Limited data July 29, 2006Posted by Michael McVey in Online Tools.
I am convinced the major difference between scientists and non-scientists is the way they deal with the data presented to them. The other day, a colleague wanted to know if it was going to rain in the late afternoon. He called up a radar image of storms in the area and saw several large cells about 100 miles north. He asked for my opinion.
One of the first things you learn in the scientific world, and it begins with basic geometry, is that a single point is not a line. You get two points and you have a line segment. You get three points and you have a pattern you can use to project with some certainty into the future.
Returning to that radar image we were looking at, there was no way to tell if the clouds were heading toward us or away from us, but the image that the rain was not overhead at that moment comforted him.
I wonder if this is what guides political reasoning. If things are relatively fine for our limited dataset then we are more likely to conclude that everyone’s dataset is doing just fine. All it takes is a tipping point and a direction for it to tip to change opinion. I guess that is why advertisers spend so many millions of dollars on advertising. Apparently, in real dollars, we are spending eleven times the amount for advertising than we did just fifty years ago. That is a lot of opinion being altered.
Back to the radar image, by pointing my colleague to www.intellicast.com and setting up a loop he was able to see that clouds were indeed moving our way. We calculated the speed and inferred that direction and range then decided there was a thirty percent chance of being rained out. Those odds were good enough for him.
People are making many decisions in education these days based on just the initial tip of the potential dataset. They are also drawing conclusions and making decisions based upon data from sources that are constantly changing. In my world of educational technology, I cannot accept many conclusions as relevant to today if researchers gathered the data before 1995. After that point, classrooms changed quickly, computers changed dramatically, and schools became wired so that the entire analytical equation changed.