Scaled Score Ranges
A Scaled Score Scale can have many different possible ranges. There is not only one range for a scaled score. It can be either a 100 or 10, a 1000 or 3. Any range is possible. However, there is an optional range. Small ranges came with discrimination problems. A scale that ranges from 0 to 4, are usually too small to have a good discrimination. Scales with an "intermediate" score discriminate better, specially in psychological tests.
A 1 to 5 scale is going to discriminate better than a scale 1 to 4, or a 1 to 6 scale. Always choose an odd number for your scale, it helps to discriminate the results way better.
Most of the qualitative descriptors in tests follow a similar pattern. Or at least, they should.
A 1 to 5 scale is going to discriminate better than a scale 1 to 4, or a 1 to 6 scale. Always choose an odd number for your scale, it helps to discriminate the results way better.
Most of the qualitative descriptors in tests follow a similar pattern. Or at least, they should.
|
|
A Scaled Score can have many different ranges. Choose wisely.
|
|
SCALED SCORE Mean
A scale score mean is the average of a group of numbers that have been scaled (for example, the average of all the results of all the students in one same test). A scale score mean is calculated by adding all the scores and dividing the result by the number of total scores.
If you already have the scaled scores it would be easier than if you have raw scores, because you will need to scale them.
If you already have the scaled scores follow these examples:
i.e. A small math class has 5 students. Their scores in the final test are: 100, 85, 75, 90 and 95. What is the mean scaled score?
We will need to add all of them up and divide the result by 5 (because we have 5 students).
100+85+75+90+95 = 445
445/3 = 89
Result: The mean of those scaled scores is 89.
i.e. We have these scaled scores: 10, 15 and 20.
We will need to add all of them up and divide the result by 3 (because we have 3 numbers).
10+15+20 = 45
45/3 = 15
Result: The mean of those scaled scores is 15.
i.e. We have these scaled scores: 5, 10, 20 and 5.
We will need to add all of them up and divide the result by 4 (because we have 4 numbers).
5+10+20+5 = 40
40/4 = 10
Result: The mean of those scaled scores is 10.
If you already have the scaled scores it would be easier than if you have raw scores, because you will need to scale them.
If you already have the scaled scores follow these examples:
i.e. A small math class has 5 students. Their scores in the final test are: 100, 85, 75, 90 and 95. What is the mean scaled score?
We will need to add all of them up and divide the result by 5 (because we have 5 students).
100+85+75+90+95 = 445
445/3 = 89
Result: The mean of those scaled scores is 89.
i.e. We have these scaled scores: 10, 15 and 20.
We will need to add all of them up and divide the result by 3 (because we have 3 numbers).
10+15+20 = 45
45/3 = 15
Result: The mean of those scaled scores is 15.
i.e. We have these scaled scores: 5, 10, 20 and 5.
We will need to add all of them up and divide the result by 4 (because we have 4 numbers).
5+10+20+5 = 40
40/4 = 10
Result: The mean of those scaled scores is 10.
|
|
This is how to calculate a scaled score mean without any effort
|
|
AP Statistics and scaled scores
AP Statistics and scaled score are two important elements for students and the exam. A student's score, which can be determined with an MRS, is sometimes represented as a number on the GRE with the digits indicating the row of the scale. Additionally, the midpoint score is often represented as an ellipse with the bottom left part being the RAS score. I first learned about the use of the statistics while doing work with a tutor for the AP Statistics exam. The scale has come into use more recently and is now being used for more sophisticated purposes such as writing detailed exam questions.
Scaled Score Descriptors
|
Strong descriptors of a scaled score goes as follow:
1-15 Very weak 15-30 Weak 30-45 Low 45-55 Medium 55-70 High 70-85 Strong 85-100 Very strong These scaled score descriptors are based on a random distribution. Other kind of distributions might need different scaled score descriptors. |
|