There's no word yet on whether they can also comprehend lies or damned lies, but infants are at least able to understand statistics. As long as there are lollipops involved.
A quick test has shown that babies as young as twelve to fourteen months can understand probability, and use it to their advantage. All it took was a couple of tests to figure out what infants knew, and what they could do, if they were properly motivated.
The researchers started off by performing a taste test. Kids were given two different lollipops, one black and one pink, to see which one they preferred. When a preference was noted, the kids were brought back for a second test. This one was a test of their tiny, soft-skulled minds.
The children were shown two jars. One jar had a lot of black lollipops and only a few pink lollipops. The other had loads of pink lollipops with only a few black ones mixed in. Researchers then picked a lollipop from each jar, in view of the infants, but with the flavor of the chosen lollipop covered. They then dropped the candies into two separate opaque cups. The kids couldn't see which flavor had been picked of the jar. When given a choice between the two cups, seventy-eight percent of the kids chose the cup that contained the lollipop that came from the more 'likely' jar. This indicates that many kids, even when they're very young, are able to make the mental connection that a random lollipop picked from a jar that had more pink lollipops, is more likely to be pink than one picked from a jar containing mostly black lollipops. This ability to understand the statistics of large sets was, up until recently, not believed possible for young kids.
The things that can be done when candy is involved.