My score on the Miller Analogies Test lives in infamy in my mind. All graduate programs required GRE scores and one required the Miller as well. My number wasn’t bad–it was above average–but it was much lower than I expected and well below my other test scores. I was a winner of high school and college academic awards, a word nerd, a book worm. I expected to do extremely well on this vocabulary and verbal puzzle test. Take a spin yourself: Foot is to boot as (cough, bout, plough, tree) is to bough *. I figured that one out, but I missed so many others.

Number Use and Abuse

My job requires me to generate number after number. These numbers help me understand why a student reads slowly or can’t remember math facts or excels in spelling but not science. They tell me how a student compares to others her age, and that helps me know whether her low grades are due to a problem with skills or because of something else. They help parents decide if tutoring is needed, if an advanced class would be too much of a stretch, if their son would fit in at a school of academic superstars.

Numbers can be harmful when they are misused.

They are harmful when they are wrongly used as shorthand for “competent” or “incapable: “Brandon has a 98 IQ so I doubt he’ll be interested in science camp.” “Anita got a 31 on the ACT and I know she’ll be a great class president.” They are harmful when they are used as descriptions of people rather than scores on a test. No one “is” a 107 IQ.

Numbers are only a part of the picture.

Consider two students. One has average scores and is focused, motivated, and organized. The other has above average scores and is distractible, indifferent, and forgetful. Which one is likely to do better with an advanced curriculum and a large amount of homework?

Snapshots, Not Portraits

Any single score is a snapshot at a specific point in time.

It takes many snapshots/scores to “make a movie” of a student’s learning profile. Think about photos of yourself. Sometimes you think, “Wow, that picture really flatters me.” Other times you say, “that looks just like me” or “that doesn’t look like me at all.” I would only have a good idea of what you usually look like if I saw photos taken under varying conditions. In the same way, I give many tests to get a variety of “snapshots.”

When Numbers Lie

When numbers don’t describe or predict performance, there is something wrong with the test or the testing situation.

Too often the focus is on the number, not on what the number is supposed to indicate. One of my friends is a scientist with a Ph.D. He has many publications and has been honored for his research skills. Yet he confided that he didn’t want to take an IQ test because he feared he wouldn’t do very well. I was puzzled. If it was pouring outside, he wouldn’t leave his umbrella behind because the forecast said there was only a 10 percent chance of rain. He would look at the actual weather, not the prediction of the weather. Why worry about his score on an IQ test? What matters are his actual accomplishments, not some number that is supposed to predict his accomplishments.

An attorney I was working with wanted some firsthand experience of the testing process. To my surprise, she didn’t do very well on a verbal fluency test. Her poor performance upset her immensely. I explained to her that the test result wasn’t useful at all since she was clearly very articulate and successful. Maybe the unfamiliarity of the test made her mentally “freeze.” Maybe her brain was fuzzy because she didn’t sleep well the night before. Maybe the test is just not useful for people without brain impairments. She was unpersuaded and disheartened. During a phone conversation weeks later she told me she had figured out strategies she should have used and asked if she could take the test again. I eventually “talked her down” by pointing out the many ways she showed verbal fluency skills every day. Her skills were good; the test was a bad source of information.

Numbers Need Context

The attorney’s experience did help her understand that her client’s low scores had to be put into context to make sense. One low score in a sea of high scores or a low score that doesn’t match up to real life performance is an “outlier”—a fluke. Outliers can happen because of “natural variability” or because of a problem with the test. More numbers mean a higher chance for outliers. If you weigh yourself every day you have experienced this. Sometimes you weigh two or even three pounds more than the day before. The next day your weight usually returns to near your baseline. Sometimes you can blame a very salty restaurant meal but other times the gain is a mystery.

When a number makes no sense, first check out the source of the number.

If a physician sees a body temperature reading of 90 degrees in a healthy, active person, she doesn’t check the patient into the hospital—she checks the thermometer. If a fastball registers 175 miles per hour the pitching coach doesn’t call up an agent—he sends the speedgun out for repair.

Some numbers are just not meaningful.

After evaluating literally thousands of students, I know which tests I give are more likely to result in “flukey” scores. I can quickly decide whether to investigate a score or shrug it off. There is usually no need to worry about a single outlier. I will never know why my Millers Analogies Test score was so much lower than my other test scores. Maybe if I had taken it on a different day I would have gotten a different score. Maybe my pencil marks were too light. What matters is that it really didn’t matter in the long run.

Even accurate, reliable scores need a broader context.

Once numbers have served their purpose, they should be discarded. What will this test score mean in ten years? In 20 years? None of my clients have quizzed me about my Millers Analogy or Graduate Record Exam scores. Think about a family member or friend you admire and professionals you trust. Do you know how fast your cousin reads? Do you know your best friend’s ACT score? Do you know your physician’s Medical College Admission Test score?

“Use Your Words”

Sometimes parents or adult clients ask me to delete scores from the report I send them. They worry that the numbers will be more of a worry and a distraction than a help. They don’t want to attach numbers to their child. I am happy to oblige. I do have to send a full report with scores to a school or agency if the client needs services or accommodations. But whenever possible I take the advice we give to toddlers and “use my words.” It’s usually enough for parents to understand that their son reads “much more slowly” than most students his age or that their daughter’s math skills are “extremely advanced for her age.” I would have been only briefly disappointed if told that my Millers score was “a bit above average.” I doubt I would even remember taking that test if not for the 52. 52. 52.

What Really Matters

The academic and thinking skills that standardized tests are supposed to measure are important. These skills help us understand the world, make good decisions, and meet our goals. They usually operate in the background. No one says, “I just love Malik. He can solve analogies at the 90^th percentile level!” or “You should avoid Sofia. Her SAT scores were below average.” When describing people we think about temperament, sense of humor, reliability, kindness, emotional maturity, wisdom.

As long as your child is in school you will have to cope with the numbers. I have to deal with hundreds of numbers in my professional role and almost as many as the parent of school-aged child. These numbers are simply points of information, not enduring qualities.  My young client is not her reading fluency score. My friend is not her IQ score. I am not my Millers Analogies Test score. I’m 100 percent sure of that.

*”Cough” is the correct answer. Focus on the change in the vowel sound and the repeated letters, not the meanings of the words. If you couldn’t figure this out, you should immediately….roll your eyes then shrug your shoulders.

Talk to me!

Is there a topic you’d like to read about? Email your suggestions to contact@sjunedye.org