Slide 27 of 62
Notes:
As statisticians well know the confidence with which we assess the nature of a population is a function of the degree to which we sample it -- larger sample sizes yielding greater confidence. Experimental subjects have often shown a distinct preference f
It is our bias to assume that population characteristics are reflected in small local samples. We assume that the likelihood that a family of six children would have the birth pattern GBBGBG is much greater than BBBBBG. Of course the likelihood is the s
In the above example, Alyssa’s data is far more compelling than Sean’s. Twice as compelling in fact. Her odds of being right are 16:1 where Sean’s are only 8:1 in spite of the fact that he more clearly had a majority of red balls in his sample.
