Tuesday, October 13, 2009

Percent defoliation results

As i said in last weeks blog, this week i would be reviewing the results from a percent defoliation exercise we did two weeks ago. We were asked to look at 60 leaves and estimate the percent defoliation of each leaf. Afterward the actual percent defoliation for each was calculated and then our "guesses" were compared to that. So how did i/we do? Well i have posted the results from lab below-








Denton won the prize for being closest to the actual amount of defoliation that had occurred. This can be determined by looking at the Regression equation. If someone was to guess all the percent defoliation correct there equation would be equal to 1, or there would be no deviation away from the one to one line (what was actual vs. what was guessed). For all samplers involved it appears we fell into the range of R squared values from .67 to .84(Denton) which sounds fairly good to me. Overall if i read into the statistics well enough the class tended to overestimate the amount of defoliation present on the leaves, which is what i did (y=1.5264x). And overestimating seems to be the trend of the class but we did have a few participants that underestimated the percent defoliation.
Posted by at

1 comment:

  1. b.mccornack (K-State)November 2, 2009 at 1:50 PM

    You're correct, the class did tend to overestimate defoliation, with the exception of a couple of people. However, at least students were consistent with their estimations, which is the reason for the high r-square values.

    How could you adjust for the overestimation in the field? In general, is it better to over or under estimate defoliation or insect damage?

    ReplyDelete

Subscribe to: Post Comments (Atom)