ARTICLE IN PRESS

Teaching and Teacher Education 23 (2007) 999–1011

www.elsevier.com/locate/tate

Ethics in classroom assessment practices: Issues and attitudes

Susan K. Greena,, Robert L. Johnsonb, Do-Hong Kimc, Nakia S. Popea
a
Winthrop University, Rock Hill, SC 29733, USA
b
University of South Carolina, SC, USA
c
University of North Carolina at Charlotte
Received 24 January 2006; received in revised form 5 April 2006; accepted 10 April 2006

Abstract

Student evaluations should ‘‘be ethical, fair, useful, feasible, and accurate’’ [JCSEE (2003). The student evaluation
standards. Arlen Gullickson, Chair. Thousand Oaks, CA: Corwin]. This study focuses on defining ethical behavior and
examining educators’ ethical judgments in relation to assessment. It describes the results from a web-based survey of
educators in which they read a brief scenario and indicated whether the student evaluation practice in the depiction was
ethical or unethical. Results showed strong agreement among the educators on fewer than half of the scenarios presented in
this study. These findings suggest that assessment is currently an educational realm without professional consensus.
r 2006 Elsevier Ltd. All rights reserved.

Keywords: Evaluation methods; Ethics; Classroom assessment

1. Introduction the students were caught cheating on a class project.

The decision split the community and the teacher
The importance of the evaluation of students is resigned (Carroll, 2002). In another instance, the
evidenced in the recent publication of The Student state law enforcement division in a southern USA
Evaluation Standards (The Joint Committee on community investigated a teacher after she used in
Standards for Educational Evaluation [JCSEE, her classroom some commercially available test
2003]). The standards are offered as ‘‘principles preparation materials that are designed to simulate
that should guide and govern student evaluations’’ the type of items and subject matter in the state test.
(JCSEE, p. xx) and ‘‘require that student evalua- Finally, the president of a southeastern USA college
tions be ethical, fair, useful, feasible, and accurate’’ recently fired two professors for their refusal to
(JCSEE, p. 3). adhere to a policy awarding freshman 60% of their
The need for guidelines on ethical assessment grade based on effort (Click, 2004). The dismissal of
practices is evident in frequent incidents taken from the professors brought unfavorable national atten-
newspaper headlines. In one incident, a biology tion to the administrator’s misguided policy on
teacher in the Midwest USA decided to assign effort and grades.
students failing grades for the science course after Guidelines for assessment practices include those
offered in The Student Evaluation Standards
Corresponding author. (JCSEE, 2003), the Standards for Teacher Compe-
E-mail address: [email protected] (S.K. Green). tence in the Educational Assessment of Students

0742-051X/$ - see front matter r 2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tate.2006.04.042
 ARTICLE IN PRESS
1000 S.K. Green et al. / Teaching and Teacher Education 23 (2007) 999–1011

(American Federation of Teachers, National Coun- such practices may be considered professionally
cil on Measurement in Education, & National unethical (p. 82).’’ While Popham does offer this
Education Association, 1990), the Standards for guideline, he neglects to place it within a larger
Educational and Psychological Testing (American ethical framework. Why, for example, does the
Educational Research Association, American Psy- ethicality of test preparation depend on what other
chological Association, & National Council on people think? Making ethical assessment practices
Measurement in Education, 1999), and the Code contingent on whether or not one may get caught
of Ethical Conduct and Statement of Commitment results in a simplistic utilitarian calculus, which is
of the National Association for the Education of not, in and of itself, ethical.
Young Children (2005). As Gipps (1994) points out, Some textbooks do briefly discuss ethical princi-
however, guidelines are often ‘‘general and almost ples that can guide ethical judgments related to
superficial’’ (p. 144). They provide rules without an assessment. For example, Airasian (2005) suggests
overarching ethical framework that can provide that the ethical standards for assessment refer to
support and guidance for exercising judgment in ‘‘some aspect of a teacher’s fairness in dealing with
specific situations. his or her pupils (p. 20).’’ Similarly, Taylor and
Several authors have provided ethical guidelines Nolen (2005) point out that because poor assess-
related to preparation for high stakes testing (e.g., ment can significantly affect students, ‘‘the ethical
Haladyna, Nolen, & Haas, 1991; Mehrens & responsibility of educators is ‘first, Do No Harm’
Kaminski, 1989; Popham, 1991). In addition, (p. 7).’’ Payne (2003) sums up ethical codes by
several studies have examined in-service and pre- stating that educators should ‘‘Assess As Ye Would
service teacher attitudes regarding the appropriate- Be Assessed.’’ These principles are not elaborated as
ness of standardized test preparation strategies (e.g., tools to assist educators in making day-to-day
Bright, 1992; Kher-Durlabhju & Lacina-Gifford, judgments, however. To summarize, the discussion
1992; Moore, 1993). However, ethical standards in the literature often describes what various
related to testing are not consistently defined or authors consider ethical practices, but these discus-
agreed upon (Thorndike, Cunnningham, Thorn- sions suffer from a lack of explicit identification of
dike, & Hagen, 1991; and see Kilian [1992] in the underlying ethical assumptions related to mak-
response to Popham [1991]). Gipps and Murphy ing decisions about behavior in the realm of
(1994) go a step farther and suggest that more assessment.
attention should be focused instead on wider equity The present authors define ethical behavior as
issues such as equality of access to instruction. The acting based on one’s judgment of an obligation—a
lack of agreement highlights the need for an duty by virtue of a relationship with a person,
overarching ethical framework from which to persons, or social institution (Kant, 1785; Rawls,
develop the capacity to make judgments about 1971). What makes a behavior ethical or unethical is
ethical assessment practices. whether that behavior is consistent with or contra-
Other authors have discussed the importance of dictory to one’s obligations. Given these conditions,
ethics in assessment (e.g., Baumgart, 1996; Gipps, there is always a gap between principles and
1994; Popham, 2000). For example, Baumgart behavior requiring judgment in a specific situation
(1996) applauds the development of ethical codes (Hostetler, 1997; Kant, 1790; Strike & Soltis, 1998),
related to assessment. He argues that education, so uniform agreement on standards in every case
reflection, and self-regulation related to ethics are would not be expected. Educators must be given the
more likely with easy access to such codes of ethics, space, autonomy, and support to learn to use their
but does not suggest principles that might govern judgment. This approach thus shows an important
such activities. Gipps (1994) points out that the distinction between ethics and best practices in
powerful impact of assessment necessitates an teaching and assessment. Ethics is addressed in
ethical framework on which to draw, but limits terms of obligations based on relationships so there
the discussion to specific issues such as consequen- is an affective component as well as a rational one
tial validity (Messick, 1993), equity, and test (Noddings, 1984; Nussbaum, 1986, 1990). More
preparation. Popham (2000) does offer an ethical specifically, obligations are felt, but best practices
guideline—‘‘when teachers engage in test prepara- are merely followed.
tion practices which, if brought to the public’s Other literature suggests that teachers may not
attention, would discredit the education profession, always be well equipped to make ethical judgments
 ARTICLE IN PRESS
S.K. Green et al. / Teaching and Teacher Education 23 (2007) 999–1011 1001

