Content uploaded by Chandra Muller
All content in this area was uploaded by Chandra Muller
Content may be subject to copyright.
Social Psychology of Education 2: 199–216, 1998.
© 1998 Kluwer Academic Publishers. Printed in the Netherlands. 199
The Minimum Competency Exam Requirement,
Teachers’ and Students’ Expectations and
University of Texas at Austin
Abstract. This paper analyzes whether the minimum competency exam requirement for high school
graduation affects students’ academic performance directly or affects the educational process by
moderating the effect of teachers’ expectations on students’ mathematics test score gains, proﬁ-
ciency levels, and high school graduation. Tenth-grade students and their mathematics teachers from
the National Education Longitudinal Study of 1988 are analyzed. Contingent, negative associations
were found between the minimum competency exam requirement and both mathematics proﬁciency
and performance. The requirement was also not found to be associated with the odds of earning
a diploma. In the case of mathematics achievement, teachers’ expectations were a more important
predictor of learning gains and proﬁciency than were students’ expectations. Students’ expectations
better predicted who earns a diploma. The minimum competency exam requirement was found to
moderate the association between teachers’ expectations and mathematics achievement but did not
affect the relation between teachers’ expectations and high school graduation.
For nearly half a century, teachers’ expectations have been recognized as central
to students’ successes or failures. Even when students are motivated to succeed
they must count on schools to teach the cognitive skills and knowledge to achieve
their aspirations. Yet, teachers with lower expectations may teach and demand less
than is needed to attain those goals. Recently, education reformers have suggested
that external assessments of student progress may provide teachers with impor-
tant motivation to ensure student academic competence in critical subject areas.
Thus, while teachers may not share students’ expectations for future educational
?A previous version of this paper was presented at the 1996 annual meeting of the American
Sociological Association in New York City. The research reported in this paper was supported by a
grant from the American Educational Research Association which receives funds for its “AERA
Grants Program” from the National Science Foundation and the National Center for Education
Statistics (U.S. Department of Education) under NSF Grant #RED-9452861, by a faculty fellow-
ship from the Charles A. Dana Center for Mathematics and Science Education, and by a summer
research fellowship from the University of Texas. Opinions reﬂect those of the author and do not
necessarily reﬂect those of the granting agencies. I wish to thank Ronald Angel, Bruce Biddle,
Kathryn Borman, Kathryn Schiller, David Stevenson, and anonymous reviewers for helpful com-
ments. Correspondence concerning this article should be sent to Chandra Muller, Department of
Sociology, University of Texas, Austin, TX 78712 U.S.A. Tel: 512-471-1122; Fax: 512-471-1748;
200 CHANDRA MULLER
attainment, the external assessment would encourage teachers to teach the material
required for progress toward students’ aspirations so long as that material is rep-
resented in the external assessment. This paper analyzes whether one common but
weak form of external assessment, the minimum competency exam requirement
for high school graduation, affects students’ academic performance directly or
affects their educational process by moderating the effects of teachers’ expecta-
tions on students’ mathematics test score gains, proﬁciency levels, and high school
The minimum competency exam requirement for high school graduation has been
in practice in about half of the U.S. states for at least two decades (Catterall, 1989).
These examinations, that students must pass before high school graduation, are
designed to ensure basic knowledge and skills of all students who obtain a high
school diploma. Initially controversial, the low-level performance standards are
now thought to be a minimal obstacle for almost all students. Individual states
develop their ownexaminations, and students are usually allowed multiple chances
to pass. One concern about the minimum competency exam is that students who
fail the exam early in high school may either drop out or lower their expectations
and that this process has yet to be measured because it takes place earlier than
is focused on in most studies of high school drop outs and graduation. Unlike
minimum competency testing, national standards and assessments have yet to be
implemented. Analyses of the effects of the minimum competency exam are im-
portant in light of the massive effort to implement national standards and reform
and the lack of knowledge about their effects.
Proponents of external assessments argue that they will provide a motivation for
teachers to encourage students to maximize learning measured by the assessment.
If well designed, external assessments would encourage teachers to work with stu-
dents in a coach-like relationship toward a common goal. Coupled with a system of
accountability, external assessments would shift the incentive structure for teachers
to one which rewards gains in student achievement (Porter, 1994). Teachers would
have a greater sense of responsibility for their students’ achievement, effectively
monitor student progress, and change weak pedagogical approaches. Perhaps most
importantly, they would be more likely to use the students’ resources, including
motivation, to help students learn (Coleman, 1995). Furthermore, a system of ex-
ternal assessment would take the burden of evaluation away from the teacher,
encouraging a more cooperative relationship between teacher and student. In short,
external assessments may shift teachers’ goals and in doing so may change the
Controversy surrounds the recent policy initiatives to set national standards
for student performance and accompany them with assessments of progress. Op-
ponents argue that it is impossible to design an exam that adequately measures
MINIMUM COMPETENCY EXAM REQUIREMENT 201
learning for all students. Additionally, they argue, it is not possible to measure
higher-order cognitive skills with such tests, encouraging teachers to teach super-
ﬁcially to the test instead of teaching advanced cognitive skills. Some opponents
fear that students will become discouraged because they think there is no chance of
passing the exam (and therefore graduating) (Catterall, 1989). Others insist that the
source of the problem is outside the school in larger economic problems associated
with poverty (Biddle, 1997). Further debate exists overuse of test results and policy
implementation. A key feature of recent initiatives is that compliance be voluntary
on the part of states or in some cases districts or schools (Baker, 1994).
