ERIC Identifier: ED481472
Publication Date: 20030900
Author: Monetti, David M.; Hinkle, Kerry T
Source: ERIC Clearinghouse on Assessment and Evaluation
Five Important Test Interpretation Skills for School
Counselors. ERIC Digest.
School counselors are often asked to administer and interpret normreferenced
tests. Certain fundamental test interpretation skills are necessary to
accurately interpret and utilize test data. The purpose of this Digest
is to outline five skills that will increase the likelihood that test information
is interpreted correctly.
"SKILL 1: UNDERSTANDING WHAT NORMREFERENCED TESTS ARE AND WHAT THEY
DO"
Normreferenced tests are assessments administered to students to determine
how
they perform in comparison to others. They are often used to classify
students for
placement and award purposes (Bond, 1996). One wellknown example is
the use of the PSAT test for determining participation in the National
Merit Scholarship award program. A student's current test performance is
compared to that of a representative sample of students, known as a norm
group, who were previously administered the test.
Norm groups can be used to create either national norms or local norms,
depending on who is included in the normative sample.
Normreferenced tests have several strengths. For example, Dombrowski
(2003) points out that normreferenced testing often allows for reliable
and objective measurement.
However, with these types of assessments, it is critical to understand
the composition of the norm group (Oosterhof, 2003). It's also important
to note that test scores on
normreferenced tests typically rise the longer the test is in use,
likely due to changes in instruction or test preparation that are made
as educators become more familiar with the form of a test (Linn et al.,
1990).
Criterionreferenced and normreferenced tests yield different, but
complementary,
pieces of information. Criterionreferenced tests, such as many of
the state high school tests, help demonstrate how a student stands in relation
to a given educational curriculum. The emphasis is not on comparison with
other students, but rather, on mastery of specific content knowledge and
skills. Criterionreferenced tests provide useful information about students'
strengths and weaknesses in various curriculum areas. It is critical that
the content domain of the assessment be clearly defined (Oosterhof, 2003).
Results of norm and criterionreferenced tests should be combined with
other formal and informal data collection methods, since no single set
of test scores is adequate to make important educational decisions (Code
of Fair Testing Practices in Education, 1988).
"SKILL 2: UNDERSTANDING THE PROPERTIES OF THE NORMAL CURVE"
As previously discussed, in normreferenced tests, a child's test performance
is
compared to a norm group. The distribution of test scores generated
by the norm group is normal (Popham, 2002). Therefore, understanding what
a normal curve is becomes critical for normreferenced test interpretation.
First, a school counselor should recognize that if he or she were to draw
a vertical line down the center of the normal curve, the distribution would
be divided into two equal halves. Because both halves are identical, statisticians
classify normal curves as symmetrical. The vertical line represents the
mean, or average, performance of the norm group. With normal curves, this
line also represents the median and the mode; however, this is not always
true of other types of curves.
Knowing the mean helps the test interpreter to identify the average
performance of the norm group, but is not sufficient to correctly interpret
an individual's performance on a normreferenced test. The school counselor
also must know how the concept of standard deviation is related to the
normal curve. Drummond (2000) indicates that standard deviation is a statistic
that defines the spread of scores around the mean.
Standard deviation helps determine how far above or below the norm group
mean an individual's score falls. For practical purposes, the normal curve
is divided into three standard deviations above the mean and three standard
deviations below the mean. In a normal curve, 34% of individuals fall between
the mean and one standard deviation above the mean, 14% of individuals
fall between one standard deviation above the mean and two standard deviations
above the mean, and 2% of individual fall between two standard deviations
above the mean and three standard deviations above the mean. Since the
normal curve is symmetrical, the percentages are the same for the standard
deviations above and below the mean.
"SKILL 3: KNOWING THE PROPERTIES OF COMMON SCORE TYPES"
To accurately and efficiently interpret normreferenced assessments,
school counselors need to be familiar with the properties of common scores
they may encounter. Those score types are Z scores, T scores, NCE scores,
and stanine scores.
* Z scores have a fixed mean of 0 and a standard deviation of 1. Thus,
if a student's test performance is one standard deviation above the mean,
the individual has a Z score of 1. If a student's test score is two standard
deviations below the mean, his or her Z score is a 2. It is possible(and
common) for students to have negative Z scores.
* T scores often have a mean of 50 and a standard deviation of 10. Thus,
if a student's test performance is one standard deviation below the mean,
the individual has a T score of 40. If a student's test score is at the
mean, his or her T score is 50.
