Public opinion polls can be done in a variety of ways. There's the standard "live operator" survey, where actual human beings dial up voters. There's the fairly new "automated" phone survey, where everything is done by computer. And there's the Internet, where folks answer on the Web.

The number crunchers at Pollster.com let you see President Barack Obama's approval and disapproval ratings any way you like.

If you generate a typical Pollster.com set of fever lines, which include all three types of polls from dozens of top survey firms, the post-Inauguration Day trends look like this:

 

Filter out the Internet and automated surveys, though, and the gap between Obama's "approval" and "disapproval" ratings widens considerably:

— Approval: 52.1%
— Disapproval: 41.6%.

What does it mean? Many pollsters believe that Internet and automated surveys are less reliable than more old-fashioned polls done using "live" operators. Internet polls, in particular, have their critics because participants are "opting in" — that is, they are pro-actively choosing to participate, which can dramatically affect the make-up of the survey pools. But some, such as the forces behind the Zogby, Rasmussen and Harris Interactive firms, argue that they've had enough years of experience now to know how to effectively "weight" the responses.

Pollster.com's Mark Blumenthal (also known as the "mystery pollster"), by the way, wrote in September that automated phone surveys have proven very reliable during campaigns when it comes to the "horse race" — that is, how candidates are doing at particular moments in time.

Blumenthal says, though, that he "would certainly not recommend an automated interview for any general population study that wants to ask more than four or five substantive questions or that involves open-ended questions that allows respondents to answer in their own words."

He tells us he hopes to have more to say on all this in coming days. We'll report back.