Re: Need hugs
Posted 6/26/2009 6:56:39 AM by glinda-goodwitch


Posted by lipglossjunky73

Chat Icon Chat Icon Chat Icon Chat Icon Chat Icon Chat Icon Chat Icon

Just curious - what type of experiments are you doing that are "unpredictable"? I'm just wondering because in the science of behavior, when research was done on animals, everything was based on how predictable they behaved!!! Chat Icon



I do what is called discrimination experiments. I train the birds to listen to a repeating background of one sound, and they peck one key when they hear that sound. Then target sounds are inserted randomly and they peck a second key when they hear the new sound, or basically the change in sound. But on 30% of the trials the same background sound is inserted. It's a sham trial. If they respond during that trial, it is recorded as a "false alarm". We throw out data that has more than a certain level of false alarm because then we assume the animal is guessing.

so really I don't care what changes they can and cannot hear. I only care that their false alarm is low enough to keep the data. And these birds are giving me high false alarms so I have to throw out the data Chat Icon

It is unpredictable because some of these changes are difficult for them, but on the edge of their ability. So depending on their mood, they will do the task with an okay false alarm level, or some days they will get frustrated and give a bad level and we have to throw the data out. Parakeets are known for being moody animals. The finches usually will just work consistently.
Print Post      Close Window