Fordham Law Review

Volume 47 | Issue 5 Article 2

1979

Competency Testing Programs: Legal and

Educational Issues
Merle Steven McClung

Recommended Citation
Merle Steven McClung, Competency Testing Programs: Legal and Educational Issues , 47 Fordham L. Rev. 651 (1979).
Available at: http://ir.lawnet.fordham.edu/flr/vol47/iss5/2

This Article is brought to you for free and open access by FLASH: The Fordham Law Archive of Scholarship and History. It has been accepted for
inclusion in Fordham Law Review by an authorized editor of FLASH: The Fordham Law Archive of Scholarship and History. For more information,
please contact [email protected].
COMPETENCY TESTING PROGRAMS: LEGAL AND
EDUCATIONAL ISSUES*

MERLE STEVEN McCLUNG**

TABLE OF CONTENTS
INTRODUCTION ...................................... 652
I. LEGISLATIVE, LEGAL, AND EDUCATIONAL BACKGROUND ..................... 653
A. Legislative Background ............................................. 653
B. Legal Background ................................................. 657
C. Some Educational and Psychometric Considerations.......... ........ 665
IL INAPPROPRIATE TEST CONTENT ......................................... 672
A. Coerced Belief ..................................................... 674
B. Invasion of Privacy............................................... 677
C. Unteachable or Unmeasurable Content ................................. 678
II. ADEQUACY OF PHASE-IN PERIOD ........................................ 679
A. Issues .......................................................... 679
B. Analysis ......................................................... 681
IV. MATCHING TEST AND INSTRUCTION .................................... 682
A. Curricular and Instructional Validity ................................. 682
B. Basic Literary Skills Versus Adult Life Skills ............................ 684
V. RACIAL AND LINGUISTIC DISCRIMINATION ................................ 687
A. Racial Discrimination ................................................ 688
B. Linguistic Discrimination ..................... ......................... 692
C. Cultural Bias .......................................................... 694
VI. HANDICAPPED STUDENTS ................................................... 698
A. Exemption ............................................................ 698
B. Differential Diplomas and Standards .................................. 699
C. Differential Assessment Procedures ....................................... 699
D. Other Legal Implications ........................................... 700
E. Individualized Determinations.......................................... 701
VII. RECOMMENDED APPROACHES ................................................. 701
A. Community Participation ........................................... 702
B. Two-Phase Implementation Procedure................................. 704
C. Approaches to Instructional Validity and Assessment ..................... 705
D. Proposalsfor Competency Testing Programs .............................. 708
1. Basic Framework .............................................. .. 708
2. Model Program Provisions ........................................ 709
CONCLUSION .... .......................................................... 711

*© Merle Steven McClung 1979.

** Mr. McClung is currently working part-time as a staff attorney at the Center for Law and
Education in Cambridge, Massachusetts and part-time as an education law consultant. The
author acknowledges with appreciation the assistance of his colleagues at the Center, especially
Robert Pressman, in developing the legal analysis that forms the basis of this Article, although
the Article does not reflect their views in all parts. As of July 1979, Mr. McClung will be the
director of the Law and Education Center, Education Commission of the States, Denver,
Colorado.
 FORDHAM LAW REVIEW [Vol. 47

INTRODUCTION

C OMPETENCY testing' is one of the most significant and contro-

versial reform movements in public elementary and secondary edu-
cation today. The movement is primarily initiated by state legislators
who maintain that competency testing programs respond to public
dissatisfaction with the large number of functionally illiterate students
graduating from public high schools. By redirecting educational re-
sources to students with poorly developed literacy and numeracy skills,
some competency testing programs have constructive potential to
improve student performance in essential basic skill areas. Many
competency testing programs, however, have been designed and/or
implemented in an inequitable manner and are likely to have more
negative than positive effects. Programs that require a student to pass
a minimal competency test as a prerequisite to a high school diploma,
in particular, have potential for discrimination against students. Some
of these programs may not only be unfair to students, they may be
illegal as well.
Whether competency testing programs can and will have more
beneficial than harmful effects remains an open question. 2 Some would
agree with the National Academy of Education committee that compe-
tency testing as a prerequisite to a high school diploma "is basically
unworkable, exceeds the present measurement arts of the teaching
profession, and will create more social problems than it can conceivably
"3
solve.
1. This Article uses the term "competency testing" in a general sense to mean tests purporting
to measure basic skills and/or life-role activities because the legal analysis is essentially the same
for both. More exact terminology is offered by William Spady who defines "competencies as
indicators of successful performance in life-role activities (be they producer, consumer, political
citizen, driver, family member, intimate friend, recreational participant, or life-long learner) and
distinguishes them from the discrete cognitive, manual, and social capacities (such as reading and
computational skills, speaking ability, and motivation) that, when integrated and adopted to
particular social contexts, serve as the enablers or building blocks on which competencies
ultimately depend." Spady, Competency Based Education: A Bandwagon in Search of a
Definition, 6 Educ. Researcher 9, 10 (1977). The terms "proficiency" and "basic skills" are often
used in the same sense as "capacities." While the legal analysis is essentially the same, the
distinction between basic skills and life-role activities can be crucial. See notes 159-64 iinfra and
accompanying text.
2. For a general introduction to the issues involved in minimum competency testing, see:
American Friends Service Committee, A Citizens' Introduction to Minimum Competency Pro-
grams for Students (1978); National Institute of Education and Education Commission of the
States, Minimum Competency Testing: A Report of Four Regional Conferences (1978) [hereinaf-
ter cited as Minimum Competency Testing]; Brickell, Seven Key Notes on Minimum Competency
Testing, 59 Phi Delta Kappan 589 (1978).
3. National Academy of Education Committee on Testing and Basic Skills, Report to the
Assistant Secretary of Education: Improving Education Achievement 9 (1978) [hereinafter cited as
National Academy Report]. This quotation refers specifically to "statewide minimum competency
standards," but elsewhere in the report the same assertion is made with respect to competency
testing generally. See, e.g., id. at 16.
1979] COMPETENCY TESTING PROGRAMS

Others would agree with the Educational Testing Service (ETS) that
the problems are not insurmountable, and that a competency testing
program establishing "a baseline below which relatively few students
would fall, yet still allow students to acquire enough skills to function
as citizens, is preferable to the status quo-an unacceptable quality of
education." 4
Whatever one's personal views, the movement is so strong that few
elementary and secondary schools in this country are likely to be
untouched by its impact--or by the concomitant legal and educational
issues considered in this Article. 5

I. LEGISLATIVE, LEGAL, AND EDUCATIONAL BACKGROUND

A. Legislative Background
A number of studies indicate that, whatever definition of literacy is
used, substantial numbers of Americans are not literate. 6 One of these
studies, published by the Department of Health, Education and Wel-
fare (HEW), concluded that an estimated one million American
youths, twelve to seventeen years old, probably could not read as well7
as the average fourth grader, and thus they could be called illiterate.
The study showed that disproportionate numbers of black youths were
illiterate (15%), and that substantial numbers of white youths were
also illiterate (3.2%).8 Not surprisingly, the study also found that the
rate of illiteracy correlated with family income, declining "from 14

4. Educational Testing Service, Information Report: Basic Skills Assessment Around the
Nation 3 (1977) [hereinafter cited as Basic Skills Assessment].
5. This Article is based in part upon the following earlier publications by the author
McClung, Are Competency Testing Programs Fair? Legal?, 59 Phi Delta Kappan 397 (1978);
McClung, Competency Testing: Potentialfor Discrimination, 11 Clearinghouse Rev. 439 (1977)
[hereinafter cited as Potentialfor Discrimination];McClung, Developing Proficiency Programsin
CaliforniaPublic Schools: Some Legal Implications and a Suggested Implementation Procedure,
in Cal. State Dep't of Educ., Technical Assistance Guide for Proficiency Assessment app. K
(1978); McClung & Pullin, Competency Testing and HandicappedStudents, 11 Clearinghouse Rev.
922 (1978).
6. The exact degree of illiteracy, however, is subject to debate. Donald L. Fisher of the
University of Michigan reports: '[H]igh schools have been accused of graduating thousands, even
millions, of functional illiterates .... This accusation is, at first glance, supported by numerous
surveys and observations. For example, the surveys referred to in this paper report that anywhere
between 2 and 13 percent of the population with a twelfth grade education are functionally
illiterate. We will argue that the upper estimate is greatly inflated, and more often than not
misinterpreted." D. Fisher, Functional Literacy and the Schools 1 (1978). Of the studies that
report racial data, however, almost all show that disproportionate numbers of blacks and
Hispanics are illiterate. Another question that is subject to debate is whether schools are more or
less successful today than in the past in teaching basic literacy and numeracy skills.
7. U.S. Dep't of Health, Education, and Welfare, Literacy Among Youths 12-17 Years 3
(1973).
8. Id. at 4.
 FORDHAM LAW REVIEW [Vol. 47

percent in the lowest income group (less than $3,000) to 0.3 percent in
the highest ($15,000 or more)." 9
The widely publicized Adult Performance Level (APL) study con-
ducted by the University of Texas at Austin 10 found that on overall
competency performance in five knowledge and four skill areas, 19.7%
of the population could be classified as "functionally incompetent" or
"adults who function with difficulty," 33.9% could be classified as
"functional adults," and 46.3% could be classified as "proficient
adults."'" The functional incompetence rate was 21.7% in reading,
16.4% in writing, and 32.9% in computation. 12 The study concluded:
"Overall, approximately one-fifth of U.S. adults are functioning with
13
difficulty.'
Some studies exaggerate the extent of illiteracy and incompetency
among minority groups. An example is the Texas study which noted
great differences between whites and minority groups. "While 16% of
the Whites are estimated to be functionally incompetent, about 44% of
the Black and 56% of the Spanish-surnamed groups are estimated to
be so. Here, as with other variables that have been discussed, the
differences are probably due to the relatively lower levels of income,
education, job status, and job opportunity found among minority
groups in this country.' 4 This was only a partial explanation, how-
ever, because the extent of minority incompetence was exaggerated by
questionable test norms, as discussed in Part V(C).
The HEW and Texas studies nevertheless identify serious shortcom-
ings of many public schools in teaching and students in learning basic
skills. There is no consensus about whether competency testing is the
best means of correcting those shortcomings. Few would disagree,
however, that functional illiteracy in today's world for large numbers
of students after twelve years of public education is unacceptable.
While higher level skills such as analysis, synthesis, and evaluation
may be beyond the reach of many students, most educators would
agree that basic cognitive skills and processes fall within the domain of
teachable and learnable skills for virtually all students.' 5 Discussion
concerning the broader goals of public education is suspect if basic

9. Id. at 6.
10. University of Texas at Austin, Adult Functional Competency: A Summary (1975) (here-
inafter cited as Texas Study]. The Texas research was turned over to the American College
Testing Program (ACT) of Iowa City, Iowa, a test developer and publisher. ACT publishes an
Adult APL Survey and a Secondary APL Survey. ACT now markets tests based on the Texas
research and also tests tailored to specifications provided by school districts.
11. Id. at 5-6.
12. Id. at 6.
13. Id.
14. Id. at 8.
15. See, e.g., Madaus & Airasian, Issues in Evaluating Student Outcomes in Competency-
Based Graduation Programs, 10 J. Research & Dev. Educ. 79, 81 (Spring 1977).
1979] COMPETENCY TESTING PROGRAMS

literacy cannot be assured for virtually all students after twelve years
6
of schooling.'
The various reports of widespread illiteracy have produced a dra-
matic reaction in the public and in state legislatures. The legislative
rush to adopt competency testing programs has been startling. The
Education Commission of the States (ECS) has issued a series of
updates since April 1976 in an attempt to keep educators informed of
developments in minimal competency testing. In an update published
in November 1977, ECS reported:
As of April 15, 1976, four states had enacted legislation and four state boards of
education had adopted regulations mandating some form of minimal competency
testing. Bills were pending in eight state legislatures and five state boards had studies
under way. Today, 11 states have enacted legislation and 20 state boards have adopted
regulations, for a total of 31. Activity is occuring in the other 19 states either at the
7
local level, by state department study or legislative hearings.

There was also a proposed bill18 to amend the Elementary and

Secondary Education Act of 196519 to require all state agencies to
establish a program of basic educational proficiency standards before
funds could be received under the Act, but Congress limited the
federal role to special financial assistance for states and local districts
20
pursuing these goals.
The development of competency testing programs has occurred in
three stages. Statewide assessment programs were first designed simply
to measure performance in certain basic skills, for the purpose of
identifying the school districts or individual schools in need of help.
Extensive statewide programs of this kind were adopted in Colorado, 2 t
23 24
Michigan,2 2 New Jersey, and other states.
16. This statement is qualified to account for some mentally handicapped and other students
who do not have normal capabilities. For a challenge to traditional assumptions of student
capabilities, see B. Bloom, Human Characteristics and School Learning (1976). Bloom's research
in mastery learning leads him to conclude that it is possible for 95% of all students to learn, at
near the same mastery level, all that the school has to teach, with only 1 to 3% at the bottom level
who cannot master the curriculum.
17. Education Commission of the States, Update VII Minimal Competency Testing I (C.
Pipho ed. 1977) [hereinafter cited as Update VII].
18. H.R. 6088, 95th Cong., 1st Sess., 123 Cong. Rec. H3093 (daily ed. Apr. S, 1977)
(introduced by Rep. Mottl).
19. Pub. L. No. 89-10, 79 Stat. 27 (current version at 20 U.S.C. §§ 236-246 (1976)).
20. See Education Amendments of 1978, Pub. L. No. 95-561, §§ 921-922, 92 Stat. 2296 (to be
codified in 20 U.S.C. §§ 3331-3332). Section 921 authorizes federal assistance to the states in
setting education proficiency standards, and § 922 authorizes federal assistance to state and local
educational agencies to help develop their capacities to conduct programs for testing the
achievement in basic skills of children in elementary and secondary schools.
21. Colo. Rev. Stat. §§ 22-7-101 to -105 (1974), discussed in Education Commission of the
States, Cooperative Accountability Project, Educational Accountability in Four States 3-7. 96-104
(1975) [hereinafter cited as Educational Accountability].
22. Mich. Comp. Laws Ann. §§ 388.1081-.1086 (1976), discussed in Educational Accounta-
bility, supra note 21, at 10-11, 108-09.
 FORDHAM LAW REVIEW [Vol. 47

During the second stage in the development of competency testing

programs, the burden of poor schooling has been shifted onto the
student by testing each student as an individual, with the demonstra-
tion of minimal competency in basic skills being a prerequisite to a
high school diploma and/or grade-to-grade promotion. A survey made
in July 1978 reported thirteen states in this "second stage" of compe-
tency testing: Arizona (as of 1976), Oregon (1978), Florida (1979), North
Carolina (1980), Utah (1980), Alabama (1981), California (1981),
Delaware (1981), Vermont (1981), Virginia (1981), Maryland (1982),
Nevada (1982), and Tennessee (1982).2- And a February 1978 survey
found: "Overall, the majority of the states, nearly 80 percent, are, or
will be establishing in the near future, requirements for high school
graduation based on competency examinations." 2 6 Two leading states
27 28
in this movement are Florida and California.
The two states nevertheless represent dramatically different ap-
proaches. Florida has statewide standards and a statewide test. In
California the State Department of Education is precluded from devel-
oping a single statewide test-the responsibility for designing a test
falls to each local district. This difference in approach has profound
educational implications, as discussed in Part I(C). What is similar
about the Florida and California mandates (requiring students to pass
a competency test as a condition to receiving a regular high school
diploma) are timelines that make it very difficult for districts to
develop competency testing programs in a careful and sound way. The
senior class of 1978-1979 will be the first to face the new Florida
graduation requirements. In California the first class will be the one
that graduates after June 1980.

23. N.J. Stat. Ann. §§ 18A:7A-4 to -16 (West Supp. 1978-1979), discussed in Education
Commission of the States, Cooperative Accountability Project, Legislation by the States: Accoun-
tability and Assessment in Education 7, 13-22 (1975).
24. See, e.g., Education Accountability Act of 1971, ch. 71-197, §§ 1-5, 1971 Fla. Laws 1148
(current version at Fla. Stat. Ann. § 229.55 (West 1977)), discussed in Educational Accountabil-
ity, supra note 21, at 8-9.
25. See National Education Association, Impact of Minimum Competency Testing in Florida,
Today's Education, Sept.-Oct. 1978, at 30, 33 [hereinafter cited as Impact].
26. Chal, Minimum Competency Testing, 22 Harv. Graduate Sch. Educ. Bull. 9, 10 (1978).
27. See Fla. Stat. Ann. §§ 229.55-.58 (West 1977). The Florida program has been criticized
on psychometric, legal, and educational grounds. See Glass, Minimum Competence and Incompe-
tence in Florida, 59 Phi Delta Kappan 602 (1978) (psychometric critique); Potentialfor Discrini-
nation, supra note 5 (legal critique); Impact, supra note 25 (educational critique); notes 63, 83-85,
172, 177, 183-88 infra and accompanying text. On October 16, 1978 a lawsuit challenging the
Florida competency testing was filed. Debra P. v. Turlington, No. 78-892 Civ.-T-H (M.D. Fla.,
filed Oct. 16, 1978), discussed at note 63 infra and accompanying text. For a favorable view of
Florida's program, see Fisher, Florida'sApproach to Competency Testing, 59 Phi Delta Kappan
599, 599-602 (1978).
28. See Cal. Educ. Code §§ 51215, 51217 (,Vest 1978 & Supp. 1979); Hart, The California
Pupil Proficiency Law as Viewed by Its Author, 59 Phi Delta Kappan 592 (1978).
1979] COMPETENCY TESTING PROGRAMS

The third stage in the development appears to be a legislative

compromise, satisfying the political need to respond in some way
to the public demand for accountability, but avoiding the difficult
decisions involved in how to use the test results. Whereas the second
stage is characterized by the state mandating that passing a compe-
tency examination is a prerequisite to receiving a high school diploma,
the third stage is characterized by the state mandating competency
testing, but leaving the decision to tie the test to promotion or diploma
to local districts. Massachusetts, 29 Connecticut, 30 Rhode Island, 3 ' and
New Hampshire 32 are in the third stage of development. Regardless of
whether the third stage approach is motivated mainly by a greater
deference to local autonomy or by an increased sensitivity to the
potential legal and educational problems, the result is the same for
local districts. Local districts in third stage states will probably feel
some pressure to tie the test results to promotion and/or diplomas.
Therefore, the legal and educational issues considered in this Article
may be helpful to local districts as well as to uncommitted states in
deciding whether to tie prospective competency testing to promotion
and/or diplomas, and, if they decide affirmatively, which approaches
will minimize the concomitant problems.

B. Legal Background
In considering whether to develop some kind of competency testing
program, one of the most crucial questions faced by a state or a local
school district is how the test results will be used. The legality of a
testing program will usually depend more upon the use of the test
results than upon the test itself. 33 Using the test results as the primary
basis for any decision that will cause serious harm to a student raises
the initial legal questions. The trigger for legal analysis is this injury.
Assuming there is injury, the following questions arise: who is respon-
sible for that injury and does that person or agency have sufficient
justification for causing that injury?