The national standards and assessments debate has been interesting, in part, be-
cause of its history. First supported by conservative Republicans, the initiative has
the full support of many Democrats, including the Clinton Administration (Biddle,
1997). The early motivation for the minimum competency exam was to ensure a
minimum level of skills and knowledge for all students who obtained a high school
diploma. This, proponents reasoned, would provide graduates with a minimum
level of “equal” access to jobs requiring those skills (Cohen & Haney, 1980).
Additionally, employers hiring students with a high school diploma would be as-
sured of their basic knowledge and skill levels. As the initiatives broadened and
were adopted by proponents with other interests, standards and assessments were
also viewed as a check on local control and a means of monitoring whether stu-
dents were taught a curriculum that ensured equal access to advanced coursework
and schooling. This last motivation about equal access is particularly interesting
in light of the early opponents to minimum competency testing, who objected
strongly because they observed especially low passing rates among minority stu-
dents (Fisher, 1980). It is important to keep in mind that the minimum standards of
the competency exam do not target performance above a low level.
These differences aside, most supporters concur that the standards and assess-
ments are a way to remove decisions about who progresses to the next level of
school from those who teach the curriculum. Additionally, regardless of the con-
troversy and divergence of interests among those who support national standards
and assessment, most analysts expect that implementation of the policy will change
teaching and classroom processes. Proponents argue that one change will be in the
relation between expectations and outcomes because of a change in incentives.
Opponents argue that curriculum and pedagogy will change what is learned and
how it is taught.
The recent initiative for national standards and assessment is a much more far
reaching external assessment policy than minimum competency exams for high
school graduation. Although voluntary, national standards and assessment involve
the development of common standards for student proﬁciency at speciﬁed grade
levels and the use of a common instrument to evaluate proﬁciency and progress.
The proposed national standards and assessments imply broader rewards and sanc-
tions for educational organizations and assessment of students at all levels of pro-
ﬁciency, while the minimum competency exam measures a relatively low level of
202 CHANDRA MULLER
achievement and applies the most serious consequences of failure to the student.
If a student fails the minimum competency exam, the student does not receive a
high school diploma. Nonetheless, the use of an external assessment rather than
teachers’ more subjective assessments may still have consequences for classroom
processes, and in particular, for the effect of expectations and motivation on student
Teachers form their expectations of students based on students’ prior performance
and their perception of student opportunity, using indicators like test scores, track
placement, student behavior, student expectations, student socioeconomic status
(SES), gender, and racial and ethnic characteristics of the student (Delpit, 1995;
Jones, 1990; Oakes, 1985; Rist, 1970; Williams, 1975). Little is understood about
how students’ expectations form and change over the course of school experience.
Wilson and Wilson (1992) found that family factors (like parents’ educational level,
SES, and race) and the students’ perception of the parents’ and teachers’ aspira-
tions were important predictors of the student’s own aspirations. Hanson (1994)
showed that High School and Beyond seniors with high expectations adjusted their
expectations downward or did not realize their goals, and that the process varied
by gender, race or ethnicity, and SES. Student and school characteristics were im-
portant predictors of student expectations. Catterall (1989) suggested that students
may adjust their expectations downward after their ﬁrst failure on the minimum
competency exam. In many states this occurs during Grade 9 or Grade 10.
Both teachers’ and students’ expectations affect students’ academic achieve-
ment. Drazen (1994), using the National Education Longitudinal Study of 1988
(NELS), found that teacher expectations for the class were a strong predictor of
student achievement; in fact, they were the best predictor in her models. Her ﬁnd-
ings were consistent with those of others about the powerful relationship between
the expectations of teachers and students’ achievement and persistence in school
(Good, 1993; Persell, 1977; Rist, 1970; Rosenthal & Jacobson, 1968). Similarly,
students’ expectations are positively associated with achievement (e.g., Coleman,
Hoffer, & Kilgore, 1982; Spenner & Featherman, 1978) and other forms of aca-
demic behavior such as fewer behavior problems, greater investment in school
(MacLeod, 1987), and less likelihood of dropping out (Ensminger & Slusarcick,
In the early and middle 1980s, schools were encouraged to change the culture
and raise teachers’ expectations. Today, campaigns proclaiming positive attitudes
may be found in many schools despite slim evidence that such practices are effec-
tive or that they represent the attitudes and behavior of school personnel. While
the language changed, those policies do little to address the mechanisms by which
teachers’ low expectations are translated into lower performance of students. (For
a thorough summary of the process see Good and Brophy, 1997.)
