It is generally good advice to avoid using any numerical value to avoid inadvertent use of that value in any analysis. Depending on what format you are doing this in (paper vs electronic. If electronic, which programming language is used) there would be different mechanisms to handle.

A checkbox also ensure a positive record when 'not performed' avoiding trusting that all incomplete answers were not performed (e.g. the question may have been inadvertently missed but the ). This means that blank answers would be captured as 'missing' and differentiated from 'not performed'. This is important as choosing not to perform something is a direct consequence of choices based on various factors and so it has a meaning.

That depends on how you are planning to analyze the scale.

If you are calculating sum scores by adding together several items, you should not include such 0-codings, since the persons answering 0 would count as 'very poor'.

Best option would be to have a checkbox. Then you can us this to select only the persons which actually performed.

It is generally good advice to avoid using any numerical value to avoid inadvertent use of that value in any analysis. Depending on what format you are doing this in (paper vs electronic. If electronic, which programming language is used) there would be different mechanisms to handle.

A checkbox also ensure a positive record when 'not performed' avoiding trusting that all incomplete answers were not performed (e.g. the question may have been inadvertently missed but the ). This means that blank answers would be captured as 'missing' and differentiated from 'not performed'. This is important as choosing not to perform something is a direct consequence of choices based on various factors and so it has a meaning.

If the individual did not perform the test, why would this non-performance be part of any scale. Shouldn't the individual simply be excluded from the analysis?

Dennis

Dennis Mazur