29. The Massachusetts State Board and Department of Education specifically state that their
proposed regulations are not for the purpose of establishing a competency test as a new condition
for promotion or graduation, but the proposed regulations do not prohibit use for this purpose.
The purpose of the regulations is "to assist students in achieving mastery of basic skills prior to
high school graduation through the provision of appropriate curriculum, instruction, and evalua-
tion." Massachusetts Dep't of Education, Proposed Regulations: Basic Skills Improvement (1978)
[hereinafter cited as Proposed Regulations].
30. An Act concerning Education Evaluation and Remedial Assistance, Pub. L. No. 78-194,
1978 Conn. Pub. Acts 258.
31. Rhode Island has not formalized its policy as of this writing.
32. See New Hampshire Dep't of Education, Guidelines for the Implementation of New
Hampshire Accountability Plan (1978).
33. For a discussion of possible legal challenges to the test itself, see ps. 11, '(C) infra.
 FORDHAM LAW REVIEW [Vol. 47

If there is no injury, then there is no legal problem. Competency

tests can be used in many ways that cause no injury to a student. For
example, competency tests could be used simply to determine the
general level of student performance in basic skills on a statewide or
district level; to identify basic skill areas in an instructional program
that need more emphasis; or to diagnose areas in which an individual
student needs specific help.
34
In such cases, there is usually no injury
and no legal problem.
On the other hand, competency tests can be used to make de-
cisions about individual students that have potential for grave injury.
For example, competency tests can be used for tracking,"5 grade
promotion, or denial of a regular high school diploma. Diploma denial,
as mandated in Florida and California, probably causes the greatest
injury to an individual student, and therefore raises the most serious
legal questions.
Some people try to minimize the potential injury to students by
saying that sending students into the world without basic skills is a
greater disservice to them than diploma denial, and that a high school
diploma based upon social promotion rather than achievement does
not mean much anyway. This argument of course assumes that an
effective educational program and academic achievement is dependent
upon diploma denial,3 6 and ignores the reality that many prospective
employers and schools use the diploma as a minimum requirement
without regard to the actual skills required for the task or possessed by
the applicant.
Others will not try to minimize the injury of diploma denial, but will
argue that schools have a responsibility to certify competence for
employers, educational institutions and society at large, or that cer-
tification is a legally sufficient justification for a competency testing
program.3 7 Another justification likely to be advanced in support of

34. Tests with inappropriate content, however, may injure a student regardless of use. See
pt. II infra.
35. For an extensive judicial analysis of the harmful effects of tracking, see Hobson v.
Hansen, 269 F. Supp. 401 (D.D.C. 1967) (Wright, J.), appeal dismissed, 393 U.S. 801 (1968),
aff'd sub noma.Smuck v. Hobson, 408 F.2d 175 (D.C. Cir. 1969). For other cases on tracking, see
note 182 infra.
36. See note 246 infra.
37. Another common argument is that it is a misnomer to talk about "diploma denial" since a
student has no legal right to a diploma, and therefore no legal basis for challenging denial of a
diploma. A diploma, like education itself, however, is a benefit that must be made available to all
on equal terms when the state decideg to provide it. Consider the famous language of the
Supreme Court in Brown v. Board of Educ., 347 U.S. 483, 493 (1954): "[Ain opportunity of an
education where the state has undertaken to provide it, is a right which must be made available
to all on equal terms." Cf. Griffin v. County School Bd., 377 U.S. 218 (1964) (unconstitutional
closing of all public schools in county to avoid desegregation). Designating a diploma as a
"privilege" rather than a "right" would not affect the constitutional analysis. See Graham v.
Richardson, 403 U.S. 365, 375 (1971) (unconstitutional denial of welfare benefits to resident
19791 COMPETENCY TESTING PROGRAMS

competency testing is that it produces more beneficial than harmful

effects. The argument is that a serious penalty is necessary to motivate
students and teachers. 3 8 In this regard, Madaus and Airasian state: "If
the minimal competencies are to be perceived as an integral aspect of
the high school curriculum, skills to be consciously acquired and
valued, the evaluation of competencies must possess two characteris-
tics: it must be rigorous and it must involve penalties for non-
mastery. ' 39 The strength of this argument will depend in part upon
how successful the competency testing program is in helping students
40
achieve the requisite skills.
In support of its claim that minimum competency testing programs
can and do work, ETS points to two school systems-Westside High in
Omaha, Nebraska and the Denver, Colorado public schools-as ex-
amples of successful programs. 4 1 ETS states that Westside 4 2 has
graduated its first competency based class, with only 1% (eight seniors)
of 762 in the senior class not passing; and that Denver, a pioneer in
competency testing during the past eighteen years, has reduced its
failure rate from 14% to 1.5%. The competency testing programs in
Omaha (Westside) and Denver certainly merit careful study.
Whether other school systems will be able to similarly reduce their
failure rate remains an open question. The high initial failure rates for
students in Florida 4 3 and other geographic areas" where competency
testing programs have been implemented are not encouraging, and the
disproportionate failure rates for black and Hispanic students are
alarming.
The use of competency tests to determine which persons are mini-
mally competent and which persons are not, and to certify this

aliens). In sum, a state cannot withhold diplomas from some students without regard to federal
constitutional safeguards.
38. The penalty, of course, falls on the student regardless of the quality of the teaching or the
educational opportunities generally provided by the school. The disparities are most obvious
between property-rich and property-poor school districts. To date, teachers have been sufficiently
organized to prevent direct accountability for student performance on competency tests, but they
are also concerned about being held indirectly accountable.
39. Madaus & Airasian, supra note 15, at 83.
40. Because individual rights are involved, a school presumably could not justify individual
injury by demonstrating beneficial effects for students generally. The evaluation must ultimately
be fair with respect to the individual denied a diploma. See notes 263-64, 267 infra and
accompanying text
41. Basic Skills Assessment, supra note 4, at 3. These and other competency testing programs
are described in 59 Phi Delta Kappan 585-656 (1978).
42. Westside is located in a relatively affluent part of Omaha, and is not confronted with the
difficult educational problems of most urban schools with large numbers of disadvantaged
children.
43. See Florida Dep't of Education, Statewide Assessment Results by Demographic Catego-
ries, Fall, 1977 Tests, Grade Eleven Highlights (Functional Literacy) (1978).
44. Similar disproportionate failure rates on competency tests for minority students are
reported by school systems in Virginia, North Carolina, and Milwaukee, Wisconsin.
 FORDHAM LAW REVIEW [Vol. 47

competence or incompetence on a permanent record, could, in effect,

create a permanent subclass of American citizens-a group of "second
class" citizens comprised primarily of blacks, Hispanics, and other
persons who do not enjoy the benefits of a white middle-class back-
ground. Some states will create three classes of citizens by offering a
certificate of attendance was well: first class citizens who are deemed
competent (and awarded a regular high school diploma), second class
citizens who are deemed incompetent but dependable workers (certific-
ate of attendance), and third class citizens who are simply labeled
incompetent (no diploma or certificate). Some will argue that such
distinctions are not created but simply identified and certified by the
state. But this kind of competency testing program will, in effect,
create many of the distinctions, and the state's role in placing its 4-
official label of incompetence on a citizen cannot be trivialized,
especially because the state through its compulsory education laws,
will penalize any 4underage
6
student attempting to avoid the testing and
labeling process.
The potential injury to students is great enough to raise sobering
questions about whether to use competency testing for diploma denial
or differentiation, even if the measurement technology were sufficiently
developed to warrant confidence in determinations of competence and
incompetence. The beneficial effects of a competency testing program
should be considerable if it is to outweigh this probable injury. The
injury is certainly serious enough to warrant special care in making
determinations of incompetence. As underscored by a federal district
court judge in another context: "When a program talks about labeling
someone as a particular type and such a label could remain with him ' '4 7
for the remainder of his life, the margin of error must be almost nil.
For purposes of this Article, some form of serious injury to the
student from the use of test results will be assumed. Denial of a high
school diploma will be the primary example. Assuming that a plaintiff
will be able to show injury from the use of the test results, another
preliminary legal issue is whether the courts will refuse to review
competency testing programs because they involve evaluations of
academic performance.
Courts have traditionally been reluctant to review cases questioning
the adequacy of educational programs or the validity of professional

45. The extent of the perceived injury of a certificate rather than a diploma Is illustrated by
comments attributed to Morris Milton, an attorney for the National Association for the Ad-
vancement of Colored People: "Kids would be better off dropping out of school at the end of their
senior year than being stigmatized by a Certificate of Attendance as certified dummies."
Tallahassee Democrat, Mar. 24, 1978, § B at 3, col. 2 (quoting Mr. Milton).
46. The compulsory education laws in most states extend to age 16. See K. Alexander & K.
Jordan, Legal Aspects of Educational Choice: Compulsory Attendance and Student Assignment
11-12 (1973).
47. Merriken v. Cressman, 364 F. Supp. 913, 920 (E.D. Pa. 1973).
1979] COMPETENCY TESTING PROGRAMS

judgments about academic performance of students. Such judicial

restraint is rooted in an appreciation of the subjective nature of most
determinations about educational quality and academic performance.
The subjective nature of academic evaluations was a major factor in
the failure of students to prevail in two recent suits against public
school systems.
In Peter W. v. San Francisco Unified School District,48 a function-
ally illiterate high school graduate failed in his $1 million educational
malpractice claim against the public high school that had presumably
provided him with an inadequate education. Although the school had
awarded him a high school diploma, he could not read and write well
enough to fill out a standard job application. In affirming the dismissal
of this case, the California Court of Appeals stressed the subjective
nature of the relevant educational standards:

[T]he wrongs and injuries involved [are not] both comprehensible and assessable within
the existing judicial framework. . . Unlike the activity of the highway or the
marketplace, classroom methodology affords no readily acceptable standards of care,
or cause, or injury. The science of pedagogy itself is fraught with different and
conflicting theories of how or what a child should be taught, and any layman
might-and commonly does--have his own emphatic views on the subject.4 9

48. 60 Cal. App.3d 814, 131 Cal. Rptr. 854 (1976).

49. Id. at 824-25, 131 Cal. Rptr. at 860-61 (emphasis added) (citations omitted). This language
has important implications for Peter IV.-type cases that may be brought against school systems
that are implementing competency testing programs. The lack of objective standards in education
is one reason why no educational malpractice case of this kind has ever been successful. See
Donohue v. Copiague Union Free School Dist., 64 A.D.2d 29, 33, 407 N.Y.S.2d 874, 877 (1978)
(citing Peter W. in affirming dismissal of case).
Competency testing programs significantly change the nature of educational standards by
setting explicit educational goals, clearly specified performance objectives related to those goals,
objective test specifications to measure those objectives, and test items keyed to those spec-
ifications. The Florida statute arguably creates an explicit right to a minimally adequate
education: "The first priority of the public schools of Florida shall be to assure that all Floridians,
to the extent their individual physical, mental, and emotional capacities permit, shall achieve
mastery of the basic skills. The term 'basic skills,' for purposes of this section, means reading,
writing, and arithmetic." Fla. Stat. Ann. § 230.2311(1) (Vest 1977). Even in the absence of such
statutory language, the officially stated purpose of most competency testing programs is to
provide a minimum level of education for all students, and objective standards are adopted to
serve that purpose.
Educators often express the concern that they may be sued whether or not they give a
functionally illiterate student a high school diploma. Competency testing programs, with their
explicit goals and objective standards, will provide a stronger basis for this concern. In special
education, where professional standards are often more established, a student recently won an
education malpractice case. A New York City jury awarded $750,000 in damages to a former
student who claimed that he had been negligently misclassified and miseducated as retarded. The
New York Appellate Division reduced the award to $500,000 but distinguished the Donohue and
Peter W. cases and let the verdict stand. Hoffman v. Board of Educ., - A.D.2d -, -, 410
N.Y.S.2d 99, 110-11 (1978). See generally Abel, Can a Student Sue the Schoolsfor Educational
Malpractice?, 44 Harv. Educ. Rev. 416 (1974); Ratner, Remedying Failure To Teach Basic
Skills, 17 Inequality Educ. 15 (June 1974); Sugarman, Accountability Through the Courts, 82
Sch. Rev. 233 (1974); Comment, Educational Malpractice, 124 U. Pa. L. Rev. 755 (1976).
 FORDHAM LAW REVIEW [Vol. 47

A similar rationale underlies the Supreme Court's refusal in Board of

Curators of the University of Missouri v. Horowitz"0 to provide
substantial procedural safeguards to medical students dismissed -for
academic reasons. Writing for a divided Court,Sl Justice Rehnquist
emphasized "the significant difference between the failure of a student
to meet academic standards and the violation by a student of valid
rules of conduct."'5 2 Academic evaluations of a student, in contrast to
disciplinary determinations, he noted, do not lend themselves to judi-
cial factfinding procedures:
The decision to dismiss [the plantiff] ... rested on the academic judgment of school
officials that she did not have the necessary clinical ability to perform adequately as a
medical doctor and was making insufficient progress toward that goal. Such a
judgment is by its nature more subjective and evaluative than the typical factual
questions presented in the average disciplinary decision. Like the decision of an
individual professor as to the proper grade for a student in his course, the determina-
tion whether to dismiss a student for academic reasons requires an expert evaluation of
cumulative information and is not3 readily adapted to the procedural tools of judicial or
administrative decisionmaking.5

Relying on cases like Peter W. and Horowitz, some educators

question whether competency testing is different from any other testing
program traditionally implemented by public schools free of judicial
intervention. In addition to claims of racial discrimination, 54 there are
at least three important distinctions: the extent of the injury, the
ultimate reliance on a single assessment, and the purportedly objective
nature of that assessment. What makes competency testing different
from most other high school testing programs is the severe nature of
the penalty for failing a single test-denial of a high school diploma or
its equivalent.5 5 Until recently, very few public high schools in this
country have made a student's diploma depend upon the outcome of
one test,5 6 and the severe nature of the penalty almost guarantees that
the testing program will be subject to intense scrutiny by students,
parents, and their lawyers.

50. 435 U.S. 78 (1978).

51. Justices Powell, White, Marshall, and Blackmun filed separate opinions. Id. at 79.
52. Id. at 86.
53. Id. at 89-90 (emphasis added).
54. See pt. V infra.
55. See notes 33-47 supra and accompanying text.
56. Some people point to the traditional Regents Examination in New York State, see N.Y.
Educ. Law § 209 (McKinney 1969), as being similar to contemporary competency testing
programs. The Regents Examination, however, is not used as a basis for denying high school
diplomas or for determining whether a student is competent or incompetent. A Regents diploma
is generally regarded as a superior diploma and has special value for college admissions. In May
1976, the New York Board of Regents approved a resolution making passage of a competency
test a prerequisite to a high school diploma beginning with the graduating class of June 1979. See
Update VII, supra note 17, at 39.
1979] COMPETENCY TESTING PROGRAMS

The most important distinction between contemporary competency

testing and traditional testing programs, however, is the purportedly
objective nature of the former. Most competency testing programs are
specifically designed to provide relatively objective criteria for making
the crucial evaluation of academic performance. Unlike the subjective
evaluations involved in the Peter W. and Horowitz cases, competency
testing programs purport to provide relatively objective measurements
of competence. They are usually characterized by clearly specified
standards: explicit educational goals, performance objectives related to
those goals, test specifications designed to measure those objectives,
test items keyed to those specifications, and finally the crucial determi-
nation of a cut-off score separating those deemed to be competent,
from those deemed to be incompetent. These efforts are supported by
psychometric specialists who perform the various psychological, statis-
tical, and validation studies necessary to construct a test instrument
that meets accepted professional standards. The result is a test instru-
ment that appears as impressive and scientific as the best of the
traditional standardized tests5 7 but is also subject to many of the same
as well as some additional shortcomings. Reliable, valid, and timely
measurement of functional competency is especially problematic, as
noted by William Spady of the National Institute of Education: "The
technology surrounding the assessment and measurement of success in
life role activities is only in its infancy, even though the rush toward
adopting [competency-based] programs is upon us." 8
Although the technology of constructing competency tests is not
sufficiently developed to warrant great confidence in its scientific
objectivity, educators implementing many competency testing pro-
grams will be making assumptions of objectivity, and in turn making
crucial decisions based on those assumptions. Those assumptions are
likely to be tested in court. As a result, courts will no longer be faced
with a challenge to a graduation decision based upon "an expert
evaluation of cumulative information"-that is, hundreds of tests in
scores of classes graded by numerous teachers using complex personal
mixes of objective and subjective criteria. Instead, courts will be asked
to review an obviously crucial decision based primarily upon the
results of one test instrument purported to incorporate test items that
are objectively related to clearly specified performance objectives and
explicit educational goals. The academic evaluation involved in a
competency test, therefore, seems, to use Justice Rehnquist's language,
readily adaptable "to the tools of judicial decisionmaking." 5 9 To so

57. "The average American parent has a great and naive faith in 'scientifically' constructed
tests." Sherrer & Roston, Some Legal and Psychological Concerns About Personality Testing in
the Public Schools, 30 Fed. B.J. 111, 114 (1971).
58. Spady, supra note 1, at 11.
59. 435 U.S. at 90.
 FORDHAM LAW REVIEW [Vol. 47

characterize competency testing is not to predict that the courts will in

fact treat academic decisions based upon competency test results
differently from traditional evaluations of academic performance; it
does suggest, however, that the courts' primary rationale for past
restraint is far less persuasive in the present situation, and that judicial
review is warranted.
Apart from these new characteristics of competency testing pro-
grams, they will be subject to review by the legal standards applied to
traditional testing programs. The most obvious of these standards are
those prohibiting racial and linguistic discrimination in testing
programs-in particular the constitutional prohibition against certain
practices that carry forward the effects of proven, past discrimina-
tion. 60 In addition, procedural and substantive due process standards 61
may provide a basis for scrutinizing competency testing programs.
The minimal competency requirement as a prerequisite to a high
school diploma is a new phenomenon in most states. Therefore, with
the exception of racial and linguistic discrimination, it is difficult to
identify the strongest legal arguments for or against it, and even more
difficult to predict the judicial response. The first direct legal challenge
to a state competency testing program, 6 2 Debra P. v. Turlington,63 was

60. See notes 179-88 infra and accompanying text.

61. See pts. Ill, IV infra.
62. Two limited procedural challenges to the Florida competency testing program preceded
Debra P. v. Turlington, No. 78-892 Civ.-T-H (M.D. Fla., filed Oct. 16, 1978): Robinson v.
School Bd., No. 78-2137-CA (L) 01 H (Fla. Cir. Ct., filed June 8, 1978); Brady v. Florida State
Bd. of Educ., No. 78-653R (Fla. Div. Adm. Hearings June 15, 1978).
In Robinson, plaintiff alleges that the Palm Beach County School Board did not provide
adequate notice of the rule-making involved in requiring students to pass a competency test by
June 1978 as a prerequisite to a high school diploma, and that the Board did not prepare an
economic impact statement for the proposed rule as required by the Florida Administrative
Procedure Act, Fla. Stat. Ann. § 120.54 (West Supp. 1974-1978).
In Brady, the court ruled that the cut-off score required by the board of education for
evaluating mastery of the competency testing program is a "rule" within the meaning of Fla. Stat.
Ann. § 120-52(14) (West Supp. 1974-1978). Because the scoring criterion was not promulgated or
adopted in accordance with the Administrative Procedure Act, the court declared it an invalid
rule, thus forcing the state to hold hearings on the rule pursuant to" the Act. The state held
hearings on the proposed rule in September 1978.
63. No. 78-892 Civ.-T-H (M.D. Fla., filed Oct. 16, 1978). The first sentence of the
complaint's preliminary statement reads: "This is a class action brought by certain black school
children in Hillsborough County challenging the defendants' administration of the Florida
Functional Literacy Examination (now denominated as Florida Student Assessment Test, Part
II), and the statutory scheme under which the test is required, as racially discriminatory and a
deprivation of due process." Complaint at 1-2, Debra P. v. Turlington, No. 78-892 CIv.-T-H
(M.D. Fla., filed Oct. 16, 1978) [hereinafter cited as Complaint]. The ten named plaintiffs seek to
represent "a class consisting of all present and future twelfth grade black public school students In
the State of Florida who have failed or who hereafter fail the Florida Functional Literacy
Examination." Id. at 3. The named defendants include Commissioner of Education Ralph
Turlington, the Florida State Board of Education and Department of Education, and the School
Board of the Hillsborough County Schools. Plaintiffs ask the court, inter alia, to enter a
1979] COMPETENCY TESTING PROGRAMS

recently filed in federal district court in Florida. The plaintiffs allege

that the Florida competency testing program as applied to black
students violates rights guaranteed them by federal statutory and
constitutional law. Debra P. is only the first in what is likely to be a
long line of legal challenges as various states begin to withhold high
school diplomas on the basis of a competency test, and a judicial
pattern and predictable case law on competency testing will eventually
emerge as a result of these challenges.