MINIMUM COMPETENCY EXAM REQUIREMENT 203
Unlike policies of the early 1980s designed to raise teacher expectations for
students, external assessments are targeted at changing the incentives for teachers
and students to work together toward student mastery of curriculum. Coleman
(1995) argued that the system by which teachers are expected to evaluate stu-
dents’ performance is fundamentally ﬂawed if learning is to take place because
it is structured in direct opposition to the way teachers are themselves evaluated
and rewarded. He claimed that it provides incentives for teachers and students to
enter into tacit agreements in which teachers exchange low academic demands for
students’ orderly behavior and compliance. The system of external assessments
would provide common incentives for teachers and students to maximize student
learning gains cooperatively because each would be evaluated on the learning that
takes place. A key aspect of Coleman’s model is that teachers and students are
evaluated on the basis of “value-added” and not simply the performance level of
students. Independent of the teachers’ expectations for students’ futures, propo-
nents of external assessments argue that teachers will do much more to raise the
achievement levels of students because the independent, external evaluation will
reﬂect directly on teachers’ job performance.
The effect of external assessments as an incentive structure for teachers should
be evident in the way teachers’ expectations are associated with student outcomes.
If external assessments motivate teachers to exploit students’ resources for learn-
ing in ways that are independent of the teachers’ own opinion about the students’
future, there should be a weaker relationship between teachers’ expectations and
outcomes when external assessments are in place. In contrast, in the absence of
external assessments, the relationship between teachers’ expectations and students’
outcomes should be strong.
THE MINIMUM COMPETENCY EXAM AND EXPECTATIONS
The minimum competency exam may change the effects of teacher attitudes and
behavior in several ways. It may motivate teachers to teach more material so stu-
dents pass the exam. This might be especially true if the minimum competency
exam is used to assess teacher or school performance because there would be
additional incentives for teachers to have students pass the exam. Alternatively, the
exam may do little to change teacher behavior, but it may circumvent the inﬂuence
of the teacher attitudes or behavior on student outcomes because it reduces the
extent to which the teacher must act as gatekeeper in the decision about which
students should graduate.
Tosummarize, Figure 1 shows the expected effects of the minimum competency
exam on the relationships between teachers’ expectations, students’ expectations,
and high school graduation and achievement. The argument is that teachers’ ex-
pectations will have observed effects on students’ expectations and on students’
outcomes when teachers act as gatekeepers, as is the case when there is no external
assessment (shown in the top half of the ﬁgure). When teachers are in a position
204 CHANDRA MULLER
Figure 1. Models of the relationship between teachers’ expectations, students’ expectations,
and students’ academic outcomes depending on the presence of a minimum competency exam
of coach, as they are more likely to be when there is an external assessment upon
which even they may be evaluated, then there will be no independent effect of
their own expectations on students’ outcomes. Teachers’ expectations may become
irrelevant either because teachers are motivated to behave in ways that do not reﬂect
their expectations or because they are not in a gatekeeping position. In each case
background, prior achievement and course taking may have an independent effect
on students’ outcomes.1
MINIMUM COMPETENCY EXAM REQUIREMENT 205
It is important to evaluate student learning separately from graduation because
the external assessment may affect one outcome without the other. Speciﬁcally,
an exam that targets high school graduation may make a difference in graduation
but not learning (because the standards are rudimentary). Similarly, the minimum
competency exam may inﬂuence the process by which teachers’ expectations are
associated with students’ expectations and outcomes. This article analyzes the
background determinants of teachers’ and students’ expectations and then eval-
uates the process by which expectations affect high school graduation and math-
ematics test scores. The analyses also evaluate the direct effect of the minimum
competency exam on students’ outcomes.
Data and Method
This study employed data from the National Education Longitudinal Study of 1988
(NELS). NELS used a nationally representative sample of 24,599 eighth-grade
students who were followed up in the tenth and twelfth grades. In each wave,
a sample of teachers were selected to provide information about the student, the
classroom environment, and the teacher. Some students had two teachers (usually
either an English or social studies teacher and a math or science teacher), while
other students had no teachers. Additionally, in each wave, students were admin-
istered a battery of achievement tests in reading, mathematics, social studies, and
science (see Ingels et al., 1994, for a complete description of the data set).
Tenth-grade students paired with a mathematics teacher were selected for the
analysis because the tenth-grade teacher data included information about teachers’
expectations. Analysis of tenth graders also allowed for the estimation of a lagged
effect of teachers’ expectations on twelfth-grade mathematics test performance
and on graduation. Only public school students and teachers were included in the
present study because classroom processes and student-teacher relationships may
be different in private schools (cf. Bryk, Lee, & Holland, 1993; Coleman & Hoffer,
1987). Additionally, data from Native Americans were excluded because they were
signiﬁcantly different from other racial and ethnic categories yet did not comprise
a large enough group to allow for meaningful analysis (Schneider & Coleman,
1993). Selection on the basis of these characteristics reduced the original sample of
16,813 public school students to 6,192.2Furthermore, students were only included
if they had non-missing data for all analysis variables described below, reducing
the sample size further to 3,442.3
As mentioned above, it was important to consider learning and attainment sepa-
rately. Mathematics achievement was measured by the Grade 12 proﬁciency level
of students, and learning was measured by regressing Grade 12 mathematics item
206 CHANDRA MULLER
response theory (IRT) scores on the scores in Grade 10. Student attainment was
measured by whether the student obtained a high school diploma as reported in the
NELS third follow-up, two years after most students graduated. Schools reported
whether students were subject to a minimum competency exam requirement to
graduate. Students and teachers each reported their educational expectations for the
student. Students also reported their average mathematics grades for the last two
years. Teachers reported the ability level of the mathematics class. Finally, student
background, including socioeconomic status (SES), race and ethnicity, urbanicity,
and gender were included in all analyses. Table 1 shows variable construction
and descriptive statistics for the combined group of students and, separately, by
whether students were subject to a high school graduation minimum competency
In this sample, students were split about evenly between those who were subject
to the minimum competency exam and those who were not. The two groups of
students differed in almost every respect measured. It was only in the proportions
of Asians, boys, and students in low- and mixed-ability groups that were not sig-
niﬁcantly different with respect to being required to take a minimum competency
exam. A much larger proportion of African Americans were subject to the re-
quirement. In addition, students who were required to take the exam had higher
educational expectations and lower mathematics achievement test scores.