related to assessment. Teachers’ knowledge base Student Evaluation Standards (p. 7). Several of those
about guidelines or ethical codes may be uncertain standards address aspects of this fundamental
because they often lack formal assessment training assumption: to serve the needs of students, to treat
or their training has become dated (Impara, Plake, students with respect, and to incorporate basic
& Fager, 1993; Plake & Impara, 1997; Stiggins, principles of fairness (Airasian, 2005). Popham’s
1999). Conflicting norms related to teaching (e.g., (1991) point that educators serve in loco parentis
institutional norms vs. ethical interpersonal norms) also lends support to the principle that assessment
often place teachers in ethical dilemmas (Colnerud, should protect students and Do No Harm to them.
1997). But, teachers’ ability to apply guidelines to a In the present conceptualization, the guiding
specific context has received limited attention in the principle Do No Harm was chosen to make the
literature. One exception is the Plake and Impara implications more concrete, focusing on the harm
(1997) study in which five items in a test of teacher that can be done when the basic premise is not
assessment competence measured teachers’ ability followed. Fairness (or protection of student rights)
to answer items related to ethical issues in student is a general principle that no one contests in the
evaluation. These items, however, all focused on abstract. However, thinking about causing harm
issues relating to standardized tests. focuses the discussion at the level of the implications
Another element that has not been thoroughly of everyday practice. Educators must be well versed
addressed in the literature is the application of in the potential impact of the practices they use
ethical principles to ongoing classroom assessment. because their assessment and evaluation may have a
In the classroom, teachers are required to weigh variety of unintended consequences for their stu-
their obligations and make judgments every day dents. For example, a teacher who uses surprise
related to assessment. Comprehensive principles items on a test that did not appear on the study
related specifically to classroom evaluation have guide may do harm by breaking the implicit bond of
not been widely promulgated, although several trust between teacher and student. A teacher who
assessment texts do discuss ethical codes and passes out tests from highest grade to lowest may do
practices related to classroom assessment (e.g., harm by breaching confidentiality. Such actions
Airasian, 2005; Taylor & Nolen, 2005). imply lack of respect for student rights and needs.
As a starting point for generating principles The second general guiding principle for class-
widely applicable to the various types of assessment room assessment is to Avoid Score Pollution. This is
that occur in classrooms, the present authors a specific iteration of the Do No Harm principle that
examined professional standards, articles addressing applies to assessment. This principle is adapted
ethical principles in education, papers related to from guidelines suggested by both Popham (1991)
ethical standardized test preparation practices, and and Haladyna et al. (1991) for ethical standardized
assessment texts. Two general guiding principles test preparation. Both authors suggest that any
seem to capture the essence of the ethical concerns practice that improves test performance without
in these documents. concurrently increasing actual mastery of the
The first is to Do No Harm (Taylor & Nolen, content tested produces score pollution. That is,
2005). This is a basic broad ethical principle that the score on the test does not represent actual
people in general use to govern their lives by virtue student achievement in the content area and is
of their being human. It is a variant of ethical ‘‘polluted’’ by factors unrelated to academic attain-
principles handed down through the centuries, like ment. If scores do not reflect mastery then harm has
the Golden Rule (or Assess as Ye Would Be been done. This situation is akin to lying. For
Assessed [Payne, 2003]). (This is a core ethical example, practicing beforehand with actual test
principle in other professions such as medicine content would produce score pollution. In essence,
[Jonsen, Siegler, & Winslade, 1998; Munson, 2000].) this is a validity issue. Test scores no longer measure
The judgment involved comes in defining what generalized mastery but simply ability to memorize
harm is, or choosing between different harms in the specific test items.
relationship between teacher and student. This The present authors believe this principle can be
principle stems from the basic premise that ethical extended to other elements of classroom assessment.
guidelines must protect the rights of individuals Any official assessment, including grades, should
affected by an evaluation. This premise is congruent reflect only the extent to which students have
with the proprietary standards in the JCSEE (2003) mastered the goals of instruction (e.g., Brookhart,
 ARTICLE IN PRESS
1002 S.K. Green et al. / Teaching and Teacher Education 23 (2007) 999–1011