C. Some Educationaland Psychometric Considerations

As of this writing, there are no court decisions or HEW guidelines
setting minimum standards for competency testing of high school
students, but both are likely in the near future. 64 Until then, decisions
and standards in related areas offer guidance. Judicial involvement in
determinations of test validity has focused primarily on situations
involving employee selection procedures. The Supreme Court 6s has
relied upon the Guidelines on Employment Selection Procedures (Fed-
eral Guidelines) 66 developed by the Equal Employment Opportunity
Commission, and upon the Standards for Educational and Psychologi-
cal Tests (APA Standards), 67 developed by a joint committee of the
American Psychological Association (APA), the American Educational
Research Association (AERA), and the National Council for Measure-
ment in Education (NCME). In fact, the Federal Guidelines defer to
the APA Standards on technical matters, as noted with some concern
by Novick:
[T]he Federal Executive Agency Guidelines on employment selection issued in 1976
refer to the joint test standards and, in fact, defer to these standards for technical
detail. Thus, executive agency guidelines effectively having the force of law are
adopting
68
professional standards that were not specifically written with this purpose in
mind.

permanent injunction restraining defendants from requiring that members of the plaintiff class
pass the examination as a prerequisite to receiving a high school diploma. Id. at 18. The
complaint's specific claims for relief are set forth in text accompanying notes 145, 172, 183-88, 251
infra.
64. HEW has circulated a draft policy on competency testing to state departments of
education for comment, but no policy has been approved as of this writing.
65. See, e.g., Washington v. Davis, 426 U.S. 229 (1976) (American Psychological Associa-
tion's (APA) Standards for Educational and Psychological Tests applied to testing practices of the
Washington, D.C. Police Department in selecting individuals for its training academy).
66. 29 C.F.R. §§ 1607.1-14(1978).
67. American Psychological Association, Standards for Educational and Psychological Tests
(1974) [hereinafter cited as APA Standards].
68. Novick, The Influence of the Law on Professional MIeasurement Standards, in Educa-
tional Measurement & the Law 41, 47 (1977).
 FORDHAM LAW REVIEW [Vol. 47

Recent decisions by the Supreme Court indicate an increasing aware-

ness of important differences between the Federal Guidelines and APA
Standards, and also a split within the Court as to which set of test
standards should be accorded greater deference. 69 This difference
concerning the respective weight that should be given to federal agency
guidelines versus professional standards is unlikely to be resolved soon,
especially because the Federal Guidelines have been replaced by the
new Uniform Guidelines on Employee Selection Procedure70 and be-
cause the APA Standards are being considered for revision in light of
their increased use in resolving legal disputes. 7 1 Whatever the ultimate
resolution, it is likely to incorporate considerable deference to profes-
sional standards on technical, psychometric questions. Therefore, some
understanding of the basic psychometric concepts and terminology that
courts have and will probably continue to rely upon in deciding
questions of fair testing is essential.
The APA Standards define the requirements for reliable and valid
tests. Simply put, reliability refers to whether the instrument measures
accurately what it measures; for example, the instrument should yield
comparable results when used at different times. 7 2 Validity refers to
whether the instrument measures what it purports to measure. There
are many different kinds of test validity, as indicated below, and each
should be considered in relation to the exact test in question. Accord-
ing to the APA Standards:
Questions of validity are questions of what may properly be inferred from a test score
... . The many types of validity questions can, for convenience, be reduced to two:
[wlhat can be inferred about what is being measured by the test [and] [w]hat can be
inferred about other behavior?

The kinds of validity depend upon the kinds of inferences one might wish to draw
from test scores. Four interdependent kinds of inferential interpretation are tradi-
tionally described to summarize most test use: the criterion-relatedvalidities (predic-
73
tive and concurrent); content validity; and construct validity.

Predictive validity is a measure of how well test items predict the

future performance of test takers. This type of assessment requires an
analysis comparing the predictions about each test taker based on the
test results with the actual functioning of the test taker at a later point
in time. For example, if a high number of students which an Adult
69. See Lerner, The Supreme Court and the APA, AERA, NCMA Test Standards: Past
References and Future Possibilities, 33 American Psychologist 915 (1978).
70. 43 Fed. Reg. 38290-309 (1978) (to be codified in 5 C.F.R. § 300.103(c); 28 C.F.R. §
50.14; 29 C.F.R. § 1607; 41 C.F.R. § 60-3).
71. An AERA, APA, NCME Review Committee has been established to make recommenda-
tions about the proposed revision of the APA Standards. See Letter from Willo P. White,
Administrative Associate, Office of Scientific Affairs, APA to Merle McClung (Feb. 23, 1978).
72. See APA Standards, supra note 67, at 48.
73. Id. at 25-26 (footnote omitted).
1979] COMPETENCY TESTING PROGRAMS

Performance Level (APL) test had predicted would be functionally

incompetent in fact turned out to be such when studied74years later,
then the test could be said to have predictive validity.
Concurrent validity is a measure of how well test results correlate
with other criteria which might provide the same type of information
about test takers. This type of assessment provides a measure of a
test's immediate predictive validity, or how well determinations based
upon test results correlate
75
with other currently available information
about test takers.
Construct validity is a measure of how well test items correlate to
the theory or constructs behind the test. This assessment indicates the
relationship between the test and actual test performance. This as-
sessment is probably the most difficult to conduct since it may be
difficult to identify the constructs upon which a test is built and
because a statistical
76
analysis of the interrelationship of test items may
be required.
Content validity is a measure of how well test items represent the
knowledge that the test purports to measure. A test with a high degree
of content validity is a test for which high test performance serves as
an index of a high degree of 77
skills or knowledge in the area which the
test purports to measure.
Two kinds of content validity not specified in the APA Standards-
curricular and instructional validity-are of special relevance to com-
petency testing. As argued below in Part IV, any competency test that
lacks instructional validity in particular should not be used as the basis
for a decision as important as diploma denial.
Two other issues with far-reaching educational and psychometric
implications are whether a state committed to competency testing
chooses (1) statewide or locally developed tests, and (2) criterion-
referenced or norm-referenced tests. The issue of statewide versus local
tests, with corresponding implications for statewide versus local cur-
riculum, raises, in a new context, the classic controversy between state
control and local autonomy, with local districts ever ready to ward 78
off
any attempted encroachment upon their traditional authority.
One advantage of a statewide test as in Florida is uniformity.
Students moving from one school district to another within the state
74. See generally id. at 26. For comment on predictive validity and competency tests, see P.
Airasian, G. Madaus, & J. Pedulla, Policy Issues in Minimal Competency Testing and a
Comparison of Implementation Models 73-74 (1978) [hereinafter cited as Policy Issues] (prepared
for and published by the Massachusetts Department of Education).
75. See generaUy APA Standards, supra note 67, at 26.
76. Id. at 29-30.
77. Id. at 28-29.
78. On this issue, however, some local districts would prefer a statewide test because of the
difficulty and expense of constructing a valid tesL In fact, a number of local school districts in
Virginia reportedly expressed this view to the state legislature and Department of Education.
 FORDHAM LAW REVIEW [Vol. 47

will not be exposed to easier, 79 more difficult, or simply different

standards, and thus employers and others throughout the state will
know what minimum standards are certified by the diploma. One
disadvantage is that a statewide test cannot be tailored to the diverse
instruction and curricula that has been offered in local districts, and it
will take a number of years for curriculum and instruction to adapt
themselves to the test.
This in turn raises another classic controversy about whether tests
should lead or follow curriculum, and whether teachers should "teach
to the test." 80 Some commentators like Herbert Rudman would not see
these as important issues with minimal competency testing because
they do not see significant differences among American schools in basic
skills curriculum and instruction. Others like Professor W. James
Popham of the University of California at Los Angeles challenge this
view:
At a rather general level of abstraction, Rudman is accurate in his claim that "there is
a relatively uniform program among U.S. school districts." But day-to-day classroom
activities are not fashioned at such a general level, nor are standardized test items
written at this level of generality. It is only when we take such general concepts as "the
ability to perform fundamental mathematical operations" and write test items or
design instructional materials that differences in educators' preferences begin to show
up. And there are differences aplenty. 8 '

Differences between local districts in basic skill instruction, and be-

tween the test and instruction, will be especially acute where the
competency test is designed to measure adult-life skills, as discussed in
more detail in Part IV(B).
Thus statewide tests, at least in the short run, are inherently
vulnerable to a charge of insufficient match between what the test
measures and what the students have been taught. Locally developed
tests, on the other hand, can be tailored to local curriculum and
instruction, but designing a reliable and valid test that meets profes-
sional psychometric standards is beyond the expertise and perhaps the
budget of most local districts. Test publishers are attempting8 2to solve
this problem by creating "item pools," as discussed herein.
Many educators, however, question whether current educational
technology is sophisticated enough to produce a sound competency
testing program, even if statewide resources are available. Others are
engaged in an intense debate over the respective advantages and dis-
advantages of criterion-referenced versus norm-referenced tests-the
former are usually preferred for competency testing programs.
79. Some reports indicate student flight to schools with lower standards. See Time, Dec. 12,
1977, at 22 (Florida juniors are fleeing to high schools in Georgia where a competency test is not
required for graduation).
80. See generally Madaus & Airasian, supra note 15, at 82-90.
81. Popham, The Standardized Test Flap Flop, 59 Phi Delta Kappan 470, 471 (1978).
82. See notes 261-62 infra and accompanying text.
1979] COMPETENCY TESTING PROGRAMS

The intensity of the debate is evident in a hard-hitting article by

Professor Gene Glass of the University of Colorado. 3 Glass criticizes
the ETS item writers for arbitrarily writing relatively difficult mathe-
matical items for the Florida test. Glass's main objection is the
arbitrariness inherent in setting standards for criterion-referenced tests.
"Teachers and their consultants attempting to define 'competencies'
and writing test items intended to reflect minimal levels of acquisition
are engaged in a bootless and potentially embarrassing endeavor." 84
Referring generally to Florida's competency testing program, Glass
concludes that Floridians should call for "an immediate suspension of
the minimal competence graduation requirements program, because it
is based on indefensible technology. The items of the test have never
been validated as measures of 'survival skills,''8 S and the pass-fail
standards were set mindlessly and capriciously.
Glass's claim of the indefensible arbitrariness of criterion-referenced
tests is challenged by many,8 6 including Professor Popham:
The cornerstone of Glass's attack on the setting of performance standards is his
assertion that such standards are set arbitrarily. He uses that term in its most
pejorative sense, that is, equating it with mindless and capricious action. But while it
can be conceded that performance standards must be set judgmental y, it is patently
87
incorrect to equate human judgment with arbitrariness in this negative sense.
Reviewing this debate, Professor Robert Linn of the University of
Illinois concludes: "The process of setting standards and of making
inferences based on those standards lacks a firm theoretical founda-
tion, and this in turn creates problems that are apt to be of more than
academic interest .... Typical current practice is not likely to provide
a basis that most standard setters would be comfortable defending in
court." 8 8
While Glass contends that criterion-referenced tests like the Florida
test are "capricious and authoritarian, "89 value judgments based on

83. Glass, supra note 27, at 602-05.

84. Id. at 603. For more detailed discussion of the state of the "science" of setting cut-off
scores, see: Burton, Societal Standards, 15 J. Educ. Measurement 263 (1978); Glass, supra note
27. But see Block, Standards and Criteria:A Response, 15 J. Educ. Measurement 291 (1978);
Hambleton, On the Use of Cut-Off Scores with Criterion-Referenced Tests in Instructional
Settings, 15 J. Educ. Measurement 277 (1978); Popham, As Always, Provocative, 15 J. Educ.
Measurement (1978) [hereinafter cited as As Always]; Scriven, How to Anchor Standards, 15 J.
Educ. Measurement 273 (1978). For an article that raises some broader questions about standard
setting, see Levin, Educational Performance Standards: Image or Substance?, 15 J. Educ.
Measurement 309 (1978).
85. Glass, supra note 27, at 605.
86. See Block, supra note 84; Hambleton, supra note 84; As Always, supra note 84; Scriven,
supra note 84.
87. As Always, supra note 84, at 298.
88. Linn, Demands, Cautions, and Suggestionsfor Setting Standards, 15 J. Educ. Measure-
ment 306 (1978).
89. Glass, Standards and Criteria, 15 J. Educ. Measurement 237, 259 (1978).
 FORDHAM LAW REVIEW [Vol. 47

comparative evidence as exemplified by norm-referenced tests impress

him as "cogent and fair." 9° This assertion is challenged by critics of
norm-referenced measures who point out, among other things, that
norm-referenced tests raise similar difficulties and are less useful than
criterion-referenced measures to diagnose learning problems and guide
instruction.9 1 Critics also note that in order to produce a bell-shaped
curve, norm-referenced tests are usually constructed to eliminate items
that most students can answer and, therefore, are designed to exag-
gerate differences in achievement among students. 92 Criterion-
referenced tests, they argue, overcome such shortcomings by measur-
ing students against objective standards rather than against themselves
93
(a driver's license test being the classic example).
The relative merits and shortcomings of criterion-referenced versus
norm-referenced measures was the subject of the 1978 AERA Presi-
dential Debate. Professor Robert Ebel of Michigan State University
argued that norm-referenced tests enjoy an established and superior
technology. 94 Professor Popham, on the other hand, argued that
criterion-referenced tests are superior for purposes of individual diag-
nosis, classroom instruction, and program evaluation. 95
Whether criterion-referenced or norm-referenced, most authorities
recognize the subjective nature of setting cut-off scores for competency
tests:
[A]lthough at first glance it appears as if the statistical technology is available for
determining cut-scores, there is no consensus on any one approach and all approaches
rely on subjective judgments of some sort.... In reality, part of what enters into the
determination of a cut-score is the 96
answer to the question "What percentage of
students can we reasonably fail?"

The largely subjective and political questions of what competencies

to test for and what is an appropriate cut-off score are two of the most
important decisions in designing any competency testing program.

90. Id.
91. See Block, supra note 84; Hambleton, supra note 84; As Always, supra note 84.
92. See generally Buros, Fifty Years in Testing: Some Reminiscences, Criticisms, and
Suggestions, 6 Educ. Researcher 9, 12 (1977) ("These statistical methods of item validation
confuse differentiation with measurement and exaggerate differences among individuals and
between grades. . . . In a 1948 paper. . . I urged the abandonment of these statistical methods
of validating test items as well as the practice of discarding items simply because they were either
passed by all or failed by all.").
93. See, e.g., W. Popham, Criterion-Referenced Measurement (1978).
94. Professor Ebel reiterates his belief that "a test of minimum competency should be a
domain-sampling test, not an objective-referenced test. It should be designed to indicate a pupil's
general level of achievement, not the particular competencies the pupil may or may not have
learned." Ebel, The Case for Minimum Competency Testing, 59 Phi Delta Kappan 546, 548
(1978).
95. See generally W. Popham, supra note 93.
96. Policy Issues, supra note 74, at 60-61; Haney & Madaus, Making Sense of the Competency
Testing Movement, 48 Harv. Educ. Rev. 462, 466-69 (1978).
1979] COMPETENCY TESTING PROGRAMS

Various individuals and groups within a given community (e.g.,

minority groups, employers) may have different views on these mat-
ters, and their support may be essential for an effective program. The
inherently subjective and political nature of setting standards for
competency tests provides a compelling argument 97
for representative
community-based participation in the process.
Some educators would not see much relevance in the debate over
criterion-referenced versus norm-referenced tests because they do not
believe that the measurement technology is sufficiently sophisticated to
design any competency testing program sound enough to make crucial
decisions such as which students receive and which are denied a high
school diploma. 98 This view is articulated in a report by the National
Academy of Education Committee on Testing and Basic Skills that
was prepared at the request of HEW Secretary Joseph Califano. 99 The
committee, 10 0 chaired by Stephen K. Bailey of the Harvard Graduate
School of Education, concluded in part:
[A]ny setting of state-wide minimum competency standards for awarding the high
school diploma-however understandable the public clamor which has produced the
current movement and expectation-is basically unworkable, exceeds the present
measurement arts of the teaching profession, and will create more social problems than
it can conceivably solve.

. . . The effort to determine and assure minimum competency standards for high
school graduation will . . . fall of its own weight, for the scaffolding of existing test
designs is too weak to carry such an emotionally-laden and ambiguous burden.
Continuing extensive efforts and funds in this direction is wasteful and takes attention
away from the major tasks of improving our schools.' 0 '

Reports such as these have dissuaded some states and school dis-
tricts from following the Florida and California examples of using
competency tests as a prerequisite to granting a high school diploma.
But others continue to express an interest in the possible merits of such
competency testing programs. Their interest is encouraged by organi-
zations such as ETS, which helped to design the Florida test.
ETS takes the position that the difficulties raised by competency
97. See pt. VII(A) infra.
98. Consider, for example, the following statement by McKenna: "As promising as the
concept of criterion-referencing seems for correcting some of the problems of standardized
measurement, criterion-referenced tests are not only complex and time consuming to develop, but
they often retain many of the deficiencies of standardized tests: poor attention to instructional
objectives; irrelevant and misleading items; and continued use for sorting, categorizing, and
predicting success rather than for diagnosing learning difficulties and improving instruction."
McKenna, A Tale of Testing in Two Cities, in The Myth of Measurability 229 (P. Houts ed.
1977).
99. National Academy Report, supra note 3.
100. In addition to Chairperson Bailey, the committee members were John B. Carroll,
Jeanne Chall, Robert Glaser, John I. Goodlad, Diane Ravitch, Lauren Resnick, Ralph W. Tyler,
and Robert L. Thomdike.
101. National Academy Report, supra note 3, at 9 (footnote omitted).
 FORDHAM LAW REVIEW [Vol. 47

testing programs are not insurmountable, and that there are strong
indications that such programs can and do work. 10 2 Another advocate
of competency testing, Professor Ebel, contends that competency test-
ing can help to reverse "two generations of warm-hearted but soft-
headed pedagogy" that relegated "the cultivation of cognitive compe-
tence" to a secondary role. "Such tests can motivate efforts to 1learn
03
and efforts to teach by providing specilic goals and rewards."
Although many states and school districts are starting to adopt a
more cautious stance, ETS and Professor Ebel appear to reflect the
majority view as of this writing. Professor Decker Walker believes that
support for competency testing among the official lay bodies control-
ling most state and local educational agencies is so strong that unani-
mous professional opposition, even if it could be mobilized (and it
obviously cannot), would make little difference. Therefore, Professor
Walker recommends that educators who see more dangers than ben-
efits in competency testing programs should devote their efforts
"to avoid the greatest dangers and maximize the positive conse-
quences. ' 10 4 In sum, competency testing programs and their concom-
mitant legal issues are likely to be with us for some time.