* NCE scores stands for normal curve equivalent scores. NCE scores have
a fixed
mean of 50 and a standard deviation of approximately 21. If a student's
test
performance is two standard deviations above the mean, the individual
has an NCE
score of 92. If a student's test score is one standard deviation below
the mean, his or her NCE score is a 29.
* Stanine scores have a fixed mean of 5 and a standard deviation of
2. The term
"stanine" stands for Standard Nine, indicating that the range of stanine
scores is fixed
from 1 to 9, with 9 representing the highest possible stanine score,
and 1 representing the possible score. If a student's test performance
is two standard deviations above the mean, the individual has a stanine
score of 9. If a student's test score is three standard deviations below
the mean, his or her stanine score is a 1. Five is the most commonly assigned
stanine score, because it falls directly on the mean the curve. If a student
is assigned a 5, he or she is performing better than half of the norm group
on the content assessed on the normreferenced test.
"SKILL 4: RECOGNIZING THE DIFFERENCE BETWEEN PERCENT AND PERCENTILE"
School counselors need to recognize that percent and percentile are
different concepts.
The term "percent" is an abbreviation of the Latin phrase per centum,
which literally
means "by the hundred." A percent represents the proportion of test
material answered correctly out of a hundred. For example, if an individual
took a 50item test and answered 25 items correctly, the percent he or
she got correct would be 50.
Percentiles, according to Drummond (2000), are one of most common tools
to help
interpret normreferenced assessments. Percentile scores range from
1 to 99 and tell
the test interpreter the percentage of individuals in the normgroup
that the test taker
outperformed. For example, if a test taker earned a score in the 74th
percentile, the
interpretation would be that 74% of the norm group performed at or
below the test
taker's score.
"SKILL 5: BEING ABLE TO TRANSLATE FROM ONE STANDARD SCORE TO ANOTHER"
Once school counselors have developed competency with the four skills
addressed
above, they will have the tools necessary to quickly and accurately
translate scores
from one common score type to another. This skill is particularly important
when
meeting with parents and students to discuss their normreferenced
test performance. School counselors need to be cognizant that parents and
students are often confused by the many different score types, the compact
layout of many normreference score reporting sheets, and most importantly,
the interpretation of the scores. This is an opportunity for the informed
school counselor to be particularly helpful.
Because the score types discussed above (Z, T, NCE, stanine) are all
based on the
properties of the normal curve, the scores can easily be converted
from one score type to another. By performing test score conversions, the
school counselor can demonstrate to the parent(s) and student how the different
score types are representative of how the test taker performed in comparison
with the norm group. For example, if a student had a Z score of 1, the
individual's performance is one standard deviation above the mean.
Eightyfour percent of the norm group performed at or below the test
taker's score. Thestudent also had a T score of 60, an NCE score of 71,
and a stanine score of 7. All of the scores are one standard deviation
above the mean. The same is true for students who have scores that fall
below the mean. If a student's normreferenced test score is two standard
deviations below the mean, the corresponding Z score is 2, the T score
is 30, the NCE score is 8, and the stanine is 1. Table 1 will help translate
four common standard score types.
Due to the increased reliance on normreferenced tests in schools, it
is essential that
school counselors be able to accurately interpret and explain test
results to various
stakeholders. While these five test interpretation skills do not guarantee
expertise, they are intended to encourage school counselors' minimum competency
with regard to normreferenced test interpretation.
REFERENCES
Bond, L. A. (1996). Norm and criterionreferenced testing. Practical
Assessment,
Research & Evaluation, 5(2).
Code of Fair Testing Practices in Education. (1988) Washington, DC:
Joint Committee on Testing Practices.
Dombrowski, S. C. (2003). Normreferenced versus curriculumbased assessment:
A balanced perspective. Communique, 31 (7): 1620.
Drummond, R. J. (2000). Appraisal procedures for counselors and helping
professionals (4th ed.). Upper Saddle River, NJ: Merrill Prentice Hall.
Linn, R. L., Graue, M. E., & Sanders, N. M. (1990). Comparing state
and district results to national norms: The validity of the claims that
"everyone is above average."
Educational Measurement: Issues and Practice, 9 (3): 514.
Oosterhof, A. (2003). Developing and using classroom assessments (3rd
ed.). Upper Saddle River, NJ: Merrill Prentice Hall.
Popham, W. J. (2002). Classroom assessment: What teachers need to know
(3rd ed.). Boston, MA: Allyn and Bacon.
