Chapter 10
Describing A Numerical Response
10.1
Pictures
Thus far in these notes, the response has been either a dichotomy or a count that follows the Poisson
or Binomial distribution. In this chapter we extend our work to counts that are neither Poisson nor
Binomial and to responses that are measurements.
Suppose the subjects are students in this class. Below are some examples of numerical re
sponses.
•
Counting:
Number of points on homework to date; number of credits this semester; number
of persons living in current household.
•
Measuring:
Height; weight; age.
As often happens in life, the boundary between these options can be blurry. For example,
consider annual income. Literally, annual income is determined by
counting
the number of dollars
earned in the year, but economists and other researchers tendtotrea
ti
tasameasuremen
t
. The
general guideline is that if a count variable has many values in a population, and no one value
dominates others in terms of relative frequency, it is usually mathematically more convenient to
treat the variable as a measurement.
Two important words are:
precise
and
accurate
.A
c
c
u
r
a
t
em
e
a
n
sc
l
o
s
e
t
o
t
h
e
t
r
u
t
h
.F
o
r
example, if I state that my dog Casey lived for 15.5 years, tha
tisaccu
ra
te
. I
fIs
ta
tetha
tmy
grandfather Wardrop lived to be 110, that is highly inaccurate.
Precise is most useful for measurements. If I state: Yesterday I ran one mile in 250.376 seconds,
this is incredibly precise (to the nearest onethousandths of a second), but ridiculously inaccurate.
If I say I ran it ‘In less than one hour’ it is accurate, but not the least precise.
Here is a good general guideline for science: measurements should be precise enough to create
variation in our population or subjects of interest, but there is no need to get carried away with it!
Precise is somewhat meaningless for counts that take on smallvalues.Forexample,itisaccu
rate to say that 2 cats live in my house. It is no more precise to say I have 2.000 cats!
103
This preview has intentionally blurred sections. Sign up to view the full version.
View Full DocumentTable 10.1: Sorted Speeds, in MPH, by Time, of 100 Cars.
Speeds at 6:00 pm
26
26
27
27
27
27
28
28
28
28
28
28
28
28
28
28
28
28
28
29
29
29
29
29
29
29
29
29
29
29
30
30
30
30
30
30
30
30
31
31
31
31
32
33
33
33
34
34
35
43
Speeds at 11:00 pm
27
28
30
30
30
31
31
31
32
32
32
32
32
32
32
33
33
33
33
33
33
33
34
34
34
34
34
34
35
35
35
35
36
36
36
37
37
37
37
37
37
38
38
39
39
40
40
40
40
40
For large counts, precision does become meaningful. For example, if forced to guess, I would
say that there are 300 million people living in the US. I suspect that this is somewhat accurate, but
clearly I am not being very precise.
10.1.1
Dot Plot
We begin with an example of measurement data, taken from a student project in my Statistics 301
class.
On a spring evening, a Milwaukee police of±cer measured the speeds of 100 automobiles. The
data were collected on a street in a “warehouse district” withaspeedlimitof25MPH
.Fiftycars
were measured between roughly 5:45 and 6:15 pm, referred to below as 6:00 pm. The remaining
50 cars were measured between roughly 10:40 and 11:20 pm, referred to below as 11:00 pm.
This is the end of the preview.
Sign up
to
access the rest of the document.
 Fall '11
 hanlon
 Binomial, Standard Deviation

Click to edit the document details