II. INAPPROPRIATE TEST CONTENT

Because competency testing is a relatively new phenomenon and
most tests are not available for public inspection, 10 5 detailed consider-
ation of test content is not possible at this time. There are indications,
however, that competency tests in some states may incorporate inap-
propriate content.
Consider, for example, the following tasks set forth in the compe-
tency test guide developed for Florida's Palm Beach County
schools. 10 6 The "Consumer Economics" part of the test guide includes
the following comprehension task: "discuss the idea that just be-
cause a rich family can afford to feed, clothe and educate a large
number of children, this does not mean that the world will be able to

102. Basic Skills Assessment, supra note 4, at 3.

103. Ebel, supra note 94, at 547.
104. Walker, The HardLot of the Professional in a Reform Movement, 35 Educ. Leadership
83, 84 (1977).
105. If the tests are not generally available as a policy matter under various freedom of
information laws, attorneys for students who are denied diplomas on the basis of a test would
probably seek a judicial order for disclosure, perhaps a protective order for limited disclosure.
Because the political viability of the tests may depend on teachers and parents inspecting them to
see if they are measuring things of value, some commentators recommend that tests be kept
secure before but not after they are administered. Policy Issues, supra note 74, at 63-65.
"Attempts to cut down on testing costs by having a completely secure test(s) to be used year after
year therefore is probably not possible or desirable." Id. at 64.
106. Palm Beach County Schools, A Note Concerning APL Goals, Objectives, and Tasks
(1975). The Palm Beach County schools contracted with ACT for an APL test and planned to use
the test as a diploma requirement one year before the June 1979 statutory mandate.
1979] COMPETENCY TESTING PROGRAMS

support their children and grandchildren."'10 7 One task in the "Occu-

pational Knowledge" part asks students "to discuss proper behavior
and attitudes for keeping a job."10 8 The "Health" part includes the
following task: "discuss the physical and psychological benefits gained
when food is served attractively in a pleasant atmosphere."'10 9 And one
task in the "Government and Law" part asks 'students to discuss the
concept of party politics including why the two-party system has been
successful."' 1 0
The following "behavioral objectives" are being considered in an-
other state In as part of a competency test to be used as a prerequisite
to a high school diploma:

Accept personal involvement in the Law (report crime, serve jury duty);
State, interpret, and obey laws;
Describe appropriate and effective ways to change laws;
Explain the importance of basic social values (honesty, pride in work, dignity);
Describe ways to practice responsible citizenship;
Demonstrate self-discipline in carrying out responsibilities of citizenship;
Justify the value of individual differences in maintaining a democratic society;
Describe the effects of the profit system on work;
Define personal qualities needed for work success (loyalty, responsibility, dependa-
bility);
Demonstrate a positive self-concept;
Understand human sexuality (physical, emotional, social factors);
Name effective contraceptives;
Write voluntarily to entertain self (creative language, captions, poems);
Develop feelings of empathy and objectivity about the arts;
Justify the value of leisure activities to others;
Demonstrate behaviors appropriate to leisure activity.

These are but a few examples of proposed behavioral objectives that

indicate test content which may be inappropriate. Detailed considera-
tion will have to await the actual test items-that is, the specific
questions and especially the answers which are deemed to be correct
and incorrect. Tests will be legally vulnerable if they include items that
involve (1) coerced belief, (2) invasion of privacy, (3) unteachable or
unmeasurable content, (4) content that is not sufficiently matched with
a school's curriculum and instruction, and (5) content that is culturally
biased. While the first three kinds of inappropriate test content dis-
cussed in this Part are likely to be an immediate issue in relatively few

107. Id. at 28.

108. Id. at 31.
109. Id. at 10.
110. Id. at 34.
111. These behavioral objectives are from materials submitted to the author with an
understanding of partial confidentiality. Because many states may include similar areas in
subsequent stages of testing, it is more important to identify the potential problems than the
particular state.
 FORDHAM LAW REVIEW [Vol. 47

states,11 2 the latter two are likely to be at issue in most states, and are
discussed in Parts IV and V(C) respectively.

A. Coerced Belief
Many of the performance objectives listed earlier appear to assume
that certain beliefs (behaviors, feelings, attitudes, opinions, and values
generally) are the correct and proper ones, and that students can fail a
test and be denied a high school diploma if their answers do not
conform to those officially deemed correct. This kind of official bias
would be questionable even without imposition of a serious penalty for
failure to provide the "proper" answers, but combined with such a
penalty the test would almost certainly constitute the kind of coerced
belief prohibited by the first amendment of the Constitution.
Perhaps the most vivid expression of the first amendment's prohibi-
tion against coerced belief is found in West Virginia Board of Educa-
tion v. Barnette,113 in which the Supreme Court held the compulsory
flag salute in public schools to be unconstitutional. "The State asserts
power to condition access to public education on making a prescribed
sign and profession and at the same time to coerce attendance by
punishing both parent and child. The latter stand on a right of
self-determination in matters that touch individual opinion and per-
1 14
sonal attitude."'
In resolving this conflict in favor of self-determination, Justice
Jackson noted the special responsibility of boards of education "for
scrupulous protection of Constitutional freedoms of the individual, if
we are not to strangle the free mind at its source and teach youth
to discount important principles of our government as mere plati-
tudes."1 ' 5 The reality of a free mind is obviously incompatible with
compelled belief. Thus Justice Jackson wrote: "[T]he State may 'require
teaching by instruction and study of all in our history and in the
structure and organization of our government, including the guaran-
tees of civil liberty, which tend to inspire patriotism and love of
country,' 116 but such teaching is dearly distinguishable from com-
pulsion of students to declare a belief as is manifested by the compul-
sory flag salute. The flag salute is unconstitutional because it "invades
112. Most states plan to limit the scope of competency testing, at least initially, to either basic
literacy and numeracy skills or to functional literacy. According to a survey by Joan Baratz, "29
states test for minimum competency in basic skills, 9 test for functional literacy and 11 have other
areas in their testing programs, including citizenship, leisure skills, life long learning and attitudes
toward school." Educ. Daily, Oct. 17, 1978, at 5.
113. 319 U.S. 624 (1943).
114. Id. at 630-31.
115. Id. at 637.
116. Id. at 631 (quoting Minersville School Dist. v. Gobitis, 310 U.S. 586, 604 (1940) (Stone,
J.,dissenting)).
1979] COMPETENCY TESTING PROGRAMS

the sphere of intellect and spirit which it is the purpose of the First7
Amendment to our Constitution to reserve from all official control. "1
The Court applied "the limitations of the Constitution with no fear
that freedom to be intellectually and spiritually diverse or even con-
trary will disintegrate the social organization."'I s
Justice Jackson further wrote:
[F]reedom to differ is not limited to things that do not matter much. That would be a
mere shadow of freedom. The test of its substance is the right to differ as to things that
touch the heart of the existing order.
If there is any fixed star in our constitutional constellation, it is that no official, high
or petty, can prescribe what shall be orthodox in politics, nationalism, religion, or
other matters of opinion or force citizens to confess by word or act their faith
therein."1 9

So too any competency testing program by which the state imposes a

penalty for failure of students to conform to specified beliefs would
raise serious constitutional questions. Such a program would, in Jus-
tice Jackson's words, invade "the sphere of intellect and spirit which it
is the purpose of the First' 120Amendment to our Constitution to reserve
from all official control.
Four points are worth emphasizing with respect to the Barnette
case. First, although the plaintiffs in Barnette were Jehovah's Wit-
nesses, the prohibition against coerced belief is not based upon or
limited to religious objections. Justice Jackson specifically stated: "Nor
does the issue as we see it turn on one's possession of particular
religious views or the sincerity with which they are held.'' t12 Thus,
when he later defined the "fixed star in our constitutional constella-
tion," that definition is broad enough to proscribe official orthodoxy in
"politics, nationalism, religion, or other matters of opinion."' 2 2 Of
course, there may be test items that meet other constitutional stan-
dards, but are constitutionally invalid solely because they violate the
establishment or free exercise of religion clauses of the first amend-
12 3
ment.
Second, test items are not rendered valid simply because the explicit
or implicit values are "good" or innocuous values beyond the reach of
political or other controversy. Thus, Justice Jackson stated with re-

117. Id. at 642.

118. Id. at 641.
119. Id. at 642.
120. Id.
121. Id. at 634.
122. Id. at 642.
123. For legal standards used in applying the religion and establishment clauses in school
settings, see: Epperson v. Arkansas, 393 U.S. 97 (1968) (religion in curriculum); School Dist. v.
Schempp, 374 U.S. 203 (1963) (Bible readings); Engel v. Vitale, 370 U.S. 421 (1962) (official
prayers).
 FORDHAM LAW REVIEW [Vol. 47

spect to the compulsory flag salute: "Whether the First Amendment to

the Constitution will permit officials to order observance of ritual of
this nature does not depend upon whether as a voluntary exercise we
would think it to be good, bad or merely innocuous."' 124 Under the
Barnette rationale, the constitutional infirmity lies in the coercion, not
the validity of the belief itself.
Third, the constitutional infirmity of a test cannot be cured by any
consensual process such as a majority vote by the legislature, school
board, parents, or students. As the Barnette Court noted, "l[t]he very
purpose of a Bill of Rights was to withdraw certain subjects from the
vicissitudes of political controversy, to place them beyond the reach of
majorities and officials .... "125 The state legislature, school boards,
community, and others who may be responsible for designing a
competency testing program must do so within the limits set by the
Constitution.
Finally, the principles enunciated in Barnette are not absolute.
Although Justice Jackson stated, after his famous "fixed star"
metaphor that "[i]f there are any circumstances which permit an
exception, they do not now occur to us,' 26
he immediately added a
footnote distinguishing situations involving military discipline.' 27 The
application of the Barnette standards will vary somewhat according to
context and circumstance. An example is the landmark case of Tinker
v. Des Moines Independent Community School District.128
In Tinker, the Supreme Court considered the disruptive potential of
student expression in determining whether school officials could
penalize students for expressing their opposition to the Vietnam War.
The Tinker Court referred to Justice Jackson's fixed star language with
apparent approval, and held that school action prohibiting the stu-
dents from wearing black armbands in school to protest the Vietnam
War was a violation of the students' first amendment rights because
there was no evidence that such action was necessary to avoid material
and substantial interference with discipline or school work. The lan-
guage of the Court underscores the constitutional prominence of free
expression:
[I]n our system, undifferentiated fear or apprehension of disturbance is not enough to
overcome the right to freedom of expression. Any departure from absolute regimenta-
tion may cause trouble .... But our Constitution says we must take this risk... and
our history says that it is this sort of hazardous freedom-this kind of openness-that
is the basis of our national strength and of the independence and vigor of Americans
who grow up and live in this relatively permissive, often disputatious, society. 2 9
124. 319 U.S. at 634.
125. Id. at 638.
126. Id. at 642.
127. Id. at 642 n.19.
128. 393 U.S. 503 (1969).
129. Id. at 508-09 (citation omitted).
1979] COMPETENCY TESTING PROGRAMS

Denial of a high school diploma for failure to provide "correct"

answers on a competency test obviously presents a different factual
situation than exclusion from school for refusing to salute the flag or
for wearing black armbands. Nevertheless, it is unlikely that the
state's interest in these kinds of questions on a competency test is
strong enough to justify infringement upon the individual's interest in
free expression or, as discussed in the next section, privacy.

B. Invasion of Privacy
Some of the listed performance objectives might also be legally
vulnerable because they invade personal privacy. The Supreme Court
has held that a right to privacy is inherent in many of the first nine
amendments to the Constitution. 130 In Merriken v. Cressman,131 a
federal district court invalidated a test instrument on privacy grounds,
stating that the right to privacy "should be treated with as much
deference as free speech."' 132 The public school defendants in Merriken
were enjoined from administering a test questionnaire designed to
identify potential drug abusers. The questions asked students for
personal information about themselves, their families, and other stu-
dents. Although the factual situation in Merriken is different from that
involved in competency testing, the case is instructive in suggesting
some privacy limitations on test instruments that are administered to
public school students.
The court found fault with the highly personal nature of the
questionnaire, 133 the suspect results gained by the testing, and espe-
cially with two questionable uses of the test results. The court was
concerned that the test results could be used by law enforcement
officials who have subpoena power.1 34 Moreover, it was concerned
that the test results would be used for questionable remedial purposes,
specifically "to change the cognitive and affective domains of potential
drug abusers and other forms of deviant behavior" 13 s by compulsory
programs which seriously infringe upon individual liberty. Applying a
balancing test involving the right of an individual to privacy on the
one hand, and the right of the government to invade that privacy for
the sake of the public interest on the other, the court found: "As the
Program now stands the individual loses more than society can gain
.... "136 "There is too much of a chance that the wrong people for

130. Roe v. Wade, 410 U.S. 113 (1973); Griswold v. Connecticut, 381 U.S. 479 (1965).
131. 364 F. Supp. 913 (E.D. Pa. 1973).
132. Id. at 918.
133. Id.
134. Id. at 916.
135. Id.
136. Id. at 921.
 FORDHAM LAW REVIEW [Vol. 47

the wrong3 7reasons will be singled out and counselled in the wrong
manner."
Some of the listed performance objectives suffer from legal
infirmities similar to those identified in Merriken. For example, a
student might refuse to acknowledge on a competency test what the
authorities deem to be responsible citizenship, appropriate ways to
change a law, positive self-concept, proper behavior for work or
leisure situations, the success of the two-party system, or feelings of
empathy and objectivity about the a.;ts. The student might assert a
privacy interest and simply refuse to disclose his or her opinions on
these matters whether or not they happen to coincide with the official
view. On the other hand, the student might answer the questions but
assert a free expression right to have different opinions regarding the
value judgments inherent in the test items.
In addition to the potential illegality of involuntary collection of
information about a student's "deviant" attitude, the use of that
information may also be illegal. Apart from misuse, if the school
cannot protect the confidentiality of the information, there is the
ominous prospect that, as in Merriken, classes may be created "to
remedy the problem." There is also the possibility with many compe-
tency programs that students will be denied a high school diploma if
the remedial attempts are unsuccessful.

C. Unteachable or Unmeasurable Content

Tests that go beyond basic skills and attempt to measure "the
affective aspects of social responsibility, good citizenship, self-concept,
and job preparedness"'' 38 run a high risk of violating a student's free
expression and privacy interests. Such tests are also legally vulnerable
not only because it is questionable that these "skills" can be measured
but also questionable whether they can be generally taught given the
current state of instructional and curricular research. Thus, Madaus
and Airasian write:
While cognitive competencies may be taught and generally attained, teaching the
affective aspects of social responsibility, good citizenship, self-concept, and job pre-
paredness to all pupils in a school calls for knowledge and techniques simply
unavailable given the current technology of instructional and curricular research ....
[Olne continually encounters behaviors such as social responsibility, career prepared-
ness and personal development described as "skill areas," with the implication that
these behaviors can be taught in much the same way that one teaches pupils to recall
number facts, balance a checkbook or read with comprehension. 139
Madaus and Airasian make a similar point about the state of the art

137. Id.
138. Madaus & Airasian, supra note 15, at 81-82.
139. Id.
1979] COMPETENCY TESTING PROGRAMS

with respect to the measurement of affective behaviors. 140

Punishment for bomething that is not one's personal fault violates a
cardinal principle of liberty. Punishment, such as denial of a diploma,
for failure to learn that which is not teachable or measurable raises
serious legal questions. This legal concept is discussed herein with
respect to tests that measure knowledge and skills not taught in the
14 1
classroom.
If a decision is made to include social responsibility, job prepared-
ness, and similar areas in a competency testing program, the perfor-
mance objectives and test items should be designed to elicit a student's
knowledge (for example, know, describe, list, explain, apply) rather
than opinions or values (for example, accept, value, appreciate, sup-
port, justify). Many of the listed performance objectives that have a
value orientation could be easily revised in knowledge terms. For
example, "accept personal involvement in the law" could be revised to
read "explain society's (the state's) interest in personal involvement in
the law." The objective is to avoid value judgments to the extent
possible, 14 2 or at least to avoid requiring implied acceptance of the
value in the answer deemed to be correct. There is an obvious and
crucial difference between asking a student to "describe what most
employers consider appropriate in a work situation," and "describe
what is appropriate behavior in a work situation."' 1 43 The latter uses
the knowledge term "describe," but is obviously written in value rather
than knowledge terms.

UT1. ADEQUACY OF PHASE-IN PERIODS

A. Issues
Many competency programs are being imposed upon students late in
their secondary education with little prior notice. Imposing the re-
quirement one year before graduation means that students will have
spent their first ten or eleven years in the school system without notice
or knowledge that passing a competency test would be a condition for
acquiring a diploma. In fact, the school district would have explicitly
approved their progress by promoting them each year, even if they did
not have basic competencies. It is likely that many if not most of those
students failing the test would have studied differently in earlier years
had they been given such notice-and teachers might have taught
differently as well. The competency test is designed to ensure that
140. Id. at 87.
141. See notes 164-67 infra and accompanying text.
142. It is not possible, even if desirable, to design a test or run a school totally free of value
judgments.
143. These examples are for general illustrative purposes. Actual test items would obviously
not be written at this level of generality.
 FORDHAM LAW REVIEW [Vol. 47

minimal competency is acquired after twelve years of schooling, but

students in this situation would not have received notice until their
tenth or eleventh year of schooling.
The exact date of notice will vary from school to school. In most
districts there will be one date when students are given general notice
of the competency requirement for a diploma, and a later date when
students are given notice of the specific performance objectives to be
measured by the competency test. Students, parents, and teachers
should be given notice of these performance objectives and the assess-
ment procedures as soon as possible after they have been estab-
144
lished.
The complaint in Debra P. v. Turlington, challenging the Florida
competency testing program, alleged in part:
Plaintiffs and members of the class they seek to represent received inadequate notice
to prepare for the test. The enabling legislation for the testing program was passed in
1976. State test standards were not established until April, 1977. The first administra-
tion of the test occurred in October, 1977 and the plaintiffs herein failed.