The minimum competency exam may affect the attitudes of teachers and students,
it may inﬂuence classroom processes, and it may inﬂuence student outcomes. Thus,
this analysis evaluates the relationships between student background, student and
teacher attitudes, student outcomes, and differences in the associations depending
on the presence of a minimum competency exam. To understand how teachers’ and
students’ expectations are associated with outcomes, one must ﬁrst consider how
teachers’ and students’ expectations are shaped. The ﬁrst analyses estimates the
determinants of teachers’ expectations, then of students’ expectations, followed
by predictions of students’ test score gains, mathematics proﬁciency levels, and
obtaining a high school diploma.
Table 2 includes models that estimate teachers’ expectations and, separately, stu-
dents’ expectations. The basic models control for background and the minimum
competency exam requirement. Academic performance and ability group are added
in the full models. Additionally, to measure whether the relationship between teach-
ers’ and students’ expectations varies depending on the presence of the exam re-
quirements, teachers’ expectations and the interaction between teachers’ expecta-
MINIMUM COMPETENCY EXAM REQUIREMENT 207
Table I. Variable Description and Descriptive Statistics for (1) Combined Sample and for Students
(2) Without and (3) With a Minimum Competency Exam Requirement (Weighted).
SESbFor most students, based on parents’ highest education, family income
and parents’ occupation. NCES constructed variable.a(1) M= –.059,
SD = .269, (2) M= –.09, SD = .262, (3) M= –.03, SD = .276
Race/ethnicity NCES constructed variable based on student report (Native Americans
excluded from analysis, European Americans are base category for
(1) Asian = 3.5%; Latino = 6.7%; African American = 11.6%, (2)
Asian = 3.1%, Latino = 2.9%,bAfrican American = 4.8%,b(3) Asian
= 3.9%, Latino = 10.3%, African American = 17.9%
UrbanicitybNCES constructed variable (base category is suburban).
(1) Urban = 24.0%; Rural = 19.4%, (2) Urban = 18.7%; Rural = 29.4%,
(3) Urban = 29.1%, Rural = 9.9%
Gender NCES constructed variable based on student report (base category is
(1) Boys = 49.5%, (2) Boys = 49.9%, (3) Boys = 49.1%
expectationsbTenth graders’ response to “Asthings stand now, how far in school do
you think you will get?” Responses ranged from 1 = less than high
school to 9 = Ph.D., M.D.
(1) M= 6.185, SD = .772, (2) M= 6.028 SD = .773, (3) M= 6.332, SD
expectationsbGrade 10 mathematics teachers’ responses to “Will this student proba-
bly go to college?” Response categories were recoded as yes = 3, don’t
know = 2, no = 1.
(1) M= 1.24; SD = 0.321, (2) M= 1.194; SD = .32, (3) M= 1.284, SD
requirement School reported minimum competency exam requirement for gradua-
tion. Fifty-one percent of students are subject to a requirement.
Ability group Teacher’s response to the question, “Which of the following best
describes the achievement level of the students in this class com-
pared with the average 10th-grade student in this school?” Re-
sponse categories including “lower achievement levels,” “widely dif-
fering achievement levels,” “higher achievement levels,” and “average
achievement levels” were combined in this analysis. (1) Low = 22.4%,
Mixed = 10.4%, (2) Low = 23.4%, Mixed = 10.2%, (3) Low = 21.5%,
Mixed = 10.7%
Students’ gradesbTenth graders’ report of mathematics grades “from the ninth grade
(1) M= 4.567, SD = .684, (2) M= 4.489, SD = .7, (3) M= 4.639, SD
Math proﬁciencybTwelfth-grade mathematics proﬁciency level. (1) M= 2.683, SD = .49,
(2) M= 2.747, SD = .483, (3) M= 2.621, SD = .496
208 CHANDRA MULLER
Table I. continued
Grade 10 math testbItem Response Theory (IRT) score on ﬁrst follow-up mathematics test.
(1) M= 44.184, SD = 5.002, (2) M= 45.003, SD = 4.949, (3) M=
43.412, SD = 5.036
Grade 12 math testbItem Response Theory (IRT) score on second follow-up mathematics
(1) M= 49.098, SD = 5.074, (2) M= 50.055, SD = 5.004, (3) M=
48.178, SD = 5.122
High school diplomabDerived from a variable constructed by NCES. Students were coded as
having received a diploma if the variable indicated such; students with
a other status (GED, certiﬁcate, currently enrolled, currently working
toward equivalency, or drop out) were coded as not having a diploma.