2004; Smith, Smith, & DeLisi, 2001). When room assessment scenarios to determine their ethical
educators modify grades or scores because of position on a range of assessment practices.
student effort, late work, or behavior problems,
for example, the scores do not accurately commu- 2. Method
nicate the level of mastery. Similarly, many teachers
do not use a blind grading system and may 2.1. Participants
unconsciously prefer certain students. These educa-
tors may unintentionally engage in score pollution The participant pool for this study consisted of
by giving less favored students lower grades than educators in the graduate and undergraduate
they deserve. Thus classroom grades are ‘‘polluted’’ programs at two major southern USA universities.
by these other factors. Students, their families, and All of these individuals were enrolled in a classroom
other stakeholders in the education system need assessment course or a course with a classroom
valid information regarding academic achievement. assessment component. Participants completed the
survey prior to the class session that addresses ethics
in classroom assessment. The pool included pre-
1.1. Conceptual framework and research questions service teachers, in-service teachers, and adminis-
trators. This study focuses on the responses of
The present study draws on both theoretical and pre-service and in-service teachers.
empirical foundations in the areas of ethics and
assessment. From the ethics literature derives the 2.2. Instrument
assumption that ethical behavior is defined as acting
based on one’s judgment of an obligation. In The current study organizes the guidelines for
addition, the notion of an ever-present gap between ethical student evaluation into a framework that
principles and ethical behavior in specific circum- addresses both classroom assessment and standar-
stances requires a focus on specific judgments as dized testing. Thus, scenarios for items on both
well as on general principles such as Do No Harm. classroom assessment and classroom issues related
From the assessment literature comes the assump- to standardized testing were developed using
tion that assessment of student achievement should resources from the assessment literature (e.g., The
accurately reflect mastery of content, should be Student Evaluation Standards [JCSEE, 2003] and
administered fairly, and should be treated confiden- The Principles for Fair Student Assessment Practices
tially. Resulting empirical studies in this tradition for Education in Canada [Joint Advisory Committee,
have focused primarily on teachers’ knowledge base 1993]) and the authors’ experiences as well as
about practices related to standardized testing. In anecdotes gathered from graduate students over
the present study, we wished to explore the several semesters.
intersection of these two bodies of literature by The researchers used a web-based survey to
examining specific circumstances requiring ethical present a series of scenarios in which an educator
judgments about a variety of classroom assessment is engaging in a student evaluation practice that
practices. may or may not be ethical. For example, one
The guiding research question was, ‘‘What is the scenario states that ‘‘A high school social studies
degree of agreement among educators about specific teacher bases students’ final semester grade on 2
ethical questions related to assessment issues that multiple choice tests.’’ In completing the survey
arise in the classroom?’’ Evidence of strong agree- item, the educator selects the Ethical or Unethical
ment would suggest that educators use similar option.
guidelines and methods of judgment in enacting
assessment in the classroom. Evidence of weak 2.3. Procedure
agreement would suggest that further exploration
and dialogue related to ethical principles and their The first draft of the survey consisted of 36
application in specific circumstances would be scenarios and 6 questions about demographic
useful. In particular, clarification of how the process information. A pilot survey was conducted with 74
between principle and judgment takes place would participants in the spring of 2004. Participants were
be needed. Thus, the present study was designed to given oral instructions during class on how to access
survey educators by presenting a variety of class- the survey web site and complete the web-based
 ARTICLE IN PRESS
S.K. Green et al. / Teaching and Teacher Education 23 (2007) 999–1011 1003

survey. The results of the pilot survey were reviewed 90% (see Table 1). Items that fell within 80%
and six items that appeared confusing were modified agreement included a teacher spending time training
or replaced. students in test-taking skills (Item 5), a teacher
The final 36-item survey was administered in the administering a parallel form of a norm-referenced
summer and fall of 2004. Of 169 responses, 114 were state test (Item 6), and a teacher using a commercially
pre-service and 55 were in-service teachers. available publication with the same format and skills
as a state test (Item 12). One item reflected high levels
3. Results of disagreement (falling within the 50–70% range):
‘‘Based on his review of the district’s mathematics
3.1. Analyzing agreement and disagreement frameworks, a teacher creates learning activities with
specific math problems that are included in the annual
Seven categories were created based on the topics achievement test (Item 4).’’
of the scenarios: Standardized Test Preparation, In the category Test Administration, as shown in
Standardized Test Administration, Multiple Assess- Table 2, almost all (96%) of the in-service and pre-
ment Opportunities, Communication about Grad- service teachers rated as unethical a teacher drawing
ing, Grading Practices, Bias, and Confidentiality. a student’s attention during testing to a problem
For the total group of 169 respondents, the that the child had missed (Item 7). The other item in
categories with highest agreement on individual this category, in which a teacher corrects a student
items among respondents were calculated by deter- who is marking items out of sequence, fell into the
mining the percentage of items in that category on high disagreement range, with 69% indicating that
which at least 80% of respondents agreed. For the practice is ethical and 31% indicating unethical.
example, 90% of educators indicated it is ethical to The rates of agreement for Multiple Assessment
for a teacher to spend a class period to train his Opportunities (see Table 3) ranged from 78%
students in test-taking skills (e.g., not spending too to 99%. Most recognized as ethical the use of
much time on one problem, eliminating impossible many forms of assessment in the classroom (99%,
answers, guessing) prior to testing. Similarly, 96% Item 21). A large percentage rated as unethical the
of educators indicated it was unethical when a questionable practice of relying on one form of
teacher provides a hint about a child’s incorrect assessment in a classroom (85%, Item 25). Reliance
answer during the administration of a standardized on too few assessments for determining grades
test. In both instances, educators were in agreement (78%, Item 10) nearly reached the 80% agreement
in their views about whether a practice was ethical level.
or unethical. We selected the 80% level because it As seen in Table 4, for the category of Commu-
indicates considerable agreement among respon- nication about Grading, 98% of in-service and pre-
dents. In addition, few items were selected by 90% service teachers rated as ethical the appropriateness
of respondents as being ethical or unethical; so little of providing students with a written policy
would be gained by such an analysis. about the calculation of grades (Item 31), infor-
We also examined the areas of disagreement. ming students about the materials that are
When 50% of teachers indicate a practice is ethical important in preparing for a test (Item 27), and
and 50% label the practice unethical, then the split stating the manner in which a task will be graded
(i.e., disagreement) can be no greater. For example, when it is assigned (Item 1). The other item in this
high disagreement occurs when 57% of teachers category, using surprise items not on the study
indicate that the weighting of homework heavily in guide, fell into the high disagreement range, with
determining report card grades is ethical, and 43% 34% of both groups rating it as ethical and 66% as
indicate such a practice is unethical. Teachers, thus, unethical.
are split on this issue. Items with percentages from As shown in Table 5, in-service and pre-service
50% to 70% were identified as displaying substan- teachers’ views about Grading Practices varied
tial disagreement among teachers. widely. Levels of agreement ranged from 57% to
86%. Items where over 80% agreement was reached
3.2. Analysis within content categories included rating as ethical a teacher basing grades for
a group project on the group product plus a heavily
The level of agreement for the individual items in weighted individual component (Item 3, 83%),
Standardized Test Preparation ranged from 53% to rating as unethical a teacher lowering report card
 ARTICLE IN PRESS
1004 S.K. Green et al. / Teaching and Teacher Education 23 (2007) 999–1011