Defendants have changed the requirements for awarding high school diplomas after
plaintiffs and members of the class they seek45to represent had completed more than ten
of their twelve years of public schooling.'
Two students from another Florida school district make the point
more bluntly. "When we got to school, they told us we have to have 20
credits to graduate," said Willie Baker. "Now they jump up and tell us
we have to pass this test to graduate." Said Carrie Howard: "They
should be teaching us a lot of things on the test, but they haven't given
it to us yet.' 14A6 person can't learn something in as short a time as they've
given US."
Educators generally agree that the crucial time for teaching, learn-
ing, and remedying basic skills is in the early years of schooling.
Accomplishing remedial objectives at the secondary level is more
difficult and expensive than in the elementary years, and, therefore, a
sound competency program should focus on the early years of school-
ing. The $10 million appropriated by the Florida legislature for
remedial instruction translates into only $246 for each student who

144. Performance objectives help teachers and students focus on the knowledge and skills to
be measured by the test. Some educators suggest that the test item specifications should also be
designed and used for instructional and remedial purposes. Test item specifications are based
upon performance objectives, but are usually more detailed in order that specific test items can be
developed. Because item specifications define what the test items measure and how it Is
measured, they could be integrated with instructional efforts without disclosing the exact test
questions. Item specifications, however, are not often subject to public disclosure and are
therefore unavailable for instructional use.
145. Complaint, supra note 63, at 36, 54. See generally id. at 1 52-57.
146. Tallahassee Democrat, Sept. 30, 1977, § A, at 1, col. 1.
1979] COMPETENCY TESTING PROGRAMS

failed the October 1977 test. 14 7 As emphasized in a study of the

Florida program sponsored by the Florida Teaching Profession and the
National Education Association:
It appears as if the current class of eleventh graders who are Black and poor were
sacrificed for the purpose of rapid implementation of the functional literacy segment of
the Accountability Act.

The critical issue is whether short-term remediation programs can be effective in

providing to those poor and Black children knowledge and skills which the schools
have not been successful in imparting over the last 11 years. 48
The point, of course, is not to withhold remedial efforts at the
secondary level, but rather to question the fairness and legality of
withholding diplomas where belated efforts are not successful.
One of the most important decisions for policymakers who are
committed to this kind of program is setting the initial date when
diplomas will depend upon the results of a competency test. Problems
of adequate notice and adequate match between test and instruction
will probably correct themselves in the long term. The task for
policymakers is to make sure that whole generations of low-scoring
students are not sacrificed in the short term for the sake of rapid
program implementation.
B. Analysis
Traditional notions of due process should require adequate prior
notice of any rule that could cause irreparable harm to a person's
educational or occupational prospects. 149 Mahavongsananv. Hall,15 0 a

147. These figures are based upon the 40,700 students who failed the October 1977 test. As of
this writing, there is a bill in the Florida legislature to increase the amount appropriated for
remedial instruction to $26 million (approximately $640 per student based on 40,700 students).
148. National Education Association, Impact of Minimum Competency Testing in Florida,
Today's Educ., Sept.-OcL, 1978, at 35-36. The cited article reprints excerpts from a report
prepared by a study panel consisting of Ralph W. Tyler (Chairperson), Stephen Lapan, Judy
Moore, L. Wendell Rivers, and Donna Skibo. The study panel, sponsored by the Florida
Teaching Profession and the National Education Association (NEA), held hearings in Florida in
the spring of 1978. The panel identified "several serious defects in implementation" that appeared
to be "largely the effects of rushing to establish the program without considering all the important
aspects of iL" Id. at 31.
149. The legal standard applied in modem substantive due process cases (under the four-
teenth amendment of the Constitution or comparable state due process provisions) is usually not
carefully set out, but state action is usually illegal if it (1) is arbitrary or capricious, (2) does not
achieve any legitimate state interests, (3) frustrates any legitimate interest the state might have, or
(4) is fundamentally unfair. See McClung, The Problem of the Due ProcessExclusion: Do Schools
Having a ContinuingResponsibilityTo Educate Childrenwith BehaviorProblems?, 3 J. L. & Educ.
491, 495-501 (1974). Whatever the exact wording of the test under substantive due process, the
general standard is that state action cannot be unreasonable, with unreasonableness being
construed narrowly (e.g., rational persons would not disagree). A stronger standard has been
applied in some situations. See St. Ann v. Palisi, 495 F.2d 423 (5th Cir. 1974).
150. 529 F.2d 448 (5th Cir. 1976).
 FORDHAM LAW REVIEW [Vol. 4 7

Fifth Circuit decision, provides some support for the proposition that
students must be provided with adequate notice of any significant
change in graduation requirements, even though the court found that
the plaintiff in that case had not been denied procedural or substantive
due process because she received "timely notice" that passing a com-
prehensive examination would be a prerequisite to a master's degree in
education.1 5 1
The legal argument for adequate notice of significant changes in
graduation requirements is stronger in the context of elementary and
secondary public schools than in postsecondary education because
courts usually apply a stricter standard of review to practices at the
elementary and secondary level where education is compulsory. Notice
of a competency test requirement for a high school diploma would
have to be much earlier than in Mahavongsanan because twelve rather
than just two or three years of education are being tested. The plaintiff
in Mahavongsanan received notice of the new requirement only six
months after starting the program. Whatever notice is considered
adequate for the competency testing requirement, notice after most of
one's educational program is already completed seems clearly in-
adequate, especially if the competency test is designed to measure
knowledge and skills not previously taught. in the district's classrooms.

IV. MATCHING TEST AND INSTRUCTION

Most persons would agree that fairness requires a school's cur-

riculum and instruction to be matched in some way with whatever is
later measured by the competency test. In other words, the test would
be unfair if it measured what the school never taught. This concept
should be considered in terms of curricular and instructional validity.

A. Curricularand InstructionalValidity
Curricularvalidity is a measure of how well test items represent the
objectives of the curriculum. 15 2 An analysis of curricular validity
would require comparison of the test objectives with the school's
course objectives. For example, if the curriculum was not designed to
teach functional competency, it would be unfair to deny students their
diploma because they did not learn to be functionally competent. In
this situation, failure on the competency test should reflect on the
schools, which are not offering an appropriate curriculum.
A competency test should also have what may be called instrctional
validity. 153 Even if the curricular objectives of the school correspond

151. Id. at 450.

152. See Cronbach, Essentials of Psychological Testing 397 (1970); Dick & Hagerty, Topics in
Measurement: Reliability and Validity 96 (1971).
153. The discussion in this Article of the various types of test validity was developed with the
assistance of Diana Pullin of the Center for Law and Education. Because neither the APA
1979] COMPETENCY TESTING PROGRAMS

with the competency test objectives, there must be some measure of

whether the school district's stated objectives were translated into
topics actually taught in the district's classrooms. While a measure of
curricular validity is a measure of the theoretical validity of the
competency test as a test instrument to assess the success of students,
instructional validity is an actual measure of whether the schools are
providing students with instruction in the knowledge and skills mea-
sured by the test. Instructional validity obviously does not require
prior exposure of the student to the exact questions asked on the
test, 154 but it does require actual exposure of students to the kind of
knowledge and skills that would enable a student to answer the test
questions. This will present difficult problems of proof in some cases;
in others, it will be relatively easy to show that the test is measuring
what the school never taught. Some possible approaches to assessing
the instructional validity of a test are discussed in Part VII(C).
It is important to note that content validity, as defined by the
APA, 15 5 does not ensure either curricular or instructional validity. The
concepts are related but distinguishable. Content validity is a measure
of how well test items represent the performance domain that the test
purports to measure (for example, adult performance skills), but it is
not necessarily a measure of how well the test items and performance
domain represent either a particular school's curricular objectives or
instruction received. Instructional validity should be the central con-
cern, because content and curricular validity mean very little in this
context if the test items are not representative of instruction actually
received by the student.
Test developers will be reluctant to evaluate the curricular or
instructional validity of their tests, but they have some kind of
professional responsibility (beyond the usual disclaimers) to ensure that
their tests do not measure skills and knowledge that were never taught
in school, particularly when they design a test instrument for an
individual user. The ultimate responsibility for the curricular and
Standards nor the educational literature generally seems to address the issue of matching the test
with actual instruction (especially important in the fair administration and assessment of any
minimal competency program), we developed the concept of "instructional validity" for this
purpose.
154. This raises the classic controversy about "teaching to the test." See Madaus & Airasian,
supra note 15, at 82-90; note 144 supra.
155. APA Standards, supra note 67, at 28. The general nature of the discussion of content
validity in the APA Standards is perhaps explained by the heavy focus on employment testing.
The APA Standards were developed during a period when competency testing as a prerequisite to
a high school diploma was not an issue. Curricular and instructional validity are irrelevant
concepts in most employment tests. Teachers certification tests, however, are an important ex-
ception. See, e.g., United States v. South Carolina, 15 Fair Empl. Prac. Cas. 1196 (D.S.C. 1977),
aff'd mem. sub. nom. National Educ. Ass'n v. South Carolina, 434 U.S. 1026 (1978). In fact, the
APA Standards were not written with the intent that they have the force of law. Novick, suprm
note 68, at 47. Ideally, the proposed revision of the APA Standards will incorporate concepts like
curricular and instructional validity for competency tests used as a prerequisite to a high school
diploma or grade-to-grade promotion.
 FORDHAM LAW REVIEW [Vol. 47

instructional validity of the test, however, should be with the school or

school system that uses the test. A school or school system that cannot
ensure the curricular and instructional validity of its competency tests
should not use them1 as56
a basis for denying promotion or a diploma to
any of its students.
Questions of curricular and instructional validity are relevant, given
either of two general purposes of competency testing: (1) to measure
the student's mastery of the school's curriculum or (2) to predict the
minimal competence required in the adult world.'5 7 The terminology
will vary from school to school, and some schools will merge the two
concepts by deciding that their curricula should be based upon mini-
mal adult competence. The different versions of adult competency tests
usually go beyond basic proficiency skills, because they seek to mea-
sure an individual's ability to apply basic skills (including literacy) to
necessary adult-life role activities such as those of a consumer, pro-
ducer, or citizen. Examples include ability to understand common
indices for comparison shopping, to understand the nutritional ingre-
dients necessary for a balanced diet, to understand a contract for a car
loan or home mortgage, to read and understand a newspaper, to fill
out a job or loan application, to complete a tax form, to balance a
checkbook, to follow a recipe in preparing a meal, to understand
proper behavior and attitudes for getting and keeping a job, to use
leisure time productively, and to participate as a citizen in the com-
munity, state, and nation.' 5 8

B. Basic Literacy Skills Versus Adult-Life Skills

Some people do not think that the distinction between basic
literacy/numeracy skills and adult-life skills is an important one since
they believe that a student with basic literacy skills will have no
difficulty passing an adult-life skills test even if he or she has not been
exposed to adult-life skills in the classroom. They argue that students
who can add and subtract a series of four-digit numbers on a basic
numeracy skills test item, for example, will be able to do the same in
the context of a tax form item on an adult-life skills test. However,
many students, especially low-scoring students, will have difficulty
with the kind of transference skills called for in an adult-life skills
item. 15 9 In fact, many of these students probably score low because

156. For a discussion of curricular and instructional validity in the context of competency
testing, see: Policy Issues, supra note 74, at 67-71; pt. VII(C) infra.
157. If the purpose of the proficiency test is to measure the minimal competence required in
the adult world, the test should also have predictive validity or an acceptable substitute. See
Policy Issues, supra note 74, at 73, 74; note 74 supra and accompanying text.
158. Some of these examples suggest inappropriate test content. See pt. II supra.
159. "One system in Massachusetts completed a study documenting that students had learned
traditional school skills, but needed work in the area of transferring or applying of those skills to
everyday life." Minimum Competency Testing, supra note 2, at 26; see Office of Program
1979] COMPETENCY TESTING PROGRAMS

they have difficulty with transference. Therefore, school districts that

plan to test for adult-life skills should have curricula and instruction
that emphasize transference as well as the other knowledge and skills
necessary to answer the adult-life skills items.
An adult-life skills item, such as adding and subtracting four-digit
numbers on a tax form, is also more difficult than its basic numeracy
components for other reasons, including the fact that (1) the forms
usually require literacy as well as numeracy skills, and (2) an official
form can be distracting and intimidating, especially if the students are
not familiar with the form. Of course, it is possible to write a basic
literacy or numeracy skill item which is more difficult than its
adult-life skills counterpart or to make an adult-life skills test easier by
allowing a lower cut-off score than for the literacy/numeracy skills test.
But in general, an adult-life skills test designed to measure the same
literacy and numeracy skills as a basic skills test will be more difficult
for most students; and a school's curricula and instruction should
account for this difference.
Schools may not have taught the life skills that are measured by
many competency tests. "Interest rates, checkbooks, tax forms, etc.,
were included in the 7-8 grade arithmetic books in the 40's and 50's but
dropped out of sight in the 60's with the advent of modern math, the
exception being courses like business arithmetic where such topics are
still covered." 160 There may be a mismatch between test and instruc-
tion even where the test is limited to basic literacy and numeracy
skills. For example, some schools may not have taught or emphasized
some basic skills such as metric conversion that are measured by many
competency tests. Even where basic skills are part of the curriculum,
Popham notes that instructional differences surface when such general
concepts as "the ability to perform fundamental mathematical opera-
tions" 161 are translated into specific test items.1 62 This problem will
be especially acute where (1) the state uses a single statewide test that
does not account for the diverse curricula and instruction developed in
individual school districts, or (2) an individual district purchases a
standardized test that was 163
developed without regard to local cur-
riculum and instruction.
Where an adult competency purpose is involved, as is the case in

Evaluation and Research, California State Dep't of Education, Getting Inside a Student Compe-
tency Test 7 (1978).
160. Policy Issues, supra note 74, at 69. For a more complete description of the ways in which
conventional curriculum and instruction are not well matched with competency based tests, see
Spady & Mitchell, Competency Based Education: OrganizationalIssues and Implications, Educ.
Researcher, Feb., 1977, at 9-10, 13.
161. Popham, supra note 81, at 471.
162. Id. at 470-71.
163. See note 261 infra and accompanying text.
 FORDHAM LAW REVIEW [Vol. 47

many school systems where various adult performance level tests have
been adopted, it is less likely that the school will in fact have taught
what is measured by the test. If this is the case, greater curriculum
revision and longer phase-in periods will be necessary. A competency
test that measures adult-life role skills that were never taught in the
school and then is used as a basis for denying a diploma is arguably so
arbitrary as to violate due process of law. 164 A competency test lacking
curricular or instructional validity might violate substantive due pro-
cess because then the students are being penalized even though they
cannot be personally faulted for poor performance on the test. Support
for this argument can be found in St. Ann v. Palisi,165 in which the
Fifth Circuit held that a school board regulation violated substantive
due process because it allowed school children to be suspended for
their parent's misconduct. In so holding, the court said: "[T]he children
do not complain that they were denied the constitutional right to an
education, but that they were punished without being personally
guilty. Thus a cardinal notion of liberty is involved and substantive
due process is applicable."'1 6 6 Since the practice established "a sig-
nificant encroachment upon a basic element of due process," the court
required the school to meet "a substantial burden" to justify this
encroachment, including proof that reasonable alternative
167
means to
achieve the stated objective were not available.
In another analogous case, a three-judge court in United States v.
South Carolina168 held, inter alia, that a study conducted by ETS for
South Carolina, to demonstrate that the content of the National
Teacher Examination (NTE) matched the content of teacher training
programs in the state, was sufficiently trustworthy to sustain the state's
burden under Title VII of the Civil Rights Act of 1964169 to show the
validity of the test, even though the validation was not related to a
particular school's instruction i7 0 or to job performance. With Justices
White and Brennan dissenting, and Justices Marshall and Blackmun
not taking part, the Supreme Court affirmed the decision of the lower
court without writing an opinion. 17 1 Since this case involved employ-
ment testing for purposes of hiring and classifying teachers, it does not
provide a firm basis for prediction about judicial review of validation
studies of competency tests used for purposes of awarding high school
diplomas.
164. See note 149 supra and accompanying text.
165. 495 F.2d 423 (5th Cir. 1974).
166. Id. at 426-27.
167. Id. at 427.
168. 15 Fair EmpL. Prac. Cas. 1196 (D.S.C. 1977), aff'd mer. sub nor. National Educ. Ass'n
v. South Carolina, 434 U.S. 1026 (1978).
169. 42 U.S.C. §§ 2000a to 2000h-6 (1976).
170. The adequacy of the match between the test and a particular school's instruction (that is,
instructional validity) was not at issue in this case.
171. National Educ. Ass'n v. South Carolina, 434 U S. 1026 (1978).
1979] COMPETENCY TESTING PROGRAMS

Plaintiffs in Debra P. include in their complaint the allegation that

the Florida statutes have never required school districts to provide,
and the defendant school district had not in fact provided, instruction
and curriculum in all areas covered by the functional literacy test.172
They allege that defendants' failure to ensure that there is a match
between the instruction and curriculum provided in the classroom and
the areas covered by the test constitutes the kind of arbitrary, caprici-
ous, and fundamentally unfair action prohibited
73
by the due process
guarantee of the fourteenth amendment.
In sum, a number of factors, such as statewide tests and tests of
adult-life skills, suggest that many competency tests may measure
what many students have not been taught. The need for some kind of
match between test and instruction seems basic to the concept of fair
play, and an initial question will be whether the test in fact reflects
curriculum and instruction. Whether courts will require instructional
and/or curricular validity of competency tests that are used as a
prerequisite to a high school diploma is a separate, unanswered
question. If so, the requisite degree of match between test and instruc-
tion and/or curriculum, 74 and the kind of evidence used to determine
that match, 175 are additional, unanswered questions.

V. RACIAL AND LINGUISTIC DISCRIMINATION

While substantial numbers of white middle-class students cannot

meet minimal competency standards, some evidence indicates that a
disproportionate number of black and Hispanic students will not be
able to meet the competency test requirements. This pattern has in fact
occurred where competency testing programs have been im-
plemented. 176 For example, the statewide assessment results released
by the Florida State Department of Education on March 16, 1978
showed that approximately 75% of the state's black students (juniors)
had failed the mathematics part of the statewide functional literacy
test. Approximately 40% of Hispanic, 30% of Asian, and 25% of white
students also failed the mathematics part of the test.' 7 7 Results of

172. Complaint, supra note 63, at 37-38.

173. Id. at 1 56-57; see note 149 supra.
174. See notes 258-62 infra and accompanying text.
175. Evidence of instructional validity might include teacher testimony, homework samples,
and materials used in the classroom. See notes 262-63 infra and accompanying text. Some
commentators are concerned that instructional validity might necessitate more extensive record-
keeping procedures. See Policy Issues, supra note 74, at 69-71; Duke, Donmoyer, & Farman,
Emerging Legal Issues Related to Classroom Management, 60 Phi Delta Kappan 305, 307 (1978).
See generally note 255 infra.
176. See note 44 supra.
177. Florida Dep't of Education, Statewide Assessment Results by Demographic Categories,
 FORDHAM LAW REVIEW [Vol. 47

earlier testing in Florida also indicated a disproportionate failure rate

among minority students.
Detailed analysis of racial and linguistic discrimination must be
tailored to the specific factual situation in each school district. Basi-
cally, any school district will fall into one of the following three
categories: (1) a district with a court finding of prior discrimination; t78
(2) a district with no such finding but vulnerable to such a finding if
the issue were to be litigated; and (3) a district with no such finding
since the district has provided equal educational opportunities to all of
its students. There are many variations on these three situations, and
each district should seek legal consultation to determine whether past
practices in the district combined with its version of competency
testing constitute racial or linguistic discrimination. Districts that have
already been found to have discriminated, or that are vulnerable to
such a finding, must be especially sensitive to the discriminatory
potential of competency testing and should design their programs to
account for possible problems.

