9. Samples and their potential for bias

An industrial/organizational psychologist is consulting with Family Practice Pediatrics, which has just lost several of its best doctors to a rival pediatric practice. The management at Family Practice would like to know what it can do to increase job satisfaction for its remaining pediatricians. The psychologist is trying to determine whether significant information will be lost if he surveys a subset rather than every single pediatrician at Family Practice.

The following table lists the job satisfaction scores for all eight pediatricians at Family Practice as well as four examples of possible four-pediatrician samples. At the bottom of the table, the variance listed in the second column for all eight pediatricians is the population variance for the eight pediatricians (that is, = SS/n = SS/8). The variances listed in the third through the sixth columns are the sample variances for each of the four-pediatrician samples (that is, s = SS / (n - 1) = SS/3).

Pediatrician	Satisfaction	Sample 1	Sample 2	Sample 3	Sample 4
A	5	x	x		x
B	6		x		x
C	8			x
D	3	x		x	x
E	7	x		x	x
F	2		x	x
G	8	x	x
H	9
Mean	6.00	5.75	5.25	5.00	5.25
Variance	5.50	4.92	6.25	8.67	2.92

What is the mean of the four sample means?

What is the mean of the four sample variances?

A sample statistic is if the average value of the sample statistic, obtained over all possible samples, is equal to the population parameter.

A sample statistic is if it consistently underestimates or overestimates the corresponding population parameter.

A sample mean (that is, M = X/n) is estimator of the population mean.

A sample variance that is computed as s = SS / (n - 1) is estimator of the population variance.