A/B test in the moderated User Testing

Just wondering if it's common and practical to show our participants another version of our design to get their thoughts? My concern is validating version A without observing version B wouldn't be a useful approach.