Table 1
Percentage of pre-service and in-service teachers indicating the ethicality of assessment practices in standardized test preparation

Item # Scenarios for standardized test preparation Respondents’ Total Pre-service In-service
answers (N ¼ 169), teachers teachers
% (N ¼ 114), % (N ¼ 55), %

5 A teacher spends a class period to train his Ethical 89.9 89.5 90.9
students in test-taking skills (e.g., not spending Unethical 10.1 10.5 9.1
too much time on one problem, eliminating
impossible answers, guessing).
12 A teacher uses scoring high on the MAT, a Ethical 88.8 86.6 89.1
commercially available publication with the Unethical 11.2 11.4 10.9
same format and skills as the Metropolitan
Achievement Test (but not the same items), in
preparation for state testing.
6 A teacher administers a parallel form of a Ethical 80.5 79.8 81.8
norm-referenced achievement test to her Unethical 19.5 20.2 18.2
students in preparation for the state testing.
The parallel form is another version of the
state test that assesses the same content;
however, the items on the parallel form are not
the same ones as on the state form of the
achievement test.
2 A teacher adds vocabulary words from a Ethical 74.0 81.6 58.2
standardized, norm-referenced verbal aptitude Unethical 26.0 18.4 41.8
test to classroom vocabulary tests.
23 An elementary teacher quizzes students in the Ethical 70.4 68.4 74.5
lunch line about the number of pints in a quart Unethical 29.6 31.6 25.5
because students had missed the item on
previous administrations of the state
standardized test.
4 Based on his review of the district’s Ethical 53.3 63.2 32.7
mathematics frameworks, a teacher creates Unethical 46.7 36.8 67.3
learning activities with specific math problems
that are included in the annual achievement
test.

grades for disruptive behavior (Item 35, 85%), Agreement for Bias Issues (Table 6) ranged from
rating as ethical a teacher considering student effort 52% to 94%. Most respondents (94%) agreed to the
when determining grades (Item 30, 85%), and rating ethicality of a teacher allowing a student with a
as ethical lowering grades for late work (Item 24, learning disability to use accommodations in testing
86%). (Item 34). They also agreed that it was unethical to
Grading Practices items with a high level of restrict the number of As in a class based on a belief
disagreement included a physical education teacher that student work is rarely perfect (Item 22, 80%
giving a student a zero as a homework grade for not agreement).
returning a form requiring a parent’s signature Two Bias items fell in the strong disagreement
(Item 8, 57% unethical), deducting more points on a range in this category. First, 59% of all respondents
test for a wrong answer than leaving an item blank rated as unethical the practice of only addressing
(Item 14, 69% unethical), changing a grade from a student strengths in narrative report cards (Item 36).
B+ to an A because a student had mastered course Second, 52% rated as unethical the practice of
objectives but not completed all homework assign- grading essay tests with the knowledge of student
ments (Item 13, 63% unethical), and weighting identities (Item 26).
homework heavily in determining report card In the Confidentiality scenarios, agreement was
grades (Item 32, 57% ethical). high, ranging between 76% and 98% (see Table 7).
 ARTICLE IN PRESS
S.K. Green et al. / Teaching and Teacher Education 23 (2007) 999–1011 1005

Table 2
Percentage of pre-service and in-service teachers indicating the ethicality of assessment practices in standardized test administration

Item # Scenarios for standardized test administration Respondents’ Total Pre-service In-service
answers (N ¼ 169), % teachers teachers
(N ¼ 114), % (N ¼ 55), %

17 While administering a standardized test, a Ethical 69.2 68.4 70.9

teacher notices that a child has skipped a Unethical 30.8 31.6 29.1
problem and is now recording all his answers out
of sequence on the answer form. The teacher
stops at the child’s desk and shows the student
where to record the answer he is working on and
instructs him to put the answer to each question
with the same number on the answer sheet.
7 While administering a standardized test, a Ethical 3.6 3.5 3.6
teacher notices that a child has missed a problem Unethical 96.4 96.5 96.4
that the student obviously knows. The teacher
stands by the child’s desk, taps her finger by the
incorrect problem, shakes her head, and walks
on to the next desk.

Table 3
Percentage of pre-service and in-service teachers indicating the ethicality of assessment practices in multiple assessment opportunities

Item # Scenarios for multiple assessment opportunities Respondents’ Total Pre-service In-service
answers (N ¼ 169), % teachers teachers
(N ¼ 114), % (N ¼ 55), %

21 A teacher assesses student knowledge by using Ethical 99.4 99.1 100

many types of assessments: multiple-choice tests, Unethical 0.6 0.89 0
essays, projects, portfolios.
10 A high school social studies teacher bases Ethical 21.9 22.8 20
students’ final semester grade on 2 multiple- Unethical 78.1 77.2 80
choice tests.
25 A second-grade teacher uses observations as the Ethical 15.4 14.9 16.4
sole method to assess what students have Unethical 84.6 85.1 83.6
learned.