A. Racial Discrimination
Some black parents in desegregated communities see a racial motive
behind competency testing. They say that competency testing was not
a concern at either black or white schools until the schools in their
district were desegregated, at which time competency testing was
introduced "to protect standards." The effect can be "resegregation"
within the school according to test results (or other forms of tracking)
since unequal educational opportunities may cause black children to
score lower than their white counterparts.
Whether or not a racial motive is involved, such practices are
arguably unconstitutional in formerly segregated districts. In compar-
able situations, the federal courts have held that practices which carry
forward the effects of prior racial discrimination are prohibited. For
example, in Gaston County v. United States,179 an action brought
under the Voting Rights Act of 1965,180 the Supreme Court held
that it was appropriate for a court to consider whether a literacy or
educational requirement had the effect of denying the right to vote
on account of race or color because the state had maintained separate
and inferior schools for its black residents who were then of voting
age. The Court stated: "[W]e cannot escape the sad truth that through-
Fall, 1977 Tests, Grade Eleven Highlights (Functional Literary) (1978). The extent to which
Florida will be able to reduce this disproportionate failure rate, and the overall failure rate, by
the time these students graduate is unclear as of this writing.
178. For example, the district may be under a court-ordered desegregation plan pursuant to a
judicial finding of purposeful segregation and/or provision of unequal educational opportunities.
179. 395 U.S. 285 (1969).
180. 42 U.S.C. §§ 1971, 1973-1973p (1976).
19791 COMPETENCY TESTING PROGRAMS

out the years Gaston County systematically deprived its black citizens
of the educational opportunities it granted to its white citizens. 'Im-
partial' administration of the literacy test today would serve only to
perpetuate these inequities in a different form."' 18 1
Similarly, the Fifth Circuit has determined that the testing necessary
for ability grouping cannot be applied to black students for the first
time in the years immediately following desegregation. 1 82 This "prior
effects" principle arguably applies in other cases of serious injury, such
as denial of grade promotion or a high school diplomd. It forms the
basis for one of the claims in the challenge to the Florida competency
testing program.
In Debra P., plaintiffs cite a lengthy history of deliberate racial
segregation of schools in Hillsborough County and in the state of
Florida generally. 18 3 They state that after the Florida Pupil Assign-
ment Law 184 was declared unconstitutional in 1962, the county and
state continued to resist the dismantling of the dual school system, as
evidenced by the various federal court findings in Mannings v. Board
of Public Instruction.185 Plaintiffs further claim that the Hillsborough
public schools were unconstitutionally segregated by race when the
plaintiffs entered school in the first grade, and as a result they received
an inferior education which contributed to their failing the test. 186 As a
result, the plaintiffs contend that they have been resegregated in
predominantly black remedial classes, and are in imminent danger of
diploma denial, permanent stigmatization, diminished employment
opportunities, loss of income, and restricted access to higher educa-
tion. 187 In sum, the plaintiffs argue that Florida's competency testing
program perpetuates the effects of past racial discrimination in viola-
tion of the equal protection clause of the Constitution and Title VI of
the Civil Rights Act of 1964.188 In summary, the legal prohibition
against carrying forward the effects of past discrimination is a prohibi-
tion against placing one injury, diploma denial, on top of another,
unequal educational opportunities-in other words, "blaming the vic-
tims" for conditions caused by earlier governmental discrimination.

181. 395 U.S. at 296-97; see Oregon v. Mitchell, 400 U.S. 112 (1970); Kirksey v. Board of
Supervisors, 554 F.2d 139 (5th Cir.), rev'g en banc 528 F.2d 536 (5th Cir.), cert. denied, 434
U.S. 968 (1977).
182. United States v. Gadsden County School Dist., 572 F.2d 1049 (5th Cir. 1978); McNeal
v. Tate County School Dist., 508 F.2d 1017 (5th Cir. 1975); Moses v. Washington Parish School
Bd., 456 F.2d 1285 (5th Cir.), cert. denied, 409 U.S. 1013 (1972); Lemon v. Bossier Parish School
Bd., 444 F.2d 1400 (5th Cir. 1971).
183. Complaint, supra note 63, at T 20-29.
184. Act of July 26, 1956, ch. 31380, 1955-1956 Fla. Laws 30 (repealed 1965).
185. 427 F.2d 874 (5th Cir. 1970).
186. Complaint, supra note 65, at 25.
187. Id. at 39.
188. Id. at T,51.
 FORDHAM LAW REVIEW [Vol. 47

Plaintiffs may have to prove a discriminatory purpose when chal-

lenging competency testing programs in school districts not recently
desegregated or found to be subject to prior discrimination. The
Supreme Court held in Washington v. Davis' 89 that the dispropor-
tionate racial impact of a police department's personnel test was not
sufficient to establish an unconstitutional racial classification without
proof that it reflected a racially discriminatory purpose. 190 The Court,
however, stated that such disproportionate racial impact could be
evidence of a discriminatory purpose. 19 '
In a concurring opinion, Justice Stevens noted that "the line between
discriminatory purpose and discriminatory impact is not nearly as
bright, and perhaps not quite as critical," 192 as the majority's opinion
suggested. "Frequently the most probative evidence of intent will be
objective evidence of what actually happened rather than evidence
describing the subjective state of mind of the actor. For normally the
actor is presumed to have intended the natural consequences of his
deeds. ' 19 3 Given the studies cited above, the natural consequence of
most competency testing programs will be racial differentiation.
The standards used to prove intent to segregate in school desegrega-
tion cases are instructive for purposes of proving discriminatory intent
in competency testing cases. In Village of Arlington Heights v. Met-
ropolitan Housing Development Corp., 194 the Supreme Court sug-
gested that evidence as to the following factors might have probative
value in proving intent: historical background, the sequence of events
leading up to the challenged decision (including substantive and pro-
cedural departures from the norm), legislative history, and testimony
from officials. 9 5 The "natural and foreseeable" standard for establish-
ing intent to segregate has been approved by at least five federal circuit
courts. 1 96 The Second Circuit in Arthur v. Nyquist 197 said:
When ... [school board] actions have the "natural, probable, and foreseeable result of
increasing or perpetuating segregation," a presumption of segregative purpose is
created. The burden of proof then shifts to defendant officials... to demonstrate that
no reasonable alternative policy would have achieved the same permissible educational
goals with less segregative effect.' 98

189. 426 U.S. 229 (1976).

190. Id. at 246.
191. Id. at 242.
192. Id. at 254 (Stevens, J., concurring).
193. Id. at 253 (Stevens, J., concurring).
194. 429 U.S. 252 (1977).
195. Id. at 267-68.
196. See United States v. Texas Educ. Agency, 564 F.2d 162, 168 (5th Cir. 1977) (standard
adopted by First, Second, Fifth, Sixth, and Seventh Circuits).
197. 573 F.2d 134 (2d Cir.), cert. denied, 99 S. Ct 179 (1978).
198. Id. at 142-43.
1979] COMPETENCY TESTING PROGRAMS

The Fifth Circuit, in United States v. Texas Education Agency, 199

held that the "natural and foreseeable" standard for establishing
segregative intent is consistent with the Supreme Court decisions in
Washington v. Davis and Village of Arlington Heights which stress
that otherwise neutral state action does not violate the fourteenth
amendment's equal protection clause solely because it has a dispropor-
tionate impact on a racial minority. Thus the Fifth Circuit concluded:
"When the official actions challenged as discriminatory include acts
and decisions that do not have a firm basis in well accepted and
historically sound non-discriminatory social policy, discriminatory in-
tent may be inferred from the ' 20 0
fact that those acts had foreseeable
discriminatory consequences.
In Washington v. Davis, the Supreme Court also distinguished
between the constitutional standard and the Title VII standard on
testing, noting that the latter was more stringent since it incorporated
an effect rather than a purpose standard. 20 1 Thus when a test or
practice disqualifies substantially disproportionate numbers of blacks
in hiring and promotion decisions, the burden under Title VII shifts to
the employer to validate the test or practice in terms of job perfor-
mance and to show that the test or practice is sufficiently job-related.
The HEW regulations implementing Title VI of the Civil Rights Act
of 1964202 incorporate a similar effect (rather than purpose) standard,
prohibiting practices which have the effect of discriminating against
individuals on the ground of race, color, or national origin.2 0 3 This
Title VI effect standard has been cited with approval and applied by
the Supreme Court in Lau v. Nichols,20 4 although recently ques-
tioned. 20 5 The Title VI regulations also incorporate the Gaston County
20 6
principle against carrying forward the effects of past discrimination,
and add to it an affirmative obligation to take steps to remedy those
2
effects. 07
199. 564 F.2d 162 (5th Cir. 1977).
200. Id. at 168.
201. 426 U.S. at 246-48.
202. 42 U.S.C. § 2000d (1976).
203. Title VI states: "No person in the United States shall, on the ground of race, color, or
national origin, be excluded from participation in, be denied the benefits of, or be subjected to
discrimination under any program or activity receiving Federal financial assistance." Id. Regula-
tions issued by HEW pursuant to this section state that a recipient of federal funds "may not...
utilize criteria or methods of administration which have the effect of subjecting individuals to
discrimination because of their race, color, or national origin, or have the effect of defeating or
substantially impairing accomplishment of the objectives of the program as respect individuals of
a particular race, color, or national origin." 45 C.F.R. § 80.3(b)(2) (1978) (emphasis added); see
id. § 80.3(b)(1)(iv)-(v), .5(e) (1978).
204. 414 U.S. 563, 568 (1974).
205. See notes 209-10 infra and accompanying text.
206. See notes 179-81 supra and accompanying text.
207. 45 C.F.R. § 80.3(b)(6)(i) (1978) provides: "In administering a program regarding which
 FORDHAM LAW REVIEW [Vol. 47

Although HEW has not taken any position on competency testing as

of this writing, application of these Title VI standards to public school
testing programs is indicated by an HEW memorandum requiring
schools to take steps "to adopt and implement procedures to insure
that test materials and other assessment devices used to identify,
classify and place exceptional children are selected and administered in
a manner which is non-discriminatoryin its impact on children of any
race, color, [or] national origin .... -208 Because virtually all public
schools are subject to Title VI regulations, competency testing pro-
grams that have a disproportionate effect on blacks or other protected
minorities should be examined in light of Title VI standards, especially
where there is evidence of racial bias in the test itself or in the
administration of the test.
Applying Title VI standards to competency testing, however, is
complicated by the Supreme Court's decision in University of Califor-
nia Regents v. Bakke. 20 9 In that landmark case setting the legal
parameters of affirmative action programs, Justice Powell and four
other justices agreed that Title VI proscribes only those racial
classifications that would violate the equal protection clause if em-
ployed by a state or its agencies. Given the Court's earlier decision in
Washington v. Davis that a disproportionate racial impact is not
sufficient to prove a constitutional violation, Bakke would appear to
undercut the effect standard incorporated in the Title VI regulations.
Justice Powell, writing for a divided Court, nevertheless indicated
approval of Lau v. Nichols2 1 0-a case in which no discriminatory
purpose was proven, and the Title VI effect standard was used to
require English-language instruction to students of Chinese ancestry
who do not speak English. Therefore, the status of the effect standard
incorporated in HEW's Title VI regulations will remain somewhat
uncertain until further clarified by the courts. Plaintiffs challenging
competency testing programs, however, should probably proceed
under the assumption that they will have to prove discriminatory
intent, except where the program carries forward the effects of proven,
past discrimination.
B. Linguistic Discrimination
Similar standards apply to competency testing programs that have a
disproportionate effect on Hispanic children. Some school districts are
the recipient has previously discriminated against persons on the ground of race, color, or
national origin, the recipient must take affirmative actiDn to overcome the effects of prior
discrimination." For illustrations of the principle, see 45 C.F.R. §§ 80.5(i)-(j) (1978).
208. U.S. Dep't of Health, Education and Welfare, Identification of Discrimination In the
Assignment of Children to Special Education Programs 3 (1975) (memorandum by M. Gerry,
Acting Director, Office for Civil Rights) (emphasis added).
209. 438 U.S. 265 (1978).
210. 414 U.S. 563 (1974).
19791 COMPETENCY TESTING PROGRAMS

imposing the competency requirements without a Spanish translation

of the test or corresponding curricular and instructional modification.
Such practices may not satisfy HEW's standards requiring public
schools to take affirmative steps to remedy the linguistic exclusion of
21
limited and non-English speaking children. '
The constitutional prohibition against carrying forward the effects of
past discrimination may also apply. Consider, for example, a school
district that-pursuant to Lau v. Nichols and related HEW
memoranda 21 2 -has implemented meaningful bilingual programs for
limited and non-English-speaking students at the elementary grades,
but has not provided adequate bilingual programs for the first group of
limited and non-English-speaking students scheduled to graduate sub-
ject to the new competency test requirement. Hispanic students who
cannot pass the test might argue that denial of the diploma is illegal
because it simply reflects unequal educational opportunities provided
by the school. Like many black students, Hispanic students in such a
situation would probably argue that denying them a diploma is in
21 3
effect "blaming the victims.
In addition to the protection provided by Title VI and the Constitu-
tion, there are separate legal standards that specifically prohibit lin-
guistic discrimination in educational programs. For example, the
Equal Educational Opportunity Act of 1974214 provides: "No State
shall deny equal educational opportunity to an individual on account
of his or her race, color, sex, or national origin, by . . . the failure by
an educational agency to take appropriate action to overcome language
barriers that impede equal participation by its students in its instruc-
215
tional programs.
The point here is not that minorities should have to meet lesser
standards of functional competence than whites. In fact, blacks and
Hispanics, who in disproportionate numbers have been enrolled in
schools that do not emphasize basic skills, are in the forefront of those
calling for increased attention to basic skills. But care must be taken to
ensure that competency testing does not lead simply to another form of
racial and linguistic discrimination.

211. See U.S. Dep't of Health, Education and Welfare, Identification of Discrimination and
Denial of Services on the Basis of National Origin (1970) (memorandum by J. Pottinger, Director,
Office for Civil Rights).
212. See id.
213. To meet this concern in part, the Massachusetts Department of Education's Proposed
Regulations for Basic Skills Improvement requires exemption of all students "identified, screened
and enrolled in transitional bilingual programs pursuant to General Laws, Chapter 71A."
Proposed Regulations, supra note 29, § 40.10(4).
214. 20 U.S.C. §§ 1701-1758 (1976).
215. Id. § 1703(0 (1976). This statute also specifically provides for a civil remedy in federal
district court. Id. § 1706.
 FORDHAM LAW REVIEW (Vol. 47

C. CulturalBias
The disproportionate racial and linguistic effect of most competency
tests may result in part from cultural bias in test content. Cultural bias
is a concept subject to many definitions and interpretations. Some
would argue that a test is culturally biased if the results are used
unfairly, while others would distinguish between cultural bias in test
content and unfair use of the test. Some would define bias to encom-
pass any situation where there is a systematic difference in test scores
among groups, whereas others would limit the term to situations where
there are items within a test that are especially more difficult for one
group than another relative to other items. These are only a few of
many possible definitions. 2 16 When the complex emotional and politi-
cal aspects of any definition are combined with the technical measure-
ment problems, it is unsurprising that no single definition of cultural
bias has won widespread acceptance.
Defining cultural bias in the context of competency and other forms
of achievement tests poses special problems. What is determined to be
cultural bias in an Intelligence Quotient (IQ) or aptitude test may not
be cultural bias in a competency or achievement test because the tests
purport to measure different things. IQ tests purport to measure innate
intelligence, and aptitude tests purport to predict an individual's
potential for future achievement-and thus these tests theoretically
should not be influenced by instruction.2 17 Competency and other
achievement tests, however, are designed to measure what skills and
knowledge an individual has learned in a particular setting, and
therefore should be sensitive to and reflective of instruction. Thus the
specific findings regarding cultural bias in legal challenges to IQ
testing 2 18 are not likely to resolve similar questions about bias in
competency testing, although the general legal analysis should be
2 19
instructive.
216. For a helpful summary and overview of the various definitions and models to determine
bias, see Linn, In Search of FairSelection Procedures, 13 J. Educ. Measurement 53 (1976). Of
special note is Linn's discussion of criterion problems. See id. at 56-57. See generally APA
Standards, supra note 67, at 43-46.
217. Intelligence and competency testing purport to measure different things, but many critics
argue that they measure essentially the same thing-achievement, or some indistinguishable
combination of inherited intellectual capacity and achievement. See, e.g., J. Schwartz, The
Illogic of IQ Tests, in The Myth of Measurability 90 (P. Houts ed. 1977).
218. See Larry P. v. Riles, 343 F. Supp. 1306 (N.D. Cal. 1972), aff'd per curiam, 502 F.2d
963 (9th Cir. 1974). Plaintiffs in Larry P. were black children who argued that the stigmatizing
EMR (educable mentally retarded) label was applied to them on the basis of culturally biased IQ
tests, which penalized unfamiliarity with white middle-class background. The court concluded
that defendants had not sustained their burden of proving a rational relationship between scoring
on the questioned IQ test and the ability of black students to learn, and, therefore, awarded
preliminary injunctive relief against future testing. Id. at 1314. Although the trial on the merits
was concluded in the spring of 1978, no decision has been handed down as of this writing.
219. Of particular interest, for example, is the legal standard that will be applied by the
1979] COMPETENCY TESTING PROGRAMS

A "culture-fair" test may be an impossible goal. Henry C. Dyer of

ETS maintains that there are only two conditions under which a test
can be culture-fair: (1) either the learning required to perform accept-
ably on the test is commonly and equally available to all people of all
cultures, or (2) the stimulus material on the test is novel to all people of
all cultures. He contends that neither one of these conditions is
obtainable. 220 This sobering realization, however, should not hinder
the development of tests which are as bias-free and culture-fair as
possible for those situations where testing is required as part of an
educational program.
Since most competency tests purport to measure school achievement
and to be sensitive to instruction, such tests should have instructional
validity. 22 1 Serious questions of test bias would be raised, for example,
if the competency test measured knowledge and skills taught to white
but not to minority students. The more difficult questions, however,
are those raised by test items that are more difficult for minority
students, in spite of being offered the same instruction, because of their
different cultural backgrounds. The point is dramatically made by
student performance on tests which are designed to measure intelli-
gence and knowledge familiar to students from minority cultures. 2 2
Not surprisingly, minority students outperform most white middle-
class students who are at a disadvantage on these tests because of their
background.
One possible solution to this problem is to try to design a test that
reflects all aspects of our pluralistic society-or at least the extent of
diversity reflected by the student population. This would have the
effect of equalizing the test by placing students at an equal advantage
or disadvantage. This approach also has the advantage of being a
more accurate measure of the skills needed to function effectively in a
pluralistic society. A functional competency test given in Miami or San
Antonio, for example, should include a number of Hispanic skill and
content items, as some cross-cultural competence is arguably necessary
for successful functioning in those cities.
The technical difficulties involved in designing a fair pluralistic test
lead some persons to favor and the goal of "culturally neutral" tests. The
problem, of course, is that a culturally neutral test is a more evasive
goal than a pluralistic test. Tests usually perceived as culturally