(1) diploma = 90.5%, (2) diploma = 91.7%, (3) diploma = 89.5%
Note: Unweighted n= 3,442 for combined sample, n= 1,680 for students without minimum compe-
tency exam requirement and n= 1,762 for students with a minimum competency exam requirement.
All statistics in this table and those that follow are weighted and adjusted for the design effect as
recommended in Ingels et al. (1994).
aAll NCES constructed variables are described in Ingels et al. (1994).
bDifference in means or proportions of students who are and are not subject to the minimum
competency exam requirement is signiﬁcant at p<.05.
tions and whether students are subject to a minimum competency exam require-
ment are included in models of students’ expectations.
Table 2 shows that teachers’ expectations for African American students are
signiﬁcantly lower than for European American students. The difference between
African American and European American students is explained with the addition
of ability group, test scores, and grades. Analyses, not shown, indicate that the
difference between teachers’ expectations for African Americans and European
Americans can be explained by variation in the students’ test scores. Table 2 also
indicates that there is no signiﬁcant difference in teacher expectations depending
on the minimum competency exam requirement.
When ability group, test scores, and grades are added to the basic model, shown
in Model 2, test scores are the best predictor of teachers’ expectations. Ability
group and grades are also, independently, associated with teachers’ expectations.
Even after prior achievement and ability group are held constant, students’ SES
is strongly associated with teachers’ attitudes. Indeed, college attendance requires
resources that are more readily available to higher SES students, and teachers may
take this into account.
Model 3 indicates that SES is also a powerful predictor of students’ expec-
tations. Race and ethnicity are different in their associations with students’ ex-
pectations compared to teachers’ expectations. African American students have
higher expectations for themselves compared to European American students. This
MINIMUM COMPETENCY EXAM REQUIREMENT 209
Table II. Regression Coefﬁcients for Predictions of Teachers’ and Students’ Expectations.
Variable Teachers’ expectations Students’ expectations
Model 1 Model 2 Model 3 Model 4 Model 5
SES 1.107∗∗∗ .802∗∗∗ 1.1∗∗∗ .633∗∗∗ .548∗∗∗
(.164) (.119) (.384) (.221) (.191)
Gender (Boy = 1) –.437∗–.455∗–.483∗∗∗ –.295∗∗∗ –.324∗∗∗
(–.044) (–.046) (–.116) (–.071) (–.078)
Asian .427 .372 .697∗∗∗ .535∗∗∗ .516∗∗∗
(.016) (.014) (.061) (.047) (.046)
Latino –.166 .211 .242 .332∗∗ .465∗∗∗
(–.008) (.011) (.029) (.04) (.056)
African American –.583∗.174 .402∗∗∗ .7∗∗∗ .927∗∗∗
(–.038) (.011) (.062) (.107) (.142)
Urban .125 .146 .06 .019 .11
(.011) (.013) (.012) (.004) (.002)
Rural –.039 –.139 –.076 –.086 –.072
(–.003) (–.011) (–.014) (–.016) (–.014)
Exam requirement .239 .218 .137∗.234∗.22∗
(.024) (.022) (.033) (.056) (.053)
Low-ability group –1.209∗∗∗ –.802∗∗∗ –.517∗∗∗
(–.103) (–.16) (–.103)
Mixed-ability group –.525 –.568∗∗∗ –.451∗∗∗
(–.033) (–.083) (–.066)
Students’ grades .327∗∗∗ .087∗∗∗
Grade 10 math test score .063∗∗∗ .029∗∗∗
Teachers’ expectations .908∗∗∗ .662∗∗∗
Teachers’ expectations –.157∗–.124
×exam requirement (–.067) (–.053)
-2 log likelihood 864.08 695.38 .162 .328 .357
or adjusted R2
∗∗∗ p<.001; ∗∗ p<.01; ∗p<.05
Note: Logistic regression is used for models predicting teachers’ expectations, and OLS
regression is used to predict models for students’ expectations.
210 CHANDRA MULLER
difference becomes more pronounced when ability group, test scores, and grades
are held constant, as shown in the last model of Table 2.
It is possible that the minimum competency requirement changes how teachers
and students evaluate the students’ future educational opportunities. Model 4 in
Table 2 shows that the minimum competency exam requirement is moderately
associated with students’ higher expectations. Furthermore, teachers’ expectations
are a strong predictor of students’ expectations. That model also indicates that
the association between teachers’ and students’ expectations is moderated slightly
when students are subjected to a minimum competency exam requirement; the
interaction term for teachers’ expectations and the exam requirement is negative
and signiﬁcant. When mathematics grades and test scores are also included, in the
last model, the interaction effect is negative but nonsigniﬁcant. Nonetheless, the
expectations of students continue to be strongly inﬂuenced by the expectations of
STUDENTS’TEST PERFORMANCE AND HIGH SCHOOL GRADUATION
The minimum competency exam requirement was designed to raise the achieve-
ment levels of students, particularly the lowest performing students, and may also
affect their educational attainment. Table 3shows coefﬁcients from models predict-
ing three student outcomes, twelfth-grade mathematics proﬁciency levels, mathe-
matics test score gains, and whether the student obtained a high school diploma.