Most of the in-service and pre-service teachers rated 36 items, did in-service teachers and pre-service
as unethical a teacher disclosing to parents the teachers provide ratings that were significantly
achievement scores of a child other than their own different. For item 4 (i.e., a teacher creates learning
(98%, Item 16), and identifying individual students’ activities with specific math problems that are
rankings on exams to others (93%, Item 20). included in the annual achievement test), pre-service
teachers were significantly more likely to rate the
3.3. Comparisons between in-service and pre-service practice as ethical than were in-service teachers
teachers (w2 ¼ 13.80, df ¼ 1, po0:001). Thus, with one
exception, teaching experience did not account for
We conducted chi-square analyses to compare the differences in labeling an assessment practice as
responses of in-service teachers and pre-service ethical or unethical.
teachers. With 36 items for comparisons, the Type
I error would be inflated. Thus, we applied the 3.4. Analysis across content categories
Bonferroni correction to our test-wise alpha of .05
to reduce our Type I, or false positive, error rate to To examine patterns in agreements and disagree-
.0014 (i.e., 0.05C36 comparisons). Only on 1 of the ments, we tabulated the number of items in each
 ARTICLE IN PRESS
1006 S.K. Green et al. / Teaching and Teacher Education 23 (2007) 999–1011

Table 4
Percentage of pre-service and in-service teachers indicating the ethicality of assessment practices in communication about grading

Item # Scenarios about communication about grading Respondents’ Total Pre-service In-service
answers (N ¼ 169), % teachers teachers
(N ¼ 114), % (N ¼ 55), %

1 A teacher states how she will grade a task when Ethical 98.2 97.4 100
she assigns it. Unethical 1.8 2.6 0
27 A teacher tells students what materials are Ethical 98.2 97.4 100
important to learn in preparing for a class test. Unethical 1.8 2.6 0
31 A middle school principal directs teachers to Ethical 97.6 98.2 96.4
give students a written policy that explains how Unethical 2.4 1.8 3.6
report card grades are calculated in their classes.
29 For the final exam, a teacher always uses a few Ethical 33.7 35.1 30.9
surprise items about topics that were not on the Unethical 66.3 64.9 69.1
study guide.

Note: All items in this table were classified as ‘‘Do No Harm’’.

Table 5
Percentage of pre-service and in-service teachers indicating the ethicality of assessment practices in grading practices

Item # Scenarios about grading practices Respondents’ Total Pre-service In-service

answers (N ¼ 169), % teachers teachers
(N ¼ 114), % (N ¼ 55), %

24 A teacher lowers grades for late work by one Ethical 85.8 86.8 83.6
letter grade for each day. Unethical 14.2 13.2 16.4
30 A teacher considers student effort when Ethical 85.2 88.6 78.2
determining grades. Unethical 14.8 11.4 21.8
3 For a group project, a teacher bases each Ethical 83.4 83.3 83.6
student’s grade on the group’s product and a Unethical 16.6 16.7 16.4
heavily weighted individual component.
33 A teacher considers a student’s growth in Ethical 79.3 78.9 80.0
assigning grades. Unethical 20.7 21.1 20.0
15 To encourage lively discussion in English III, a Ethical 74.6 72.8 78.2
teacher counts class participation as 30% of the Unethical 25.4 27.2 21.8
final grade.
32 A teacher weights homework heavily in Ethical 57.4 57.9 56.4
determining report card grades. Unethical 42.6 42.1 43.6
8 A physical education teacher gives a student a Ethical 42.6 42.1 43.6
zero as a homework grade for not returning a Unethical 57.4 57.9 56.4
form requiring a parent’s signature.
13 As a teacher finalizes grades, she changes one Ethical 37.3 33.3 45.5
student’s course grade from a B+ to an A Unethical 62.7 66.7 54.5
because tests and papers showed the student had
mastered the course objectives even though he
had not completed some of his homework
assignments.
14 To minimize guessing, a teacher announces she Ethical 30.8 29.8 32.7
will deduct more points for a wrong answer than Unethical 69.2 70.2 67.3
for leaving the answer blank.
 ARTICLE IN PRESS
S.K. Green et al. / Teaching and Teacher Education 23 (2007) 999–1011 1007

Table 5 (continued )

Item # Scenarios about grading practices Respondents’ Total Pre-service In-service

answers (N ¼ 169), % teachers teachers
(N ¼ 114), % (N ¼ 55), %

19 A middle school history teacher offers extra Ethical 26.0 21.9 34.5
credit opportunities to all his classes except the Unethical 74.0 78.1 65.5
advanced class.
28 A teacher uses student peer ratings as 40% of the Ethical 25.4 21.1 34.5
grade on an oral report. Unethical 74.6 78.9 65.6
11 An accounting teacher gives a student an F for Ethical 24.9 24.6 25.5
the course because the student missed the final Unethical 75.1 75.4 74.5
exam.
35 A teacher lowers report card grades for Ethical 15.4 18.4 9.1
disruptive behavior. Unethical 84.6 81.6 90.9

Table 6
Percentage of pre-service and in-service teachers indicating the ethicality of assessment practices items in bias

Item # Scenarios about bias Respondents’ Total Pre-service In-service

answers (N ¼ 169), % teachers teachers
(N ¼ 114), % (N ¼ 55), %

34 A teacher allows a student with a learning Ethical 94.1 93.9 94.5

disability in the language arts to use a tape- Unethical 5.9 6.1 5.5
recorder when the student answers the essay
questions on social studies tests.
26 A teacher always knows the identity of the Ethical 48.5 42.1 61.8
student whose essay test she is grading. Unethical 51.5 57.9 38.2
36 To enhance self-esteem, an elementary teacher Ethical 40.8 39.5 43.6
addresses only students’ strengths when writing Unethical 59.2 60.5 56.4
narrative report cards.

