Scaling methods available in iWise

Linear

This scaling method allows you to set a new top and bottom mark. The original marks will then be remapped to fit into this range. You should check you are happy with the resulting average and standard deviation when using this method.

Average and Top

This method is similar to Linear Scaling, but you now control the new average and top mark. Be careful that you don't force the bottom marks too low when using this method.

Multipoint

This method is like performing a series of linear mappings between midpoints in your set of marks. You can give up to 6 mapping points. All marks between the minimum and the first mapping point will be linearly scaled to the new minimum and mapping point. This process is then repeated for each mapping point. This method is nice in that you can control both the new top and bottom, but also have control over the new average and standard deviation. NB: You can only scale one task using this method and can’t use it to combine tasks. The reason for this is that each component task will have a different spread of marks, so the one set of mapping points will only make valid sense to the one source task. If you want to use this method on a combined tasks first combine the component tasks using  ‘Out of’ scaling and then scale the combined task with the multipoint scaling.

Add Tests Together (no weighting)

Use this method if you just want to add the marks of one or more tasks together. NOTE: if you use this method the new out of mark is also the combined total of the constituent tasks. This method will treat all weightings as the same, irrespective of the number entered. You would usually only use this method if you did something like split an exam up for marking and enter the section marks as separate columns and then wanted to re-combine the parts to get the final mark for the exam.

Out of

If you combine more than one test together using this method, the final result be set to the new 'out of' value. If, for example, you have a test out of 54, and a test out of 100 and want to combine them to be a test that is out of 80, use this scaling method. NOTE - this test will not moderate the tests before combining them. Use parametric, linear or Mean & SD to moderate and combine tests.

Parametric

This scaling method allows you to set the new bottom, top and average. Basically, it uses a parabola to map the test scores. One warning with this scaling method - if you have a small number of students in the test, or you set the average inappropriately; you can end up with the top student no longer receiving the highest mark.

Mean and Standard Deviation

Use this method to set a new average and new standard deviation for your test marks. The standard deviation is a measure of how spread your marks are. 12.5% is a typical standard deviation (eg, if you have a test out of 50, a standard deviation of 6 would be appropriate).