Your suspicions are correct, you did not read the spoon-fed link correctly (no surprise!!).
Time for another aeroplaneeee...open wideeee!!
Only 13 participants reported having gone to a rally or meeting at any point in the study; therefore, we did not attempt to analyze attendance as a separate dependent variable. Instead, we reasoned that having attended a meeting or rally represents the highest level of willingness to participate in the Tea Party, and thus retained 6 as the maximum score possible on the item. The support and participation willingness items formed a reliable scale at Time 1 (α = .91), Time 2 (α = .92), and Time 3 (α = .92). For the analyses, scores on these two items were transformed such that both ranged from 0 to 4, and the mean was calculated for each time point. If the Tea Party support item is dichotomized around the midpoint to create supporter and nonsupporter categories, we find that 25% of our participants were Tea Party supporters. This is compared to 18% in a representative sample collected by the New York Times in early April 2010 [8].
So, I would guess....and mind you I am no expert on such things.....that the "13" that were considered "highest willingness to participate in the TP".....were part of the 25% of the 250 that finished the survey (which would be @ 63...but again, I am not an expert on such advanced calculations).
Now I suppose the next argument is going to be the one about how statistical analysis of data sets is not perfect...YADDA YADDA YADDA......and as we have seen, I'm absolutely POSITIVE that your views on such matters is TOP NOTCH......considering how you were able to find the data IN THE FIRST PLACE.....and further backed up by your IMPRESSIVE MATH SKILLS.
But then again.....I might be wrong....because....you know....I am not the expert here.
Oh...ps..."boo-yeah".....or something like that....whateverthatmeans.