18 A teacher who knows a student had a bad week Ethical 24.9 24.6 25.5
because of problems at home bumps the Unethical 75.1 75.4 74.5
student’s participation grade up a few points to
compensate for his bad score on a quiz.
22 Two teachers teach different sections of the same Ethical 20.1 18.4 23.6
course. Because of his belief that students’ work Unethical 79.9 81.6 76.4
is rarely perfect, one teacher gives very few
grades of ‘‘A’’.

category that displayed high agreement (i.e., 80% or agreement among responses with only 4 of the 13
higher) and those that displayed high disagreement items reaching 80% agreement), followed by Bias,
(i.e., 50–70%). As shown in Table 8, the categories with 2 of the 5 items reaching 80% agreement.
with high agreement on individual items among Overall, 17 of the 36 items fell into the high
respondents included Communication about Grading agreement range (80–100% agreement).
(with 3 of the 4 items reaching at least 80% In examining the seven content categories, we
agreement), Confidentiality (2 of the 3 items), and also noted items with high levels of disagreement.
Multiple Assessment Opportunities (also 2 of 3). The Table 8 shows that 5 of the 7 categories contained
category of Grading Practices had the lowest items with high levels of disagreement, Standardized
 ARTICLE IN PRESS
1008 S.K. Green et al. / Teaching and Teacher Education 23 (2007) 999–1011

Table 7
Percentage of pre-service and in-service teachers indicating the ethicality of assessment practices in confidentiality

Item # Scenarios about confidentiality Respondents’ Total Pre-service In-service

answers (N ¼ 169), % teachers teachers
(N ¼ 114), % (N ¼ 55), %

9 A teacher discloses to the parents of a student Ethical 75.7 76.3 74.5

their child’s score on an intelligence test. Unethical 24.3 23.7 25.5
20 To motivate students to perform better, a science Ethical 6.5 5.3 9.1
teacher always announces that he is passing out Unethical 93.5 94.7 90.9
scored tests to students in order of points
earned, from the top score to the bottom score.
16 To calm the fears of distraught parents, a Ethical 2.4 1.8 3.8
teacher compares their child’s achievement Unethical 97.6 98.2 96.4
scores with the results of the student’s cousin
who is also in the class.

Note: All items in this table were classified as ‘‘Do No Harm’’.

Table 8
Categories ranked by percent of items showing high agreement (80%+) and high disagreement (40–70%)

Category Items showing high Items showing high

agreement disagreement

Communication about grading 3 (75%) 1 (25%)

Multiple assessment opportunities 2 (67%) 0
Confidentiality 2 (67%) 0
Standardized test preparation 3 (50%) 1 (17%)
Standardized test administration 1 (50%) 1 (50%)
Bias 2 (40%) 2 (40%)
Grading practices 4 (31%) 4 (31%)

Note: Rows do not sum to 100% because some items did not meet the criterion of 80–100% for high agreement or the criterion of 50–70%
for high disagreement.

Test Administration (1 of 2 items), Bias ( 2 of 5 ing more variation in responses for Score Pollution
items), Grading Practices (4 of 13 items), Commu- items.
nication about Grading (1 of 4 items), and Standar-
dized Test Preparation (1 of 6 items). Overall, 9 of 3.6. Summary of results
the 36 items fell into the high disagreement range
(40–70% agreement), with 10 items in the medium In summarizing results for all respondents,
agreement range (71–79% agreement). Communication about Grading had the highest levels
of agreement, with 75% of the items in the high
3.5. Agreement analysis on Do No Harm and Score agreement range and one item in the high disagree-
Pollution items ment range. The categories Confidentiality and
Multiple Assessment Opportunities also had high
In examining items related to the two principles, levels of agreement, with a majority of items (67%)
five out of seven Do No Harm items had high in the high agreement range and no items falling in
agreement 71.4%, and only one had high disagree- the high disagreement range. The categories with
ment (14.3%), In addition, Do No Harm items were the greatest split between respondents were Stan-
all in the top two agreement categories. Only 12 out dardized Test Administration, Bias, and Grading
of 28 Score Pollution items had high agreement Practices, with equal numbers of items in the both
(43%) and 8 had high disagreement (29%), suggest- high agreement and the high disagreement ranges
 ARTICLE IN PRESS
S.K. Green et al. / Teaching and Teacher Education 23 (2007) 999–1011 1009

