first be educated

about these assessment issues, then urged to endorse the use of

instructionally supportive standards-based tests.

If you prefer a less activist role, then set out to do a solid job of

educating your own colleagues about why it is that traditional stan-

dardized achievement tests ought not be used to evaluate schools.

Then let your colleagues know why it is that instructionally support-

ive tests should be used for that purpose. And don’t forget to pass on

information about these issues to your students’ parents. Parents of

school-age children have a major stake in seeing that the standard-

ized tests used with their children will help, not harm, the instruction

those children receive.

Popham-Teach Better ch10.qxd 8/6/2003 2:51 PM Page 146

C o p yr

ig h t ©

2 0 0 3 . A

ss o ci

a tio

n f o r

S u p e rv

is io

n &

C u rr

ic u lu

m D

e ve

lo p m

e n t. A

ll ri g h ts

r e se

rv e d .

I n s t r u c t i o n a l l y S u p p o r t i v e S t a n d a r d s - B a s e d T e s t s 1 4 7

1 0 . 1 SAMPLE DESCRIPTION OF A STATE-LEVEL, INSTRUCTIONALLY SUPPORTIVE ACCOUNTABILITY TESTING PROGRAM

Overview. State X’s accountability tests, administered annually in grades 3–8 and grade 10, each assess the state’s most significant content standards.

Content assessed. The state has identified approximately 100 content standards in every subject that is assessed. However, committees of state educators and non- educator citizens have prioritized those content standards so that each grade level’s accountability test assesses between five and nine of the highest-priority standards. The state’s curriculum authorities made a serious effort to devise a defensible grade- to-grade articulation for each content standard across the grade levels tested.

Assessment descriptions. Every state-assessed content standard’s key elements are spelled out in brief, state-published assessment descriptions that most teachers rely on when devising their instructional plans.

Results reporting. Because the accountability tests provide annual reports regarding each student’s standard-by-standard achievement levels, the state’s teachers can readily determine which parts of their standards-focused instruction are working and which parts are not. Students and their parents also receive annual reports regarding each student’s mastery of every state-assessed content standard.

Test-focused professional development. The state’s teachers and administrators are routinely provided with professional development programs related directly to the statewide accountability program. One theme of these programs addresses the best ways for teachers to improve their instruction based on analyses of their stu- dents’ per-standard test performances. A second professional-development theme deals with how teachers can construct their own classroom assessments to measure those state-approved content standards that, although important for children to master, cannot be accurately measured in the time that is available for state-level testing.

Sample assessments. This year, for the first time, state authorities have distrib- uted an optional set of classroom assessments designed to measure students’ mas- tery of the state-assessed content standards for each grade level tested. Teachers are encouraged to use these optional assessments during the school year in a formative evaluative fashion, that is, to determine how students are progressing toward mas- tery of the state-assessed content standards.

Measures to assure curricular breadth. State officials have described several pro- cedures for monitoring a school’s curricular breadth. The state requires each school’s staff to use one of those monitoring mechanisms annually to supply evidence regarding the richness of the curricular targets addressed instructionally in that school’s classrooms. State officials have also made available to teachers a substantial number of optional classroom assessments that teachers can employ to measure their students’ mastery of “state-approved but state-unassessed ” content standards.

Summation. Although the state’s districts and schools are rigorously evaluated each year on the basis of the accountability tests’ results as well as other evidence (such as teachers’ pretest/post-test data and students’ attendance rates), the state’s teachers appear to be reasonably well satisfied with the accountability program because, given their understanding of what is to be promoted instructionally, most teachers have assembled evidence that they are doing a solid job of teaching. Based on students’ assessed progress in mastering the state-tested content stan- dards, the state’s teachers appear to be successful.

Popham-Teach Better ch10.qxd 8/6/2003 2:51 PM Page 147

C o p yr

ig h t ©

2 0 0 3 . A

ss o ci

a tio

n f o r

S u p e rv

is io

n &

C u rr

ic u lu

m D

e ve

lo p m

e n t. A

ll ri g h ts

r e se

rv e d .

Recommended Resources

Commission on Instructionally Supportive Assessment. Building tests to sup- port instruction and accountability: A guide for policymakers. Washington, DC: Author.

Kiernan, L. J. (Producer). (1996). Reporting student progress [Videotape]. Alexandria, VA: Association for Supervision and Curriculum Development.

Linn, R. L. (2000, March). Assessments and accountability. Educational Researcher, 29(2), 4–16.

Popham, W. J. (2001). The truth about testing: An educator’s call to action: Alexandria, VA: Association for Supervision and Curriculum Development.

T E S T B E T T E R , T E A C H B E T T E R1 4 8

INSTRUCTIONALLY FOCUSED TESTING TIPS

• Inform your colleagues about the attributes of standards-based

tests that make such tests instructionally supportive: standard-by-

standard reporting, accurate assessment of only the highest pri-

ority content standards, and clear assessment descriptions for

each measured content standard.

• Influence policymakers to replace, or at least supplement, tra-

ditionally constructed standardized achievement tests with in-

structionally supportive standards-based tests.

Popham-Teach Better ch10.qxd 8/6/2003 2:51 PM Page 148

C o p yr

ig h t ©

2 0 0 3 . A

ss o ci

a tio

n f o r

S u p e rv

is io

n &

C u rr

ic u lu

m D

e ve

lo p m

e n t. A

ll ri g h ts

r e se

rv e d .