Some people have trouble understanding the low signal-to-noise ratio over the short term period of a decade or so during the current climate change. While the loudest ones shouting about the currently-cold January likely already understand the issue but don't care about the truth, some people can be persuaded. For these people, baseball is a good analogy - even a casual baseball fan understands in a single game, the worse team sometimes beats the better one, and only in the long run do you get to know the truth.
So comparing a recent time period to an older one is like looking at a matchup between two baseball teams, the Hot Sox and the Cold Sox. The Global Warming Baseball Analogy is that the Hot Sox has been gradually improving the quality of its players relative to the Cold Sox, so Hot is tending to win against Cold. Comparing a single matchup of a recent year to an older one is just like looking at a single baseball game - one game alone has only a slight chance of reflecting which is the better team. And to compare one month like January to the past is like judging two baseball teams based on a single inning. It doesn't prove anything.
And anyway, as Deltoid points out in the link above, January 2007 was the warmest January ever. The same people screaming about the drop relative to January 2008 usually scream about 1998 being warmer than recent years (according to some instrumental records). However, January 2007 was warmer than January 1998. If the skeptics deny this is important, then it can only be because comparing two January months is insufficient to prove anything.
I'd agree with them, so then we can quit the useless comparison to January 2008, and I'd extend it to saying comparing 1998 to 2008 is almost as useless. Multiyear averaging gives the validity of longer-term stats that a baseball fan would understand. Five-year averaging (graph at the Deltoid link) has been going upward with only slight interruptions ever since the 1970s, and 1998 didn't change that upward trend.