Using a canonical tag to eliminate ID variables?

[QuickLearnTraining]

My research on seomoz has resulted in conflicting ideas regarding the canonical tag. One article says avoid it, the other says embrace it.

We have fixed a majority of our architecture problems using redirects for duplicate content, however, when we send out newsletters we still have these pesky tracking ids. I figured out how to remove them from analytics, but am unsure of how this affects our SEO. An example of one of our links is: https://www.quicklearn.com/transcript/?utm_source=news101011&utm_medium=e&utm_campaign=newclass&nlid=news101011&UID=2287

The original url being www.quicklearn.com/transcript/ the custom (non-Google) variables being nlid and uid.

Is this a problem? Do I need rel cononical tags on each and every page?

[AlanMosley]

Are these links on your web site or are they in a newsletter sent out via email?

anyhow, there is nothing wrong with the canonical tag, if its used for the right reasons, but often a 301 redirect is better.

Not understanding your situation fuilly, my advice is.

first, will a 301 do the job?

if not can i use a canonical