The last two days, the television media has been dominated by the story of Dean’s possible comeback in NH. The impetus for these stories was the Zogby Tracking polls which showed Dean pulling even with Kerry, statistically speaking. The problem that the big media didn’t seem to want to address was that several other tracking polls had very different forcasts. Almost all had Kerry’s margin of victory at +10 percentage points. The Boston Globe has Kerry up by 17% and the ARG has had Kerry with as much as an 18% lead over Dean as recent as yesterday. The reason for this huge discrepancy is that the Zogby poll that everyone has been citing factored in “leaners”, i.e. of the undecided who leans where. One the one hand this methodology can help to forcast where the undecided are going to fall in the critical moment, but on the other there is a certain risk in assuming that these “undecided” can describe which way they are truly leaning with any sort of accuracy.

Its important to note that Zogby’s latest tracking numbers do not factor in leaners and thus reflect similar percentages to the other tracking polls. This latest poll shows Kerry leading with 37% and Dean in second with 24%. That’s a pretty large margin.

The outcome of today’s contest will tell us something about the methodology of factoring in leaners and also about the ability of big media to take differring methodologies into account when reporting on polls.