Do You Have to Reverse the Transformation?

You can transform variables, and then now all of a sudden they actually satisfy assumptions. So if you square it, for example, you are simply saying one unit goes up by roote, two units instead of two. You know what i mean? Because the power of transformation is that sometimes even your x variable can be transformed. And again, the relationship holds. It's just that when you transform them, you can apply linear regression, and then you can analyze. But hold onu without that, you can't analyze. The analysis step is broken. Icary.