The Chicago Tribune/WGN-TV conducted a few polls recently.  First, they did a polling statewide of registered voters regarding the Kirk/Giannoulias race finding that it is a statistical tie at 34-34 with 22% undecided.  There is some troubling with this polling in my opinion.  First, the Trib/WGN neglected to release their crosstabs, methodology, or polling service used.  Second, a paragraph from the Kirk/Giannoulias article makes me question the "statewide polling":

Nearly four in 10 collar-county voters still had no opinion on whether to view Kirk favorably or unfavorably, despite near universal name recognition. A similar share of conservative voters, 36 percent, have no opinion on whether they like the GOP candidate, and one in five are undecided in the Senate contest. That could leave an opening for the Libertarian candidate to pick up support.  

One, while it's good that the polling is tracking geographical differences between the voters, it makes me question whether or not the poll conducted was truly statewide.  Did the polling of the 600 registered voters question 200 Chicago, 200 collar-county, and 200 "downstate" voters?  "Nearly 4 in 10" suggests there was enough polled in the collar counties to make an assumption about the electorate and their voting preferences.  Two, the number of undecided voters are assumed to be so disaffected by Kirk that they will vote 3rd party.  You have to love the Trib's speculative journalism here because they don't make the same assumption in print that disaffected Giannoulias voters will vote Green Party.  

The Brady/Quinn poll has its problems as well.  According to the polling of 600 registered voters, Brady's lead is at 5% over Quinn 37-32 with 19% undecided.  Maybe it's editorial oversight, but there is no mention of this poll being a statewide poll.  Furthermore, there is no indication of geographical, i.e. mentioning a breakdown of collar county voters as in the Kirk/Giannoulias poll raising my suspicions.  In addition, the Trib neglects to include the crosstabs, methodology, and polling service used for the Brady/Quinn poll. 

Be careful when you read polls because even though it is not an exact science, there are those polling agencies like Rasmussen and Public Policy Polling (PPP) that are very experienced, very careful and very specific on many of the things that the Trib/WGN left out.  The reason these places include their crosstabs, methodology and polling service used is because if you don't, you run the risk of having your data called out and questioned and it gives people a chance to examine the legitimacy of the polling.  Just ask the now discredited (because they were inventing data) R2000/DailyKos poll how that works out.  There will be a new round of polling soon from both of these pollsters in which at least one of them, Rasmussen, will start including "leaners" in the lead tallies after Labor Day.  "Leaners" are those polled who will not pick a candidate out right but indicate they are leaning towards one candidate or another.  Even with "leaners" factored in, Brady's lead was still 8 points in the last Rasmussen polling.