The results in Table 3 indicate that the minimum competency exam requirement
has a negative association with mathematics test score gains and proﬁciency levels,
which tends to disappear when ability group and prior grades are controlled. It is
not signiﬁcantly associated with whether students graduate from high school. Per-
haps the last ﬁnding is the most surprising because the test is speciﬁcally designed
to inﬂuence high school graduation. One might argue that it would not affect the
entire sample because it is targeted only at students achieving at the lowest levels.
However, analyses, not shown, of students who performed in the lowest 15% of
their eighth-grade mathematics achievement test indicate similar patterns.
Teachers’ and students’ expectations each appear to make a difference in stu-
dents’ outcomes. In general, teachers’ expectations best predict mathematics achieve-
ment, and students’ expectations best predict high school graduation. The ﬁrst two
models, in which mathematics proﬁciency levels are predicted, indicate that teach-
ers’ expectations are a strong predictor of proﬁciency levels even after ability group
and students’ grades are controlled (shown in Model 2). Teachers’ expectations
also moderately predict test score gains between Grades 10 and 12.
Students’ expectations predict mathematics proﬁciency well but do not have an
independent association with test score gains. In contrast to the stronger inﬂuence
of teachers’ expectations on mathematics achievement, compared to that of the
inﬂuence of students’ expectations, students’ expectations are highly associated
with high school graduation. As shown in Model 5 of Table 3, the basic model
MINIMUM COMPETENCY EXAM REQUIREMENT 211
Table III. Coefﬁcients from Predictions of Mathematics Test Score Gains, Mathematics Proﬁciency
Levels, and High School Graduation (Weighted).
Mathematics Grade 12 math High school
Variable proﬁciency test score diploma
Model 1 Model 2 Model 3 Model 4 Model 5 Model 6
Grade 10 math .864∗∗∗ .832∗∗∗
test score (.852) (.82)
SES .166∗∗∗ .186∗∗∗ .181 .385∗.322 .408
(.091) (.102) (.01) (.02) (.048) (.061)
Gender (Boy = 1) .205∗∗∗ .214∗∗∗ .74∗∗∗ .777∗∗∗ –.29 –.254
(.077) (.081) (.027) (.028) (–.03) (–.026)
Asian .145 .138 .577 .566 .925 .95
(.02) (.019) (.008) (.008) (.035) (.036)
Latino –.344∗∗∗ –.281∗∗∗ –.162 –.132 .066 .14
(–.065) (–.053) (–.003) (–.002) (.003) (.007)
African American –.625∗∗∗ –.502∗∗∗ –.224 –.204 –.709 –.561
(–.15) (–.121) (–.005) (–.005) (–.047) (–.037)
Urban .126∗.146∗∗ .351 .494∗∗∗ –.067 –.038
(.04) (.047) (.011) (.015) (–.006) (–.003)
Rural –.1 –.145∗∗ –.659∗–.714∗∗ .393 .377
(–.03) (–.043) (–.019) (–.021) (.032) (.03)
Exam requirement –.149∗∗∗ –.053 –.781∗∗∗ .209 –.283 –.159
(–.056) (–.02) (–.028) (.008) (–.029) (–.016)
Students’ .116∗∗∗ .069∗∗∗ .114 .03 .342∗∗∗ .299∗∗∗
expectations (.182) (.109) (.017) (.005) (.146) (.127)
Teachers’ .606∗∗∗ .435∗∗∗ 1.325∗∗∗ 1.374∗∗∗ .67∗∗ .689
expectations (.397) (.284) (.084) (.087) (.119) (1.22)
Low-ability group –.734∗∗∗ –1.02∗∗∗ –.601
(–.231) (–.031) (–.051)
Mixed-ability group –.329∗∗∗ –.859∗∗ –.479
(–.076) (–.019) (–.03)
Students’ grades .139∗∗∗ .593∗∗∗ .203∗
(.194) (.08) (.077)
Teachers’ expectations –.103∗–.907∗∗∗ –.401
×exam requirement (–.069) (–.059) (–.073)
–2 log likelihood .361 .434 .844 .85 241.78 233.75
or adjusted R2
∗p<.001; ∗∗ p<.01; ∗p<.05
Note: OLS regression is used to predict mathematics proﬁciency and test- score-gains models, and
logistic regression is used to predict high school diploma models. Standardized coefﬁcients in are
212 CHANDRA MULLER
predicting high school graduation, both teachers’ and students’ expectations are
signiﬁcantly associated with graduation. However, the association between teach-
ers’ expectations and graduation is not signiﬁcant in the full model, Model 6, which
includes students’ grades.
Students’ background, including SES, gender, and whether they live in an urban
area, are associated with mathematics achievement but not with whether they earn
a diploma. Also, African American and Latino students have lower mathematics
proﬁciency levels compared to European American students. There are no other
observed racial or ethnic differences in outcomes.
Table 3 also shows that teachers’ expectations are associated with test score
gains and mathematics proﬁciency levels somewhat differently depending on the
presence of an exam requirement. The interaction term for teachers’ expectations
and the minimum competency exam requirement is negative and signiﬁcant. This
indicates that the effect of teachers’ expectations on students’ test score gains and
proﬁciency level is moderated by the presence of a minimum competency exam re-
quirement. In other words, the effect of teachers’ expectations on the mathematics
outcomes is less for students who are subject to the exam requirement.