and no more than 50% of the items in the high grades for late work (Item 24, 86%). Again, this is
agreement range. In addition, Do No Harm items an issue of score pollution because the grade begins
reached higher levels of agreement than Score to convey information about behavior (i.e., tardi-
Pollution items. Finally, in-service and pre-service ness in completing work) rather than providing
teachers responded similarly to all but one item. information about a student’s understanding of the
subject matter. These findings suggest that the issue
4. Discussion of score pollution in classroom grading should be
highlighted in future work.
Each year the media report the instance of an The area of score pollution is important to
educator who violates some norm of acceptable continue to address in future research and in the
assessment practices. The assessment literature classroom because of the complexity and nuance
provides some guidance for teachers in terms of involved. For example, to avoid score pollution, a
ethical and unethical practices in standardized grade on a group project should involve an
testing; however, few resources in assessment individual component and not just one grade for
directly address the ethics of classroom assessment the entire group (Woolfolk, 2003). If the group
practices. This initial effort to develop scenarios component is weighted significantly higher than the
that address principles of ethical assessment practice individual component, score pollution can result
indicates that pre-service and in-service teachers at because grades then have a stronger probability of
these universities had strong agreement on less than not reflecting the individual student’s mastery of the
half of the scenarios presented in this study. These learning goals. Similarly, teachers often include
findings suggest that assessment is a realm without dimensions, such as neatness or class participation,
professional consensus. Thus, courses related to in their grading schemes that are not directly related
instruction and assessment should address the needs to the mastery of the learning goals. If the
of educators in these areas. Strike (1990) suggests percentage devoted to such factors has minimal
that specific moral concepts are unlikely to be impact on the overall grade, such practices could be
addressed in pre-service teacher education curricu- ethical. However, if such factors are weighted
lum. He recommends explicit instruction in ethical heavily enough to change a grade, they may result
concepts, grounded in the activities that are vital to in score pollution. Score pollution is one issue where
the teaching profession. Thus, conversations about theory meets reality in the classroom. The college
principles such as Do No Harm and Avoid Score incident where professors were fired for refusal to
Pollution as they relate to daily assessment practices allocate 60% of their grade to effort is a case in
in schools could be productive. In addition, research point.
has shown that support for teachers embedded in While we also see the guiding principle of Do No
everyday details of their ongoing activities is an Harm as a valuable one, we recognize it is not
effective approach for influencing beliefs and unproblematic. What constitutes harm is itself often
practices of teachers (Borko, Mayfield, Marion, a matter of judgment within context. Even within
Flexer, & Cumbo, 1997). These findings suggest that the realm of assessment, there are a variety of harms
discussions in schools around ethical principles and that could be done to students. There is the
classroom assessment could also be useful. potential educational harm done as the result of
One area that teachers experienced disagreement assessments that fail to accurately measure the
was that of grading practices, the area that created knowledge or skills that they claim to measure.
the firestorm in a Midwestern community. Only 4 of There is also the potential emotional harm done to
the 13 items reached high agreement, and two of students in the form of anxiety or other stress that
these items contrasted with the recommendations of high-stakes assessments often bring about. There is
the assessment community. Most assessment experts also the potential for harm of the teacher–student
agree that grades should reflect only the extent to relationship. Teacher–student trust can be damaged
which students have mastered the goals of instruc- by assessments that the student perceives as unfair
tion (e.g., Brookhart, 2004; Smith et al., 2001), yet a or unfounded.
strong majority of respondents (85%) rated as At times, Do No Harm emerges in practice as a
ethical considering effort or late work when choice between harms. A teacher may have to
calculating grades (Item 30). Similarly, a majority choose between a high-stakes exam that yields
of teachers rated as ethical the lowering of students’ important data about student performance and the
 ARTICLE IN PRESS
1010 S.K. Green et al. / Teaching and Teacher Education 23 (2007) 999–1011

emotional stress that such exams cause in students involve nuances, the findings from the current study
(and teachers). We see this situation (the choice provide a picture of the degree of agreement with
between harms) as potentially more ‘‘normal’’ than these insights by this group of pre-service and in-
the avoidance of harm altogether. service teachers. This study has been an effort to
Clearly, the Do No Harm formulation of assess- initiate a dialogue and discussion of these important
ment ethics is not, in and of itself, without issues. As the frequent newspaper anecdotes attest,
complexities and problems. We believe, however, a more consistent sense of ethical assessment
that it serves as a useful beginning to approaching practices among teachers, as well as the general
the issues of our research. public, is needed. Implementing ethical assessment
To further expand the point of complexity, in the classroom is not simply a matter of
educators’ responses probably reflect the extent to advocating ‘‘best practice.’’ A number of teachers
which respondents are bound by normative rules in have paid a significant price for violating what has
their own teaching situations. For example, some been perceived, rightly or wrongly, as ethical
school districts might purchase test preparation assessment practice in their communities.
materials for use in the schools, and thus teachers
might consider using such materials as ethical.
When presented with the scenario ‘‘A teacher uses
Scoring High on the MAT, a commercially available
References
publication with the same format and skills as the
Metropolitan Achievement Test (but not the same Airasian, P. (2005). Assessment in the classroom: A concise
items), in preparation for state testing,’’ teachers in approach (2nd ed.). Boston, MA: McGraw-Hill Company.
such school districts might indicate the practice is American Educational Research Association, American Psycho-
ethical. In other settings, however, educators who logical Association, & National Council on Measurement in
use test preparation materials will find such Education. (1999). Standards for educational and psychological
testing. Washington, DC: American Educational Research
practices considered to be unethical and their Association.
actions open to sanction (Natrona County School American Federation of Teachers, National Council on Mea-
District, 2001; North Carolina Department of surement in Education, & National Education Association
Public Instruction, 2000; South Carolina Depart- (1990). Standards for teacher competence in educational
ment of Education, 2005). assessment of students. Retrieved November 2, 2004 from
http://www.unl.edu/buros/article3.html.
One of the limitations of the current research Baumgart, N. (1996). Codes of ethics on assessment practice:
method was the requirement that respondents Confronting a quandary or inculcating ethical character?
endorse a practice as either ethical or unethical Overview of two sample documents. Assessment in education:
based on limited information. The use of only two Principles, policies & practice, 3.
Borko, H., Mayfield, V., Marion, S., Flexer, R., & Cumbo, K.
options implies no possibility of a gray area and
(1997). Teachers’ developing ideas and practices about
requires limited, noncontextual responses. The mathematics performance assessment: Successes, stumbling
contexts established in the scenarios require inter- blocks, and implications for professional development.
pretation, and educators might arrive at differing Teaching and Teacher Education, 13, 259–278.
decisions due to nuances in their interpretations of Bright, E. (1992). Teachers’ views of ethical standardized test use.
the context. Some educators might consider the Paper presented at the annual meeting of the National Council
on Measurement in Education. San Francisco, CA.
nuances that are implicit in a scenario and see a Brookhart, S. (2004). Grading. Upper Saddle River, NJ: Pearson.
given assessment practice as being ethical in some Carroll, D. (2002). Piper teacher who resigned in plagiarism
situations and unethical in others, creating a gray dispute is a hit with media. The Kansas City Star. Retrieved
area for the ethicality of a practice. The scenarios, August 3, 2004 from http://www.kansascity.com/mld/kansas-
however, create an opening for discussion of the city/news/local/2634836.htm?1c.
Click, C. (2004). Benedict fires two professors bucking effort-
ethics of practice. In future research, the use of based grading. The State. Retrieved December 15, 2004
open-ended responses and/or a range of factual from http://www.thestate.com/mld/thestate/news/breaking_
variations for each type of scenario could yield news/9447674.htm.
interesting comparisons. Colnerud, G. (1997). Ethical conflicts in teaching. Teaching and
Teacher Education, 13, 627–635.
The designation of items as ethical or unethical in
Gipps, C. V. (1994). Beyond testing: Towards a theory of
Section 4 was based on a survey of the assessment educational assessment. Washington, DC: The Falmer Press.
literature and the opinion of leaders in the assess- Gipps, C., & Murphy, P. (1994). A fair test? Assessment,
ment field. Given that some of these issues do achievement and equity. Buckingham: Open University Press.
 ARTICLE IN PRESS
S.K. Green et al. / Teaching and Teacher Education 23 (2007) 999–1011 1011

