Comments on: 5 Smartest Animals In The World http://www.environmentalgraffiti.com/featured/are-these-the-smartest-animals-in-the-world/1777 for environmentalists who don't take themselves too seriously Fri, 27 Nov 2009 14:41:33 +0000 http://wordpress.org/?v=2.8.6 hourly 1 By: Dale http://www.environmentalgraffiti.com/featured/are-these-the-smartest-animals-in-the-world/1777/comment-page-1#comment-264607 Dale Sun, 13 Sep 2009 10:45:52 +0000 http://www.environmentalgraffiti.com/?p=1777#comment-264607 Hans was also able to do higher math, including sums and products. He would also give the correct answer if the question were posed by someone other than his master, and if no-one were tapping on the board. When the case was finally "cracked," it turned out that "der kluge Hans" was responding to cues that people gave as his tapping approached the right answer ... it involved leaning slightly forward and looking in general more attentive. These are natural cues that almost everyone gives, so it didn't have to be someone the horse knew. The explanation came from the observation that Hans could not give the correct answer if there were no people at all in his line of vision. It was not a fraud worked by von Osten; he was as convinced as anybody that he had managed to teach his horse arithmetic. This case is taught in statistics and experimental design courses as the "Clever Hans Artefact." If a research experiment involves a subjective endpoint, especially a behavioral one, the subject may respond to cues from other people in a manner that has not been anticipated or controlled for. This may lead to an erroneous conclusion. For Clever Hans, it was perceived as "This horse can do arithmetic," when the truth was "This horse is really good at picking up non-verbal cues from people." In the study I use to illustrate this point, the reported observation is "This diet makes ADHD better," when an equally plausible explanation is "Parental expectations and associated cues influenced behavior during the experimental period." It can actually be a VERY difficult bias to exclude. Hans was also able to do higher math, including sums and products. He would also give the correct answer if the question were posed by someone other than his master, and if no-one were tapping on the board. When the case was finally “cracked,” it turned out that “der kluge Hans” was responding to cues that people gave as his tapping approached the right answer … it involved leaning slightly forward and looking in general more attentive. These are natural cues that almost everyone gives, so it didn’t have to be someone the horse knew. The explanation came from the observation that Hans could not give the correct answer if there were no people at all in his line of vision. It was not a fraud worked by von Osten; he was as convinced as anybody that he had managed to teach his horse arithmetic.

This case is taught in statistics and experimental design courses as the “Clever Hans Artefact.” If a research experiment involves a subjective endpoint, especially a behavioral one, the subject may respond to cues from other people in a manner that has not been anticipated or controlled for. This may lead to an erroneous conclusion. For Clever Hans, it was perceived as “This horse can do arithmetic,” when the truth was “This horse is really good at picking up non-verbal cues from people.”

In the study I use to illustrate this point, the reported observation is “This diet makes ADHD better,” when an equally plausible explanation is “Parental expectations and associated cues influenced behavior during the experimental period.” It can actually be a VERY difficult bias to exclude.

]]> By: Joelax02 http://www.environmentalgraffiti.com/featured/are-these-the-smartest-animals-in-the-world/1777/comment-page-1#comment-84690 Joelax02 Tue, 26 Aug 2008 18:32:10 +0000 http://www.environmentalgraffiti.com/?p=1777#comment-84690 What about dolphins? What about dolphins?

]]>