Discussion and Conclusion
Proponents of external standards and assessments generally argue that students will
learn more. These results do not support such an effect; indeed, my results suggest
that the requirement is negatively associated with achievement test scores. Early
critics of the minimum competency exam feared that it would cause more students
to drop out, which is also generally unsupported by evidence (Catterall, 1989).
Catterall (1989) suggested that there may be no measurable effects of the minimum
competency exams on dropping out because the level is rudimentary. Additionally,
political pressures to raise graduation rates have resulted in weak policies such as
those which allow students to retake the exam many times in an effort to pass. The
results I have presented are consistent with Catterall’s suggestions.
Proponents of external standards and assessments argue that the teacher-student
relationship will be more cooperative with teachers acting as coach rather than
gatekeeper. The minimum competency exam requirement may moderate the rela-
tionship between teachers’ expectations and mathematics achievement in that the
associations between expectations and the two measures of achievement are weaker
for students subjected to the requirement. These ﬁndings are partially consistent
with the model depicted in Figure 1; however, the relationship between teachers’
expectations and mathematics performance seems to be quite strong in all circum-
stances. One would also expect that the impact of teachers’ expectations on whether
a student earns a diploma would also be moderated by the requirement. This was
not found; however, teachers’ expectations did not generally have a strong effect
on who earns a high school degree.
MINIMUM COMPETENCY EXAM REQUIREMENT 213
An important qualiﬁcation of these results is that only public school students
who are enrolled in mathematics courses were studied(because mathematics teacher
data were used). Moreover, it is impossible with these data to know how the exam
might work as a moderating mechanism. It might cause teachers and students
to work cooperatively toward a common goal, as Coleman (1995) suggested, or
it might also lower students’ expectations around Grades 9 or 10, as Catterall’s
(1989) ﬁndings suggest. The positive association I found between the minimum
competency exam and students’ expectations is consistent with Catterall’s conjec-
ture. If students shift to more realistic expectations, then this could also explain
the strong association of expectations with graduation.4On the other hand, one
would also expect their expectations to be more closely associated with test scores
in that circumstance and exploratory analyses, not shown, do not indicate such a
The two groups of students, those who are and are not subject to the require-
ment, differ in ways that were not accounted for in these models. Efforts to control
other factors (not shown), such as other teachers’ and students’ attitudes or school
poverty level, revealed signiﬁcant associations between the factors and outcomes
but did not change the basic ﬁndings described here. Schools are only one part
of a larger social system which supports adolescent development. It is difﬁcult
to measure the pervasive effects of such factors as poverty on achievement and
conception of opportunity in a study such as this. Certainly, further research in the
area is both possible and important.
As a form of external assessment, the minimum competency exam is especially
weak because it targets only low-level proﬁciency skills and because the negative
consequences fall disproportionately on students. Furthermore, the NELS measure
itself is weak. The actual minimum competency exam requirement varies among
states, thus, students may be subject to different standards and testing procedures
even when they are subject to an exam requirement. Nonetheless, as a policy tool,
the minimum competency exam requirement is widely used, about one-half (51%)
of all NELS public school students are subject to it, and therefore its effects are
important to consider.
The students to whom the minimum competency exam is targeted may be an
especially challenging group to reach. Certainly, some may be graded lower by
teachers for misbehavior, and if they perform well enough on the exam may gradu-
ate despite their teachers’ disapproval. The exam may inﬂuence attainment without
affecting learning. It may be a much more complex problem to raise the levels
of mathematics achievement among low-performing students who are about to
graduate from high school. Additionally, the minimum competency exam could
easily affect students, especially those who are low performing, before the tenth
grade when the students in this analysis were ﬁrst interviewed.
It is unclear whether ﬁndings would be similar for other forms of external as-
sessment. Teachers are clearly sensitive to their students’ performance on standard-
ized tests, as revealed in the analysis of the determinants of teachers’ expectations.
214 CHANDRA MULLER
Muller, Katz, and Dance (in press), using multiple qualitative studies of teach-
ers and students, found that teachers who were subject to more rigorous forms
of external assessment were overwhelmingly concerned with the testing process
and results. Thus, it is possible that broader standards would produce stronger
results. Importantly, they also found that students who were excluded from the
testing process because of special designations were essentially neglected by many
An especially perplexing problem for policy makers has been how to encourage
teachers to take full advantage of student resources for learning. Especially tragic
is the case of students who, though motivated to learn, are taught by teachers who
have low expectations for them and therefore demand less than is needed for the
students to attain their goals. As a tool to inﬂuence teachers, external standards and
assessment in the form of an examination appear to have potential; however, the
nature of the inﬂuence is unclear. Importantly, little is known about the structures
that might support effective standards and assessments or what other resources
might be required for schools to meet the needs of students successfully.
1. This process would operate only for students subject to the assessment. For instance, if certain
students are exempted from an external assessment, as students in special classes might be, then
we would not expect the teacher to play a role of coach even if the broader state policy included
external assessments. In this case, the mechanism used to exclude the student from testing might
have an enhanced negative effect on the student’s opportunity to learn. Measuring this type of
exclusion is not possible with the data I used, although students with “linguistic, mental, or
physical obstacles to participation” were excluded from data collection (Ingels et al., 1994, p.
