MAT 2379, Introduction to Biostatistics, Section 7.

1 (Part 1) 1

MAT 2379, Introduction to Biostatistics

Chapter 7. Introduction to statistics
7.1 Random Sampling and Data Description

We will start by examining the question:

What is a statistical problem?
For example, a researcher is examining a group of 50 people of age 18+ and is interested in examining the factors
which are associated with the development of heart disease. These factors include: age, weight, smoking status, and
the family history.
This problem has all the characteristics of a statistical problem:
1. Associated to this problem, there is a large group of objects about which we have to draw a conclusion. This
group of objects is called the population. In the example above, the population consists of all people of age
18+.
2. We are interested in certain characteristics of the members of the population. These are called variables and
are denoted by the capital letters X, Y, Z, etc. In the example above, X =age, Y = weight, Z =smoking status.
3. The population is too large to study. So, we must draw conclusions by studying only a portion of the population
called a sample. The number of objects in the sample is called the sample size and is denoted by n. In the
example above, n = 50. The observed values of the variable X (say age) are denoted by x1 , . . . , xn .
In Section 7.1, we will examine questions about the behavior of variables:
• What graphical methods to use to summarize the observed values?
• About what values does the variable fluctuate?
• What is the variation in the observed values?
Variables are of 2 types:
1. Categorical variables. These are variables whose values fall in several categories. For instance: blood type,
sex, color of a flower.
2. Quantitative variables. These are variables which take numeric values. They are of two subtypes:
a) discrete variables, whose values can be listed (e.g. age, number of seeds which germinate)
b) continuous variables, whose values lie in a certain range but cannot be listed (e.g. weight, height, blood
pressure).
Example: In each case below, specify what is the population, the sample size, the variables of interest (and their
type).
populatio
hospital, the birth weight and mother’s blood type were recorded for 15sample
a) In a large
esfg
b) A pediatrician measured the height increase over 1 year for 25 patients.
newborns. size

c) At a center blood donation, the cholesterol level was recorded for 100 donors.
yy
Part 1: Frequency Distributions

T
Below will introduce various graphical methods for summarizing a data set.
a) The Bar Charts
A bar chart is a graphical method used for categorical variables. The vertical bars have the same width and
are separated by some arbitrary space. The heights of the bars represent the frequencies for each category of the
variable. Alternatively, one can draw a bar chart of the relative frequencies:
the frequency
The relative frequency =
n
Example 1. (Color of poinsettias) A sample of 25 poinsettias were classified according to their color as follows:
MAT 2379, Introduction to Biostatistics, Section 7.1 (Part 1) 2

Color Frequency Relative frequency

Pink 8 0.32
Red 11 0.44
White 6 0.24
Total 25 1.00
Here are the bar charts of frequencies and relative frequencies:

b) Histograms

A histogram is a graphical method used for quantitative variables. The scale of the variable determines the
placement of the bars. Usually, the vertical bars have the same width. Unlike the bar charts, the bars are not
separated by some arbitrary space.

b1) Histograms for discrete variables. In this case, the width of the bars is (usually) 1 unit.

Example 2. (piglets) A company who owns a large number of pig farms is interested in the distribution of the
number X of surviving piglets per sow. X is a discrete variable which can take the values: 0, 1, . . . , 20. A sample of
10 sows is selected. We record the observed values of X for this sample:
sons
i 1 2 3 4 5 6 7 8 9 10
xi 5 8 6 5 9 7 6 9 9 8
pi9
Below is the table of frequencies and relative frequencies for this data:
Number of Frequency Relative Frequency
Surviving Piglets ProbabilityDensity
5 2 0.2
6 2 0.2
7 1 0.1
8 2 0.2
9 3 0.3
Total 10 1.0
Here are the histograms of frequencies and relative frequencies:
frequency relativefrequencies
area

Mfg

also probabilityhistograms 4,41

MAT 2379, Introduction to Biostatistics, Section 7.1 (Part 1) 3

b2) Histograms for continuous variables. In this case, we have to arrange the data into groups (called bins)
and count how many values lie in each bin. One has to select the number of bins (this is a delicate issue, which will
not be discussed here). Usually the bins have the same width.

Example 3. (height) A sample of 15 college students were asked how tall they were. Here is the data (in inches):

66.5 61.2 63.9 62.7 65.1 68.7 64.3 73.3 69.3 66.5 70.1 71.3 68.1 67.4 66.7

We arrange the data into 7 bins of equal width: bin 1 contains the values between 61 and 63, bin 2 contains the
values between 63 and 65, etc. The following table gives the frequencies and the relative frequencies:
Height Frequency Relative Frequency Probabilitydensity
61-63 2 2/15
2 15 22 lengthof bin
63-65 2 2/15

it
65-67 4 4/15
67-69 3 3/15
69-71 2 2/15
71-73 1 2/15
73-75 1 1/15
Total 15 1.0
N A
Here are the histograms of the frequencies and relative frequencies:

probabilitydensity

4 15 2
area
p

iii iaea.TK iimA

shape of thehistogram

t.IN
MIA skewed totheright

in

skewed tothe left