Larry P. court in light of the Supreme Court's decision in Univ. of Cal. Regents v. Bakke, 438
U.S. 265 (1978).
220. H. Dyer, Race and Intelligence: An Examination of the Scientific Evidence by Four
Authorities (1963).
221. See pt. IV(A) supra.
222. An example of this kind of test is The Black Intelligence Test of Cultural Homogeneity
(BlTCH)-100-A Cultural Specific Test, designed by R. L. Williams and available from Williams
& Associates, Inc., 6374 Delmar Blvd., St. Louis, Mo. 63130.
 FORDHAM LAW REVIEW [Vol. 47

neutral in our society are almost inevitably tests that favor the pre-
dominantly white middle-class culture, in other words, the majority
culture. 223 Another argument against a pluralistic test is the practical
one that success is defined by the majority culture, and all students
regardless of background should be provided with the education and
training necessary to function effectively in our predominantly white
middle-class culture. Standard English and other values of the ma-
jority culture are the ultimate indicators of success, according to this
argument, and schools should help all students to achieve this goal.
This practical argument is voiced by many minority parents and
students.
If majority culture values are assumed, then the question of fairness
remains, given the inherent advantage that majority culture students
have over minority students even on instructionally valid tests. One
approach would be to minimize minority disadvantage on these tests
by providing additional, special instruction in majority-culture aspects
of the test that are especially difficult for minority students. Bilingual
instruction for limited and non-English speaking students is the most
obvious example, and comparable programs for other minority stu-
dents could be developed.
Pretesting can help to determine which aspects of a competency test
are especially difficult for minority students. Several statistical ap-
proaches have been developed to determine which items in a test are
especially more difficult for black and Hispanic students. Noting the
complex and somewhat uncertain nature of these statistical ap-
proaches, the California State Department of Education suggests a
simplified statistical method to identify especially difficult items.2 24 For
example, this method could isolate (1) test items which reflect concepts
either nonexistent within or unique to one culture, (2) items which
include words that have different meanings for different groups, and
(3) items with content that reflect experiences more familiar to one

223. The Texas Study, supra note 10, is culturally biased almost by definition, since the
University of Texas researchers defined incompetent adults as those whose level of mastery of
competency objectives is associated with "(ilnadequate income of poverty level or less,"
"[i]nadequate education of eight years of school or fewer," and "[u]nemployment or low Job
status." Id. at 5. This definition, for example, would include mothers receiving welfare from the
Aid to Families with Dependent Children Program, regardless of their actual level of literacy or
functional competence. The researchers admit that functional competency is "a construct which Is
meaningful only in a specific societal context.. . . [It] is culture-bound." Id. at 2. The APL test
thus does not measure an individual's competence in functioning in that part of society in which
he or she lives every day, but instead attempts to measure a person's competency by the test
designer's conceptions of what is required for successful functioning in middle-class America. The
ability to survive in a ghetto, for example, is not measured by the test, and, therefore, the test
results exaggerate the extent of functional incompetence among blacks and other minorities.
224. See California State Dep't of Education, Technical Assistance Guide for Proficiency
Assessment app. M (1977) [hereinafter cited as Technical Assistance Guide]. Compare California's
approach with the regression and other models discussed in Linn, supra note 216.
19791 COMPETENCY TESTING PROGRAMS

group than another. The Department defines these items as biased,

and recommends either revising or deleting the items. If all or most
.items measuring a particular skill prove especially difficult, another
approach would be to provide special instruction in these skill areas.
Any of these approaches-revision, deletion or special instruc-
tion-may resolve the problem, but in some cases it will not. Special
instruction may not be sufficient in some cases to overcome minority
disadvantage in these areas. Revision or deletion may not be practical
if minority students score uniformly lower on most or all items on the
test.
At this point some would seek to avoid responsibility with a genetic
explanation. 22 5 Schools would probably contend that the differential
test results are not caused by inadequate educational programs, but by
22 6
the students' poor attitude and/or low socio-economic status. Stu-
dent attitude is certainly a factor in test performance, but the school
may share some responsibility for poor attitude,2 2 7 and the causal
relation and interaction between attitude and school is probably as
complex as that inherent in the nature-nurture controversy. The most
common explanation for poor test performance is low socio-economic
status. One report observes: "Undoubtedly factors associated with low
socio-economic status contribute to the low test performance of chil-
dren from these backgrounds. However, very seldom if ever are these
socially related factors taken into account
'22 8
by those who develop and
generate mass educational programs.
The ultimate question raised by any discussion of cultural bias in
competency testing is the extent of the school's responsibility, if any, to
overcome the disadvantage that many children bring with them when
they first enroll in the public school system. This disadvantage is
usually the effect of low socio-economic status and what Justice Powell
in the Bakke case more generally called "societal discrimination". 2 9

225. See generally The IQ Controversy (N. Block & G. Dworkin eds. 1976) (reprinting
articles by Jensen, Hernstein, Jencks, Kamin, and others on whether intelligence has a substan-
tial genetic component).
226. See, e.g., Larry P. v. Riles, 343 F. Supp. 1306, 1310-11 (N.D. Cal. 1972), aff'd per
curiam, 502 F.2d 963 (9th Cir. 1974) (defendant's argument that the disproportionate number of
black children in ElMR classes simply reflects poor infant care, nutrition, and other environmen-
tal factors).
227. For example, teacher expectation can affect student performance and create a "self-
fulfilling prophecy." For a dramatic illustration of this point, see R. Rosenthal & L. Jacobsen,
Pygmalion in the Classroom (1968). Teachers were informed that tests indicated that certain
pupils would do well and others poorly, this in fact proved to be the case even though the pupils
had been chosen at random. For some of the literature stimulated by the Rosenthal and Jacobsen
study, see J. Elasoff & R Snow, Pygmalion Reconsidered (1971); Rist, Student Social Class and
Teachers Expectations: The Setf-FufiUing Prophecy in Ghetto Education, 40 Ha.'. Educ. Rev.
411 (1970).
228. Impact, supra note 25, at 35.
229. 438 U.S. at 307-10.
 FORDHAM LAW REVIEW [Vol. 4 7

Equal educational outcomes, or at least a lack of culturally dispropor-

tionate outcomes,2 3 ° may seem utopian to many, but that goal does not
seem visionary with respect to the minimum skills and knowledge
measured by most competency tests.
In any case, it seems clear that competency tests which are culturally
biased against racial and linguistic minorities will be subject to legal
challenge. 23 1 The unresolved issue is defining "cultural bias". Will the
term be limited to obvious forms of bias such as items containing
offensive terms and stereotypes? Will it extend to societal discrimina-
tion? Or will the line be drawn somewhere in between?

VI. HANDICAPPED STUDENTS

Many of the issues discussed above regarding test content, phase-in
periods, instructional validity, and racial and linguistic discrimination
apply to handicapped as well as non-handicapped students, but special
concerns are raised by competency testing programs for handicapped
students .232
A. Exemption

Some policymakers may favor total exemption of handicapped stu-

dents from competency testing programs because of the special diffi-
culty of designing fair assessment procedures for handicapped stu-
dents, and also because of the emphasis on Individualized Education
Programs (JEP) and the annual evaluations of the IEP required by the
federal Education for All Handicapped Children Act of 1975.233 If
handicapped students are exempted from competency testing pro-
grams, they should be given the option of taking the test. Failure to
provide such an option may deny such students the benefits of an
educational program in violation of section 504 of the Rehabilitation
Act of 1973.234 If policymakers decide not to exempt handicapped

230. Equal educational outcomes imply that all students reach a certain educational level.
Lack of culturally disproportionate outcomes is a less ambitious goal because it would allow for
some students to fall below the specified educational level as long as proportionate numbers of
students of all cultures and races were affected. The "effect standard," see note 203 supra and
accompanying text, uses disproportionate outcomes as the trigger for legal analysis.
231. See Morgan v. Kerrigan, 401 F. Supp. 216, 234 (D. Mass. 1975) (nondiscriminatory
curriculum ordered as part of school desegregation plans), aff'd, 530 F.2d 401, cert. denied, 426
U.S. 935 (1976); United States v. Texas, 330 F. Supp. 235, 249 (E.D. Tex.) (same), modified, 447
F.2d 441 (5th Cir. 1971). Under Title VI, the curriculum as well as other parts of the school
program should not be designed in such a way as to exclude minority children from the benefits of
participating in federally funded educational programs. 42 U.S.C. § 2000d (1976).
232. This discussion is based upon McClung & Pullin, supra note 5.
233. 20 U.S.C. §§ 1401-1461 (1976). The statute's implementing regulations are codified at 45
C.F.R. § 121a.1-.754 (1978).
234. 29 U.S.C. § 794 (1976). The statute's implementing regulations are codified at 45 C.F.R.
§§ 80.1-13, 81.1-.131, 84.1-.99 (1978).
1979] COMPETENCY TESTING PROGRAMS

children from the competency testing program, the), should consider

the issues summarized in this Part before deciding upon the exact
nature of the requirement.
B. DifferentialDiplomas and Standards
In assessing the fairness and legality of competency testing programs
for handicapped students, the two most fundamental questions are (1)
whether there should be differential diplomas for handicapped and
nonhandicapped students, and (2) whether there should be differential
standards for handicapped and nonhandicapped students.
A differential diplonma for handicapped students may be defined as a
diploma that is distinguishable in any way from that awarded to
nonhandicapped students who pass the competency test, for example,
in the color, shape, or wording of the diploma. Differential standards
for handicapped students may be defined as standards that are differ-
ent from (usually less stringent than) the standards that nonhandi-
capped students are required to meet. The basic rationale for differential
standards is that they are designed to meet the special needs and
capabilities of handicapped students.
The three general approaches likely to be considered by policymak-
ers are (1) same diploma and same standards, (2) same diploma and
differential standards, and (3) differential diploma and differential
standards. These three approaches could be applied with respect to all
handicapped students, or handicapped students on an individual basis.
For reasons discussed in Part VI(D) and (E), individualized determi-
nations are recommended.
C. DifferentialAssessment Procedures
Another important decision for policy-makers is whether differential
assessment procedures should be adopted for handicapped students.
This could take the form of modifying the paper-and-pencil test given
to regular students, and/or developing
235
methods of assessment that do
not require a paper-and-pencil test.
The modified paper-and-pencil test is used in Florida where blind
and visually impaired students may take Braille or large-print versions
of the statewide test given to nonhandicapped students. Modifications
of the test instrument, however, are not made for children with other
types of handicaps, and for some handicapped children this will raise
serious discrimination and equal protection questions under state and
236
federal law.

235. Direct or simulated performance measures are among the possibilities. See notes 263-64
infra and accompanying text.
236. Various bills to amend the Florida statute to account for these problems have been
introduced in the Florida legislature. No amendment has been enacted, however, as of this
writing.
 FORDHAM LAW REVIEW [Vol. 47

Similar legal questions would also be raised for a state that makes no
special provisions of this kind for handicapped students, and yet denies
diplomas to handicapped students who do not pass the test. The
argument would probably be that an unmodified test instrument
discriminates against handicapped students, especially those children
with sensory or motor problems, on the basis of their handicapped
conditions. 237 This argument will be stronger where the modifications
are relatively easy to make.
The use of alternatives to a paper-and-pencil test is illustrated by the
California statute which provides that local districts may adopt differ-
ential assessment procedures for handicapped students. 238 The Cali-
fornia State Department of Education emphasizes "that assessment of
student23performance
9
may be based on multiple criteria, not just a test
score."

D. Other Legal Implications

Although the federal Constitution, statutes, and regulations do not
provide clear guidance as of this writing, legal considerations may
preclude any uniform approach applied indiscriminately to all hand-
icapped students. What is fair for one handicapped student may be
unfair for another because individual circumstances vary so greatly.
Some handicapped students (for instance, a student whose only handi-
cap is a speech impairment), who need to be treated like nonhandi-
capped students but are not, will probably claim violations of the equal
protection clause, section 504, and the federal statutory presumption
that handicapped students should be integrated into the regular edu-
cational program to the maximum extent appropriate.2 40 Other hand-
icapped students (for example, many, but not all, multi-handicapped
persons), who need to be treated differently from nonhandicapped
students but are not provided with differential standard and/or as-
sessment procedures, will probably claim violations of the equal pro-
tection clause, section 504, and the federal statutory presumption that

237. See generally 45 C.F.R. §§ 84.35(b)(3), 121a.532(c) (1978) (regarding test Instruments
used for placement purposes for handicapped persons).
238. Cal. Educ. Code § 51215 (West Supp. 1979) provides: "Differential standards and
assessment procedures may be adopted for pupils who: (a) Are enrolled in special education
programs ...and (b) Have diagnosed learning handicap:; or disabilities that would preclude them
from attaining the district's regular proficiency standards with appropriate educational services
and support.
"Any such differential standards shall be included in the individualized education program
developed for the pupil ...." For an explanation of this law and discussion of various Issues
involved in setting differential standards, see Technical Assistance Guide, supra note 224, app. L.
239. Technical Assistance Guide, supra note 224, at II-15.
240. See 20 U.S.C. § 1412(5) (1976); 45 C.F.R. § 84.34(a) (1977).
1979] COMPETENCY TESTING PROGRAMS

to provide equal treatment to persons with unequal needs is unfair in

24 1
some circumstances.

E. Individualized Determinations
Legal and policy considerations suggest that decisions about the
nature and extent of participation of handicapped students in any
competency testing program should be made on an individual basis.
For some handicapped children the fairest approach would be to use
the same test, standards, and assessment procedures as those used for
nonhandicapped students. For other handicapped children, the fairest
approach would entail minor modifications. For many children with
severe handicaps, the fairest approach would be completely different
tests, standards, and assessment procedures, or even total exemption
from the competency test requirement.
Individual decisions of this kind could be made a part of the process
of developing the IEP mentioned in Part VI(A). Policymakers in some
states, including Missouri, 24 2 Massachusetts, 243 and California, 24 are
recommending or requiring individual decisions about competency
programs for handicapped children based upon the IEP. In California
the State Department of Education recommends, among other things,
that student proficiency standards for handicapped students be set
individually rather than for the group as a whole; that the committee
with the responsibility to develop an individual's plan should describe
the performance standards in basic skills for which the student may
receive a diploma; and that the committee "should prescribe attainable
' 245
standards which enhance learning.

VII. RECOMMENDED APPROACHES

The preceding parts of this Article have discussed a number of legal

and educational implications of requiring students to pass a compe-
tency test as a condition to receiving a high school diploma. For some
policymakers, the problems inherent in trying to develop a fair and
legal competency testing program will persuade them to forego an

241. In comments to the final regulations under § 504 of the Rehabilitation Act, 29 U.S.C. §
794 (1976), HEW notes that "different or special treatment of handicapped persons, because of
their handicaps, may be necessary in a number of contexts in order to ensure equal opportunity."
42 Fed. Reg. 22676 (1977). The regulations require providing aid, benefits, or services when
necessary to afford the handicapped effective opportunities. 45 C.F.R. § 84.4(b)(iv)(1978).
242. See Minimum Competency Testing and The Handicapped:Missouri'sPlan, 59 Phi Delta
Kappan 367 (1978).
243. See Proposed Regulations, supra note 29, § 40.10(2).
244. See note 239 supra.
245. Technical Assistance Guide, supra note 224, at 11-15.
 FORDHAM LAW REVIEW [Vol. 47

approach based upon diploma denial, and to explore some nonpuni-

tive, educational alternatives.2 46 Other policymakers, however, while
being concerned about the legal and educational problems, may never-
theless decide to proceed with a diploma-based competency testing
program. With these policymakers in mind, this Article will suggest
some approaches to competency testing that are designed to minimize
the educational and legal problems to the extent possible.

A. Community Participation
The decision to implement a competency testing program raises
several questions. Should the testing program, for example, be de-
signed to measure only the basic proficiency skills such as reading,
writing, and computation, or should it go beyond this by measuring a
student's ability to apply these skills in adult-life role activities such as
those of a consumer, producer, and citizen? Should satisfactory per-
formance on the test, whatever it measures, be a minimum standard to
be used in conjunction with other criteria, or should it be the exclusive
criterion of satisfactory performance resulting in a high school diploma
regardless of age or course credits?
The author's opinion is that single-criterion evaluation of students,
teachers, and public education sells public education short. For one
thing, the state of the art is not sufficiently developed to warrant such
exclusive reliance on competency-based evaluations. More importantly,
the primary goals of public education are or should be broader than
those reflected by minimal competencies, and students, teachers, and
public education generally should not be evaluated exclusively by these
narrow measures. Even given a strong commitment to a broader view
of public education, there is always the danger that the minimal
standards will become maximums rather than minimums. As in other
areas such as drivers' licenses and bar examinations, however, the
need for establishing minimum standards has been demonstrated, and
safeguards can be developed to ensure that broader goals are not
bypassed.
These issues are subject to considerable difference of opinion, as
illustrated by early exit programs that allow students, regardless of
age, to graduate from high school upon passing a minimum compe-
tency test. Competency testing programs obviously involve implicit or
explicit decisions about performance objectives and educational goals,

246. For example, the National Academy Report, after criticizing competency testing as a
prerequisite to a high school diploma, states: "However, the Panel is in agreement that a series of
standardized tests at the lower grade levels used for diagnosing individual student weaknesses,
pinpointing remediation needs, and building public pressures if school-wide performances in basic
skills continue over time to be consistently low, could be positive influences on student learning"
National Academy Report, supra note 3, at 9.
1979] COMPETENCY TESTING PROGRAMS

and these in turn have important implications not only for curriculum
and instruction but also for other school practices such as grouping and
discipline. 24 7 Given the crucial importance of these decisions, a model
program should provide for representative 24 8
community-based partici-
pation in the decisionmaking process.
There may also be a legal rationale for community involvement
since the standard-setting process is essentially an arbitrary and
judgmental one. Criterion-referenced or objective-referenced measures
may appear objective and scientific, but in fact they are based on
subjective judgments and subject to the charge of arbitrariness, as
illustrated by Glass's critique 24 9 of the Florida competency test2 50 and
also by 251
the corresponding claim by plaintiffs in Debra P. v. Tur-
lington. Determining the cut-off score for a norm-referenced test is
also, of course, somewhat arbitrary. This does not mean that the
arbitrariness is legally vulnerable arbitrariness. Courts are no strangers
to arbitrary standards; in fact, they set many themselves to aid
resolution of difficult issues. Although most courts would be reluctant
to conclude that a cut-off score on a competency test is illegally
arbitrary,2 5 2 the kind of arbitrariness inherent in both criterion-
referenced and norm-referenced tests may focus judicial attention on
the process of setting the educational goals, performance objectives,
and cut-off scores. A court might invalidate a test if the standards have
not been "legitimized" by community participation or some other
representative process.
Whether based on legal or political considerations, community-based
participation is required by many competency testing statutes.2-5 3 But
community participation in turn is likely to aggravate another

247. See, e.g., Spady, supra note 1, at 9-14; Spady & Mitchell, supra note 160, at 9-15.
248. Compare Shepard, Setting Standards and Living Wills Them, 18 Fla. J. Educ. Research 28,
28-32 (1976) (recommending that standard setting be an iterative process involving various
audiences).
249. Glass, supra note 27.
250. See As Always, supra note 84; notes 83-85 supra and accompanying text.
25 1. "The state defendants have arbitrarily determined what scores will be sufficient to pass
the test." Complaint, supra note 63, at T 46.
252. Plaintiffs, nevertheless, will probably argue that the cut-off score for classifying persons
as either competent or incompetent is so arbitrary as to be unconstitutional. Their general
argument is articulated by Judge Davis: "When a program talks about labeling someone as a
particular type and such a label could remain with him for the remainder of his life, the margin of
error must be almost nil." Merriken v. Cressman, 364 F. Supp. 913, 920 (E.D. Pa. 1973) They
will probably produce expert testimony questioning the educational and psychometric validity of
making important judgments about an individual on the basis of a single test
253. Cal. Educ. Code § 51215 (West Supp. 1979), for example, provides: "Standards of
proficiency shall be adopted by the governing board with the active involvement of parents
broadly reflective of the socioeconomic composition of the district, administrators, teachers,
counselors, and, with respect to standards in secondary schools, pupils-" See also Proposed
Regulations, supra note 29, § 40.02(2).
 FORDHAM LAW REVIEW [Vol. 47

problem-the kind of mismatch between test and instruction men-

tioned in Part IV. By including broad participation in the standard-
setting process, states and school districts are likely to end up including
some new or at least redefined educational goals and performance
objectives. Thus there is always a potential conflict between the
educational and perhaps legal need for some kind of match between
test and instruction and the community's political need to set stan-
dards that do not simply rubberstamp the educational status quo. In
order to resolve this conflict, states and/or school districts may want to
consider the kind of two-phase implementation procedure discussed in
the next section.

