Thank you all for coming here today. As you know, your children have

taken a lot of standardized tests throughout the years, which were designed to predict

their performance in a certain area. One such test is the prediction for ACT performance.

You should know that these predicting tests are not precise enough to tell you

exactly what your child will score on do on the actual test, but they can give

you a ballpark estimate. These estimates, though not exact, can help students

to figure out their areas of weakness and strength, and help them to better

prepare for future performances in classes.

As a foreign language teacher, there are not very many of these

prediction tests available for me to use. However, there are tests that predict

a student’s language ability and using these, I can determine which students

learning a foreign language may come more naturally to, and those who may need

more hands-on assistance.

The standardized tests which all your children had to take throughout

grade school were designed to measure the effectiveness of teaching in this

school compared with other public schools in the state. The way this test works

is that the students are graded based on their answers, and a cumulative score

is given which is designed to reflect the ability of those students. However,

these tests are not perfectly precise and there exists what is known as

standard error of measurement.

You may have heard this term in the news talking about political polls

or surveys, but you may not have known what it meant. A textbook definition for

“standard error of measurement” is “An estimate of the consistency of an

individual’s test performance(s) represented by an index of how likely it is

that another test performance … would yield a comparable score” (Popham, 2017,

p.410). This means that it reflects the consistency of your child’s score on a

certain test if this test were taken by your child several times (Popham, 2017).

Since this scenario is very unlikely, the standard error of measurement helps

us helps us to estimate how much the score would vary if a student would take a

certain test several times (Popham, 2017). We can never really know the exact

score of your child but only get an estimate, and the standard error of

measurement helps us estimate the true range of a student’s score (Jensen, 2015).

For example, your child scores a 28 on an ACT predictive test; however,

the standard error for this test may be 3 points, meaning that the student may

actually score anywhere between 5 and 31 based on this test. Generally, we can

say that the smaller this range is, the more precise assessment we have (Jensen,

2015). Despite the discrepancy of the actual score and the predicted score,

these tests are still a useful tool for both students and teachers alike and

when repeatedly tested using these tests, the range may be decreased. For

example, your child who scored a 28 on the predictive ACT exam with a standard

error of measurement of plus/minus 3, when repeatedly tested using the predictive

exam, later scored a 30 and a 29, may narrow the range to between 27 and 31

because repeated performance maintains the student in this range. Further, a

student who takes a practice act exam will know which subjects he or she is

weakest at and strongest at and can apply this knowledge to their preparation

for the next exam and increase their score based on the test of the predictive

exam.

Let me give you another example to help you grasps the concept of the

standard error of measurement. Let’s say a student takes an IQ test and scores

a 124. We may be tempted to think that

this student actually has an IQ of 124 and is very intelligent, but the student’s

actual IQ may be lower or higher than what was predicted on the exam, depending

on the standard error of measurement of the assessment. If the standard of error

was only one point, 124 would be an accurate assumption of the student’s IQ. However,

if the standard of error was 10 points, the students actual IQ may be much different

than the test reflected (Popham, 2017).

Therefore, because of these standards of errors, we

should not rely these predictive tests as facts and see them as an exact

science, but rather as tools which can be used to assist students learning and

teachers teaching abilities.