______________________________________________________

Compared with the complexity of the standard score scales, there is an intriguing simplicity in the definition of the grade equivalent scale.

_____________________________________________________________________

back to norms menu          next screen
```

```

_____________________________________________________________________

A student correctly answers 35 questions on a reading test. The raw score of 35 is converted to a grade equivalent of 4-5. This "means" that the student made the score of the average 4th grader in the 5th month of the school year. Sounds good, doesn't it.

Remember back when one of your friends knew someone whom you just had to meet and date because (s)he was "just your type". Remember the glowing description provided by your friend. And remember the major difference between the description and the reality.

Grade equivalents are like that. They sound like an appropriate score, especially for use with achievement tests. They sound as if they would be much easier to interpret than the other norm-based scores.

They are NOT!

_____________________________________________________________________

back to norms menu          next screen
```

```

_____________________________________________________________________

The basic rule for the grade equivalent scales is simple. DON'T USE THEM!! Grade equivalent scores are more easily and more often misinterpreted than any of the other converted scores.

The reason is that the grade equivalent score is usually not what it appears to be. Test publishers are unlikely to conduct norming studies during each month of the school year. Often, normative data are gathered only twice in a school year, once in the fall and once in the spring.

When norms were gathered only twice a year, the publisher will only know what the average performance is at any grade level in only two of the months. In the example on the preceding screen, it is very unlikely that the publisher can provide real data about the average performance of a 4th grader in the 5th month of the school year.

The publisher will probably have norms for the 2nd month or 3rd month and for the 7th or 8th month in each grade. The conversions for the months in between come from a process called INTERPOLATION. Given the average performance in the 2nd or 3rd month and the average performance in the 7th or 8th month, the publisher then guesses what the average probably was for the months in between.

_____________________________________________________________________

back to norms menu          next screen
```

```

_____________________________________________________________________

Interpolation causes some problems in the interpretation of grade equivalent scores. But an even greater problem comes from the process of EXTRAPOLATION. There is seldom a different standardized achievement test battery for each grade level. More often there will be five test levels, typically designed for use in grades 1-2, 2-4, 4-6, 6-9, and 9-12.

A fifth grade student takes the mathematics test designed for use in grades 4-6 and does quite well. The grade equivalent score is 10-6.

Does this mean the fifth grader could do tenth grade mathematics? Unlikely!!

Does this mean that the fifth grader made the average score of a 10th grader in the 6th month? Students in the tenth grade don't take the test completed by this fifth grader. So how would anyone know what their average would be?

Extrapolation is a process in which the test publisher draws curves up and down from the actual average scores and again "guesses" what the performance might be at various grade levels.

_____________________________________________________________________

back to norms menu          next screen
```

```

_____________________________________________________________________

Almost all grade equivalent conversions come from interpolation or extrapolation. At best the grade equivalent score communicates whether the student is above, at, or below the average performance for the grade level.

That information is also provided by percentiles, stanines, and the standard score scales without as much danger of misinterpretation.

_____________________________________________________________________

back to norms menu          next screen
```

```

_____________________________________________________________________

REVIEW QUESTION

a. above average

b. below average

c. cannot tell from the information provided

_____________________________________________________________________

```

```

_____________________________________________________________________

REVIEW QUESTION

Interpolation and extrapolation are used extensively to create:

a. DIQ scores

c.. z scores

_____________________________________________________________________

```

```

_____________________________________________________________________

If you missed either of these questions, you may, at some time, want to review the grade equivalent file.

_____________________________________________________________________

```

```

_____________________________________________________________________

YOU ARE CORRECT.

_____________________________________________________________________

back to norms menu          next question
```

```

_____________________________________________________________________

No, the correct answer is "c. cannot tell from the information provided.

_____________________________________________________________________

back to norms menu          next question
```

```

_____________________________________________________________________

YOU ARE CORRECT.

Only the grade equivalent score is created with extensive use of interpolation and extrapolation.

_____________________________________________________________________

```

```

_____________________________________________________________________

Grade equivalents are the only scores listed which make extensive use of interpolation and extrapolation.

_____________________________________________________________________

```

```