B. Two-Phase Implementation Procedure

In attempting to resolve the questions of legality and fairness raised
by this Article, the persons responsible for developing a competency
program may want to adopt a two-phase implementation procedure.
Phase I would emphasize that the competency standards for the first
classes of students affected by the new requirement should be formu-
lated on the basis of the curriculum and instruction to which these
sludents have already been exposed for most of their educational
careers in the district's classrooms. In other words, during phase I the
focus would not be on which proficiency skills the governing board and
community think should be taught and emphasized in the district's
schools but rather on those proficiency skills that have been taught and
emphasized previously by the district. Iii some schools there will be no
actual distinction between what should be emphasized and what has
been emphasized, but in other schools the distinction will be substan-
tial. The distinction will probably be reflected in the difference be-
tween basic literacy and numeracy skills emphasized by most schools
and the adult-life skills emphasized by most competency tests.
Thus, phase I standards are essentially past standards-those that
have already been emphasized in district classrooms. Students, par-
ents, and teachers should still be notified of phase I standards as soon
as they are adopted. Even though the standards will not have
changed, both the assessment procedures and the penalty for failure to
meet the standards, denial of a diploma, are new; therefore, advance
notice is essential. 254 Phase I standards should be retained until the
new curriculum and instruction, if any, have been implemented and
adequate notice is given to students, parents, and teachers.
New curriculum and new standards, if any, are the focus of phase
II. Given the severe nature of the penalty to be imposed on students
who cannot meet the district's phase I standards, a state or school
district might, as a policy matter, want to reevaluate its traditional
254. See notes 144-48 supra and accompanying text.
1979] COMPETENCY TESTING PROGRAMS

standards and develop new, more appropriate standards. The plan-

ning for phase II should begin simultaneously with the implementation
of phase I.
The suggested two-phase implementation procedure is based on a
number of legal and policy conclusions discussed in this Article.
Perhaps the most important of these conclusions are: (1) the need for
some kind of match between test and instruction (instructional valid-
ity), and (2) the importance of the distinction between measures of
basic literacy/numeracy skills and adult-life skills.
For those who agree with these conclusions and supporting rationale,
the two-phase implementation procedure will help to ensure fairness
for all students and will thereby reduce the risk of legal challenge. This
approach by itself, however, cannot resolve all concerns about fairness
and legality. The most serious of these concerns is the potential
discriminatory effect of competency testing programs on racial and
linguistic minorities.
If a state or school district decides to develop a second phase, it
might want to consider following the five minimum steps suggested in
Part VII(D)(1) as a basic framework for setting new graduation
requirements-starting with substantial community involvement and
concluding with a careful analysis of the psychometric and instruc-
tional validity of the assessment procedures. Some suggestions concern-
ing instructional validity and assessment follow.

C. Approaches to Instructional Validity and Assessment

The assessment procedures, whether developed as part of phase I or
phase II, should meet professional psychometric standards and have
instructional validity. This Section will describe some approaches to
instructional validity and consider direct performance measures as an
alternative or supplement to the traditional paper-and-pencil test.
In developing or selecting an assessment instrument, steps should be
taken to ensure the instructional validity of the instrument." s Al-
though content validity is a distinguishable concept,2 5 6 it is sufficiently
related to offer guidance in determining the instructional validity of a
255. The approaches discussed herein to assess the instructional validity of a competency test
reflect preliminary research and thoughts. Better approaches might be developed. For example,
Laura Wagner of the Stanford Graduate School of Education suggests that the instructional audit
developed by Professor Daniel Duke of Stanford University, or some adaptation of the instruc-
tional audit, might be useful in establishing the instructional validity of a competency test. See L.
Wagner, The Duke Instructional Audit: An Aid in Curricular Decision-Making (Nov. 22, 1978)
(paper presented to National Council for the Social Studies). Perhaps methods to determine
instructional validity, including the above approach, can be developed to generate the necessary
information without significantly increasing recordkeeping or infringing upon classroom instruc-
tion.
256. See note 155 supra and accompanying text.
 FORDHAM LAW REVIEW [Vol. 47

proposed instrument. One approach is suggested by ETS's study 2 7 of

the use of the National Teacher Examination (NTE) for certification
purposes in North Carolina. The study examined the content domain
of the NTE in relation to the content domain that should be measured
if scores are used for initial certification of teachers. The researchers
determined that the content domain that the test should measure was
the content of teacher education programs offered by North Carolina
colleges:
There are four sets of data that are measures of the correspondence between the
content of the test and the content of the teacher training program: (1) the percentage of
questions classified as content appropriate; (2) the number of content areas identified as
omitted from the test; (3) the judgments about relative emphasis on specified subject
matter in the test and in the teacher training program; and (4) the overall judgments
with respect to whether the test parallels the curriculum. Criteria were established
with respect to each of these sets of data to identify the degree of relationship between
the test content and the program content. These criteria were applied individually to
each test in the Common Examinations and each of2 5the Area Examinations, and the
interrelationships of criteria were also considered.
This kind of content validation of the NTE was also conducted by the
ETS with respect to teacher education programs in South Carolina, 59
and secured judicial acceptance in United States v. South Carolina.2
This approach could be adapted to measure instructional validity by
relating test content to instruction in specific schools rather than to
curriculum on a statewide basis. 2 60 Furthermore, a sound approach
would support judgments about instructional validity with evidence of
actual instruction.
Schools will have to play a key role in the development of the
assessment instrument even if, and perhaps especially if, they contract
with an independent test publisher. Popham has noted that compe-
tency testing poses problems for test publishers because the tests will
have to be based upon curriculum to a far greater extent than the
traditional standardized tests, thereby making the publishers' instru-
ments usable only by districts with very similar curricula. 2 6 1 This
problem will be especially acute in states where each district is
responsible for developing its own standards.
257. Educational Testing Service, Report on a Study of the Use of the National Teacher
Examinations by the State of North Carolina (1976).
258. Id. at 159.
259. United States v. South Carolina, 15 Fair Empl. Prac. Cas 1196 (D.S.C. 1977), aff'd
mem. sub nom. National Educ. Ass'n v. South Carolina, 434 U.S. 1026 (1978); see notes 168-71
supra and accompanying text.
260. Readers interested in this approach will want to refer to the complete report for details
on how these determinations were made and what degree of match between test and program was
considered sufficient. See Educational Testing Service, Report on a Study of the Use of the
National Teacher Examination by the State of North Carolina 5 (1976) (chart and related
explanation). The reference to the report herein is for illustrative purposes only and is not
intended as an endorsement of the approach described in the report.
261. Popham, Customized Criterion-Referenced Tests, 34 Educ. Leadership 258, 258-59
(1977).
1979] COMPETENCY TESTING PROGRAMS

One way in which test publishers are attempting to solve this

problem is by creating "item pools." Districts that have specifications
based on their performance objectives can match their specifications
with those for which items and exercises have already been prepared
by the test publisher. Whatever approach is used in test development
or test selection, the district will need to make important determina-
tions about whether the proposed instrument is sufficiently related to
curriculum and instruction. The following advice that Melville offers
school districts in selecting an achievement test is generally applicable
to competency tests:
[Tihe most important elements in good test selection [are]: How well does the test
measure what it was selected to measure? Are the things being measured important? Is
there a proper balance of areas covered? Are all the important areas to be taught
included? No one-salesman, neighboring school colleague, learned reviewer-can
answer these questions. Only the classroom teacher and his department colleagues can
supply the answers, and they can only do so by looking carefully at the actual test
questions one by one.
The task of making a careful examination of a test cannot be simplified very much.
It is a time-consuming job. It can be more effectively accomplished, however, if
approached systematically. One approach, applicable primarily to the selection of an
achievement test, follows.
The objective of the test analysis is to ascertain (1) what kind of materials are
included in the test and (2) the relevance of these materials to the teaching program. In
order to appraise test items in a meaningful way, one should ask of each item: What
particular skill, understanding, type of material, or subject matter does this item seem
to be covering? Since judgments regarding the relevance of an item are necessarily
subjective, a rough point-value scale for judging relevance should suffice. Items
considered inconsequential or trivial would be rated 0 to 1; those closely related to
local objectives and student capabilities would be rated 4 or 5. A work sheet ensures
that the analysis is done systematically. Four column headings are 26
needed; item
number, item content, skills involved, and relevance of the item.
The approach suggested by Melville could be adapted to make deter-
minations of test relevance to curriculum and instruction. The district
should ensure a sufficient match between curriculum and instruction
by eliminating tests or test items that are not sufficiently matched with
curriculum and instruction.
A sound approach to instructional validity would not rely exclu-
sively on the subjective judgments of a panel of administrators or
teachers about whether specific test items were taught in the district's
classrooms. These judgments should be supplemented with, and sup-
ported by, comparisons of test items and evidence of a district's
curriculum and instruction. Such comparisons would include evidence
such as curriculum guides, textbooks and materials used in the class-
room, lesson plans, student homework samples, and perhaps inter-

262. S. Melville, Selecting an Achievement Test (four-page summary of the script from a film
with the same title, available from the Cooperative Test Division, Educational Testing Service,
Princeton, N.J.).
 FORDHAM LAW REVIEW [Vol. 47

views with a representative group of students and teachers. A school

district concerned about the instructional validity of its assessment
procedure might want to arrange for an outside agency to examine
such evidence and prepare a report of its findings and conclusions prior
to final approval of the assessment procedure.
Madaus and Airasian write that many of the minimal cognitive
competencies for graduation involve application of basic literacy and
numeracy skills to real-life situations, for example, checking the accu-
racy of bills and sales slips, using the public library, using the town
and state offices. "These competencies are most validly measured by
the most direct means possible, situational or performance exam-
inations which determine if the student can actually perform the
behaviors. ' '26 3 Although they recognize that direct measurement is
often costly and time consuming, the two authors conclude that
"indirect paper-and-pencil tests, measuring knowledge about the com-
petency areas, are not enough. Any indirect, or surrogate, measure-
ment must be validated against direct performance measures. '264 Serious
questions about fairness and legality would be raised if a student could
show that he or she was denied a diploma on the basis of performance
skills that he or she could demonstrate by direct assessment but not by
the indirect paper-and-pencil method. This problem also suggests that
if a student cannot pass a paper-and-pencil competency test, perhaps
that student should be given a direct performance test to be sure that
he or she does not have the requisite skills before denying him or her a
diploma.
Field testing of the proposed test can serve several important
purposes, including determination of its likely effects on blacks, His-
panics, and the student population as a whole. If substantial or
disproportionate numbers of students in any category cannot pass the
proposed test, the state or school district may want to take another
look at the adequacy of the match between the test and instruction in
its classrooms and at the reasonableness of its proposed cut-off score.
Where limited and non-English-speaking students have difficulty with
the field test, the test could be given in the student's native language in
order to determine whether the student's difficulty on particular test
items is with language or the skills measured by the items. Such
assessments can be used to help provide low-scoring students with
appropriate remedial instruction before the final assessment.

D. Proposalsfor Competency Testing Programs

1. Basic Framework
Before new graduation standards are used as a basis for denying any
student a diploma, a sound and careful competency testing program
263. Madaus & Airasian, supra note 15, at 86.
264. Id.
1979] COMPETENCY TESTING PROGRAMS

would at a minimum include the following as its basic framework:

1. Provide for substantial parental and other community involve-
ment in developing and adopting educational goals, performance ob-
jectives, and assessment procedures.
2. Provide sufficiently detailed advance notice of the new educa-
tional goals, performance objectives, and assessment procedures to
students, parents, and teachers.
3. Make necessary changes in curriculum to reflect new educational
goals and performance objectives.
4. Provide for a sufficient phase-in period of the new curriculum
and instruction. Sufficient phase-in may take years, depending upon
the degree of curricular/instructional change.
5. Take steps to ensure that the assessment procedure meets profes-
sional psychometric standards and has instructional validity, as de-
fined in Part IV(A).

2. Model Program Provisions

Policymakers may also want to consider the following model pro-
gram provisions in attempting to design a fair and equitable program.
A committee or committees of representative community, school,
and professional persons should be formed to review the proposed test
in order to eliminate any items with inappropriate content, including
(1) coerced belief, (2) invasion of privacy, (3) unteachable or un-
measurable content, (4) content that is not sufficiently matched with a
school's curriculum and instruction, and (5) content biased on grounds
of race, language, culture, national origin, sex, handicap, etc. Particu-
lar emphasis should be given to legal representation for a review of the
first two categories, to psychometric and classroom-teacher representa-
tion for the last category.
The proposed assessment procedure should be pretested, and the test
results used to help determine (1) a reasonable cut-off score or proce-
dure, and (2) whether the test may be biased against any racial,
linguistic, or other group. A disproportionate failure rate on the test
generally or on specific items for any particular group is an indication
of possible bias. Statistical approaches can be used2 6to
5
determine which
items are especially difficult for various groups.
A model program should provide for multiple learning, evaluation,
and remedial opportunities. Students who will have difficulty passing
the test, as identified by the pretesting, should be given sufficient
instruction and/or remedial instruction to help them pass the test.
Provision should be made to insure that students who do not pass the
test are not tracked in all courses just because remedial instruction is

265. See note 224 supra and accompanying text. But see Linn, supra note 216 (limitations of
statistical approaches).
 FORDHAM LAW REVIEW [Vol. 47

necessary, especially if remedial classes would involve racial segrega-

tion or resegregation of students.
Where adequate education and/or remediation for some students is
not possible in the time available before the final test, such students
should not be required to pass the test as a prerequisite to receiving a
regular high school diploma. Exemption of black students from this
requirement may be appropriate, for example, where the school dis-
trict is under a court desegregation order or agreement, or where there
has been some other finding of prior discrimination by the school
against these students. Another example would be school districts that
have initiated, but have not had time to implement adequately biling-
ual programs for limited and non-English speaking students.
A model program should provide for differential standards and/or
assessment procedures for handicapped students. These standards and
procedures should be determined on an individual basis by the team
that formulates an individualized educational program for the student
pursuant to the Education for All Handicapped Children Act of
1975.266
A low score on a competency test should be used as the trigger for a
more careful evaluation of the student's performance rather than as the
final determination of failure. The program should utilize a committee
to determine by other methods whether the student possesses the
requisite knowledge and skills, especially if the student has achieved a
near-pass score. 267 Simulated or direct performance measures could be
used where appropriate to make sure that the indirect paper-and-
pencil measure is accurate before denying any student a diploma.
Former students who failed the test should be given an option of
earning the high school diploma by further remedial education and test
opportunities at any later point in their lives.
Since learning is a two-way street, a model program should provide
for students and educators to share responsibility for performance
rather than place the full burden on the students. For example, if
passing a competency test is one criterion for a student's earning a high
school diploma, perhaps a teacher's success in helping students pass
the competency test should be one criterion in the various forms of
teacher evaluation, for example, tenure decisions. Such evaluation is
feasible if students are tested at the beginning and end of each year in
each course, and allowance is made for mitigating circumstances such
as student absences and transfers.
266. 20 U.S.C. §§ 1401-1461 (1976); see notes 240-45 supra and accompanying text.
267. Instead of using a single cut-off score to determine minimally acceptable performance on
a competency test, some educators suggest a cut-off "band," whereby students with near-pass
scores within the band are evaluated by other means before making a final determination about
performance.
1979] COMPETENCY TESTING PROGRAMS

Finally, administrators should be responsible for making the neces-

sary changes in school practices, and the public should be responsible
for financing such changes. If these corresponding responsibilities for
teachers, administrators, and the public seem unreasonable or unfair,
then perhaps the reasonableness and fairness of denying diplomas to
students based upon the results of a competency test should be
reconsidered. Developing a workable model of shared responsibility %ill
obviously be more difficult than placing the entire burden on students,
but it will also be fairer.

CONCLUSION

This Article has identified a number of legal and educational prob-

lems raised by most programs that make passing a competency test a
prerequisite to receiving a regular high school diploma. The potential
of these programs for unfairness, combined with the severe nature of
the penalty-which could in effect relegate a student to second-class
citizenship-makes legal challenge likely and special care by state
legislatures and school districts essential.
Since litigation is costly and difficult for all parties, states and school
districts that are committed or inclined to a diploma-based competency
program may want to seek independent legal opinion about the legality
of their proposed program before it ends up in court. In the absence of
more specific materials, this Article can be used to anticipate plantiffs'
arguments. The next step is to evaluate the strength of these argu-
ments in light of the exact competency testing program proposed, and
the exact factual situation presented by the state or school district. If
independent legal opinion indicates serious problems, the state or
school district may want to modify the proposed program to meet the
legal concerns. States and school districts that are not already commit-
ted to a diploma-based competency program may want to consider
alternative approaches that raise fewer legal and educational prob-
lems.
The ultimate arbiters of the legality of competency testing programs
are, of course, the courts. As various states and districts begin to
withhold diplomas on the basis of competency tests, and legal chal-
lenges are heard by the courts, a judicial pattern and predictable case
law on competency testing will eventually take shape. This case law
should be read with care as it will set the basic parameters of
educational policymaking.
It should be emphasized, however, that whatever the shape of these
legal parameters, they will reflect only the minimum standards essen-
tial to fairness under our legal system. Policymakers must meet, but
 FORDHAM LAW REVIEW

are not limited to, these minimum standards in pursuing the goal of
educational equity for students. Some of the legal arguments, even if
unsuccessful in the courts, suggest standards of fair play. For example,
even if the eventual case law on competency testing does not require
some kind of match between what is measured by the test and what
has been taught in the school, most persons would probably agree that
this is an essential component of any fair competency testing program
where high school diplomas are at stake. Similarly, prior notice to
students and gradual phase-in periods will probably strike most per-
sons as important even if the courts decide that they are not legally
required.
The discussion set forth at the outset of this Article indicates
substantial disagreement about the advisability of competency testing
as a prerequisite to a high school diploma. Whatever one's personal
views about the competency testing movement, most readers would
probably agree that, except in unusual cases, functional illiteracy after
twelve years of public education is simply unacceptable. Perhaps this
concern can be mobilized to develop more equitable programs and to
provide an adequate education for all children.