Normalizing File: EQ vs Reducing dB


I’ve just watched ‘How to Record a Podcast with Adobe Audition’:

The interviewee connects via Skype and you address her spike (timestamp 30:55) using parametric EQ, which takes a few tweaks. I’m sure there is a very valid reason, but why do it this way instead of highlighting the spike, then bring it down on the dB wheel (and then normalizing after doing so)?



The spike contains many frequencies and is only a visual indication of gain or volume level.
A scan with a parametric eq reveals the offending frequency and then it’s just a matter of isolating and reducing the correct one without destroying all the others.