2. The students with mathematics teacher data had slightly higher achievement test scores and
slightly higher SES than the sample as a whole.
3. Most students were excluded because of missing data on either the graduation status or second
follow-up mathematics achievement test.
4. When high school graduation models are run separately depending on the presence of a mini-
mum competency exam requirement (not shown), students’ expectations are a signiﬁcant pre-
dictor of graduation only when the requirement is in place. In this case, however, the difference
between the coefﬁcient for students’ expectations is not signiﬁcant when students who are
subject to the requirement are compared with those who are not.
Baker, Eva L. (Ed.). (1994). Educational reform through national standardsand assessment [Special
issue]. American Journal of Education,102, 383–580.
Biddle, Bruce J. (1997). Foolishness, dangerous nonsense, and real correlates of state differences in
achievement. Phi Delta Kappan,79(1), 8–13.
Bryk, Anthony S., Lee, Valerie E., & Holland, Peter B. (1993). Catholic schools and the common
good. Cambridge, MA: Harvard.
Catterall, James S. (1989). Standards and school dropouts: A national study of tests required for high
school graduation. American Journal of Education,98, 1–34.
MINIMUM COMPETENCY EXAM REQUIREMENT 215
Cohen, David K. & Haney, Walter (1980). Minimums, competency testing, and social policy. In
R. M. Jaeger & C. K. Tittle (Eds.), Minimum competency achievement testing. Berkeley, CA:
McCutchan, pp. 5–22.
Coleman, James S. (1995). Achievement-oriented school design. In M. T. Hallinan (Ed.), Restruc-
turing schools: Promising practices and policies. New York: Plenum, pp. 11–29.
Coleman, James S. & Hoffer, Thomas (1987). Public and private schools: The impact of communi-
ties. New York: Basic Books.
Coleman, James S., Hoffer, Thomas, & Kilgore, Sally B. (1982). High school achievement: Public,
Catholic, and private schools compared. New York: Basic Books.
Delpit, Lisa (1995). Other people’s children: Cultural conﬂict in the classroom. New York: The New
Drazen, Shelley M. (1994). What school resources inﬂuence achievement? Unpublished manuscript,
Cornell University, Department of Psychology.
Ensminger, Margaret E. & Slusarcick, Anita L. (1992). Paths to high school graduation or dropout:
A longitudinal study of a ﬁrst-grade cohort. Sociology of Education,65, 95–113.
Fisher, Thomas H. (1980). The Florida competency testing program. In R. M.Jaeger & C. K. Tittle
(Eds.), Minimum competency achievement testing. Berkeley, CA: McCutchan, pp. 217–238.
Good, Thomas L. (1993). Teacher expectations. In L. Anderson (Ed.), International encyclopedia of
education (2nd ed.). Oxford: Pergamon.
Good, Thomas L. & Brophy, Jere E. (1997). Looking in classrooms (7th ed.) New York: Longman.
Hanson, Sandra L. (1994). Lost talent: Unrealized educational aspirations and expectations among
U.S. youths. Sociology of Education,67, 159–183.
Ingels, Steven J., Dowd, Kathryn L., Baldridge, John D., Stipe, James L., Bartot, Virginia H.,
Frankel, Martin R., & Quinn, Peggy (1994). National Education Longitudinal Study of 1988
second follow-up: Data ﬁle user’s manual (NCES 94–374). Washington, DC: U.S. Department
Jones, Edward E. (1990). Interpersonal perception. New York: W. H. Freeman.
MacLeod, Jay (1987). Ain’t no makin’ it. Boulder, CO: Westview Press.
Muller, Chandra, Katz, Susan Roberta, & Dance, L. Janelle (in press). Investing in teaching and
learning: Dynamics of the teacher-student relationship from each perspective.Urban Education.
Oakes, Jeannie (1985). Keeping track: How schools structure inequality. New Haven, CT: Yale
Persell, Caroline H. (1977). Education and inequality: The roots and results of stratiﬁcation in
America’s schools. New York: The Free Press.
Porter, Andrew C. (1994). National standards and school improvement in the 1990s: Issues and
promise. American Journal of Education,102, 421–449.
Rist, Ray C. (1970). Student social class and student expectations: The self-fulﬁlling prophecy in
ghetto education. Harvard Educational Review,40, 411–451.
Rosenthal, Robert & Jacobson, Lenore (1968). Pygmalion in the classroom. NewYork: Holt,
Rinehart, & Winston.
Schneider, Barbara & Coleman, James S. (Eds.). (1993). Parents, their children and schools. Boulder,
CO: Westview Press.
Spenner, Kenneth I. & Featherman, David L. (1978). Achievement ambitions. Annual Review of
Williams, Trevor (1975). Educational ambition: Teachers and students. Sociology of Education,48,
Wilson, Patricia M. & Wilson, Jeffrey R. (1992). Environmental inﬂuences on adolescent educational
aspirations: A logistic transform model. Youth & Society,24, 52–70.
216 CHANDRA MULLER
Chandra Muller is an assistant professor of sociology at the University of Texas at Austin. Her
research interests include parental involvement, external assessments as a form of incentive, mathe-
matics achievement, and the transition into higher education.