Haladyna, T. M., Nolen, S. B., & Haas, N. S. (1991). Raising Natrona County School District (2001). Board policy # 6302:
standardized achievement test scores and the origins of test Test ethics. Casper, Wyoming: Retrieved March 21, 2006,
score pollution. Educational Researcher, 20, 2–7. from http://mail1.ncsd.k12.wy.us.
Hostetler, K. D. (1997). Ethical judgment in teaching. Boston: Noddings, N. (1984). Caring: A feminine approach to ethics and
Allyn and Bacon. moral education. Berkeley: The University of California Press.
Impara, J. C., Plake, B. S., & Fager, J. J. (1993). Educational North Carolina Department of Public Instruction (2000). Testing
administrators’ and teachers’ knowledge of classroom assess- code of ethics. Retrieved March 21, 2006, from http://
ment. Journal of School Leadership, 3, 510–521. www.ncpublicschools.org/docs/accountability/testing/policies/
JCSEE. (2003). The student evaluation standards. Arlen Gulli- testcode080100.pdf.
ckson, Chair. Thousand Oaks, CA: Corwin. Nussbaum, M. (1986). The fragility of goodness. Cambridge:
Joint Advisory Committee. (1993). The principles for fair student Cambridge University Press.
assessment practices for education in Canada. Edmonton, Nussbaum, M. (1990). Love’s knowledge. New York: Oxford
Alberta. Retrieved July 15, 2004 from http://www.acposb. University Press.
on.ca/fair.pdf. Payne, D. A. (2003). Applied educational assessment (2nd ed.).
Jonsen, A., Siegler, M., & Winslade, W. (1998). Clinical ethics. Belmont, CA: Wadsworth Publishing.
New York: McGraw Hill. Plake, B., & Impara, J. (1997). Teacher assessment literacy: What
Kant, I. (1785). Foundation for metaphysics of morals (Thomas do teachers know about assessment? In G. Phye (Ed.),
Kingsmill Abbott, Trans.). In W. Kaufmann, & F. Baird Handbook of classroom assessment: Learning, achievement,
(Eds.), From Plato to Nietzsche (pp. 834–877). Englewood and adjustment. San Diego: Academic Press.
Cliffs, NJ: Prentice-Hall. Popham, W. J. (1991). Appropriateness of teachers’ test-
Kant, I. (1790). Critique of judgment (J.H. Bernard, Trans.). New preparation practices. Educational Measurement: Issues and
York: Hafner Publishing Company, Inc. 1951. Practice, 10(4), 12–15.
Kher-Durlabhju, N., & Lacina-Gifford, L. (1992). Quest for test Popham, W. J. (2000). Modern educational measurement:
success: Preservice teachers’ views of ‘‘high stakes’’ tests. Practical guidelines for educational leaders. Needham, MA:
Paper presented at the annual meeting of the Mid-South Allyn & Bacon.
Educational Research Association, Knoxville, TN. Rawls, J. (1971). A theory of justice. Cambridge, MA: Harvard
Kilian, L. J. (1992). A school district perspective on appropriate University Press.
test preparation practices: A reaction to Popham’s pro- Smith, J. K., Smith, L. F., & DeLisi, R. (2001). Natural classroom
posals. Educational Measurement: Issues and Practice, 11(4), assessment: Designing seamless instruction and assessment.
13–16. Thousand Oaks, CA: Corwin Press.
Mehrens, W. A., & Kaminski, J. (1989). Methods for im- South Carolina Department of Education. (2005). PACT
proving standardized test scores: Fruitful, fruitless, or administration manual: Spring 2005. Columbia, SC: Author.
fraudulent? Educational Measurement: Issues and Practice, Stiggins, R. J. (1999). Evaluating classroom assessment training
8(1), 14–22. in teacher education programs. Educational Measurement:
Messick, S. (1993). Validity. In R. Linn (Ed.), Educational Issues and Practice, 18(1), 23–27.
measurement (2nd ed., pp. 13–103). Washington: American Strike, K. (1990). Teaching ethics to teachers: What the curriculum
Council on Education. should be about. Teaching and Teacher Education, 6, 47–53.
Moore, W.P. (1993). Preparation of students for testing: Teacher Strike, K., & Soltis, J. (1998). The ethics of teaching (3rd ed.).
differentiation of appropriate and inappropriate practices. New York: Teachers College Press.
Paper presented at the annual meeting of the National Council Taylor, K., & Nolen, S. (2005). Classroom assessment: Supporting
on Measurement in Education, Atlanta, GA. teaching and learning in real classrooms. Upper Saddle River,
Munson, R. (2000). Intervention and reflection: Basic issues in NJ: Pearson Education, Inc.
medical ethics. Belmont, CA: Wadsworth Publishing. Thorndike, R., Cunnningham, G., Thorndike, R., & Hagen, E.
National Association for the Education of Young Children (1991). Measurement and evaluation in psychology and
(2005). Code of ethical conduct and statement of commitment. education (5th ed.). New York: McMillan.
Retrieved December 27, 2005, from http://www.naeyc.org/ Woolfolk, A. (2003). Educational psychology: Developing learners
about/positions/pdf/PSETH05.PDF. (4th ed.). Upper Saddle River, NJ: Merrill Prentice-Hall.