Chapter 1:

Statistics and Samples

What is biometry
In order to say make
statements about life, a
biologist must measure
biological phenomena
Biometry in Medicine
• Clinical Trials
Biometry in Dentistry
Biometry in Environmental
Science
Biometry in Wildlife Biology
Data Collection
Variables
• Variables
Data
• Data are measurements
of one or more variables
made on several
individuals
 Populations
• Data represents populations
• Populations
 Populations
• Defining a populations depends
on the intentions of the
researcher
• Must consider the organism at a
spatial location and/or temporal
scale
• The human population of this room
• The squirrel population of WTAMU
• Gazelles on the Serengeti in the last
decade
 Populations and Sampling
• Characteristics that describe
populations such as their
averages, proportions or
measures of variation are called
___________
• E.g. The average human height, I
mean the average for all 7.5 billon
humans!
 Populations and Sampling
• A critical issue to understand is
that we can never measure all
the individuals in a population
• Could you measure the height of
every human on earth?
• Because we cannot measure all
the individuals in a population
we can never know the true
values for their parameters
 Populations and Sampling
• This fact means we must rely on
sampling
• Sample – subset of the population
which has been collected for
measurement
• We then apply statistics –
methods to describe and
measure aspects of populations
from samples
 Populations and Sampling
• Statistics relies on _________,
the process of inferring an
unknown quantity of a
population using sample data
• Using estimation we can
calculate an estimate, aka a
statistic
• Parameters are the truth
• Estimates is an approximation of
the truth
 Populations and Sampling
• Scientists frequently employ
_____________________– a
specific claim regarding a
population parameter inferred
from an sample estimate
• E.g. “the mean height for humans
is greater than the mean height
for humans 200 years ago”
You never really “prove” anything
in science
• At the smallest scale:
• All estimates have
uncertainty
• Data collection is
prone to error
• You can never know if
your estimate
matches the
parameter
• At the largest scale:
• All scientific ideas are
subject to
reevaluation
You never really “prove” anything
in science
• This is among the
greatest problems
with the general
publics understanding
of science
You never really “prove” anything
in science
• Science relies on the
preponderance of the
evidence
• Science cannot know
the “truth”!
 An example with cats
• Whitney an Mehlhaff
(1987) reported on the
injury rates of cats that
had plummeted from
buildings in NYC
 An example with cats
• They could not measure all
cats that had fallen out of
buildings
• Thus, they sampled by
recording injuries and
building heights from cats
brought into the vet
hospital
An example with cats
• They found
that injuries
increased up
until 7-8
stories
• After that
the cats
injury rate
decreased
drastically
An example with cats

• From their sample, they inferred that any cat

that fell 9 or more stories would have an injury
rate on par with a cat that fell 3 stories
An example with cats

• Why did they see this pattern?

• The authors proposed that after the cat attains
terminal velocity (6 stories) it will relax and be
more capable of absorbing the impact
An example with cats

• Understand that this explanation cannot be

inferred by their data directly
An example with cats

• All their data says is that cats falling out of

buildings will be increasingly injured to about 9
stories, after that injuries will decrease
An example with cats

• The authors are implying __________ to explain

their pattern
An example with cats

• _________ - connects one process (the cause)

with another process (the effect)
• A made B happen
An example with cats

• This is acceptable for a scientist to do, to an

extent
• This can lay a foundation for ideas to be
examined in the future
Correlation does not imply
causation
• The proposed explanation must be reasonable,
and more importantly, testable
• If you go overboard, this can get totally
ridiculous…
Properties of good samples
• Estimate will always differ somewhat from the true
population parameters
• ___________– is the difference between an
estimate and the population parameter being
estimated
• Our goal as biometricians is to minimize error as
much as possible
Properties of good samples
• Sampling error can be broken down into precision
and accuracy
Properties of good samples
• The spread of
estimates
resulting from
sampling error
indicates the
________ of
the estimate
Properties of good samples
• Using the target
example, this is
called group size
• The red box
shows very
precise samples,
i.e. small group
size
Properties of good samples
• ______________
is the systematic
discrepancy
between the
average of multiple
estimates and the
true population
parameter
Properties of good samples
• Using the target
example, this is like
hitting the bullseye,
where the bullseye is
the unknown
parameter
• The red box shows
accurate samples, i.e.
in the center of the
target
Properties of good samples
• You want to be
accurate and
precise
Properties of good samples
• You do not
want to be
inaccurate and
imprecise
How to collect a good sample
• A representative sample is a small quantity of
something that accurately reflects the larger entity
How to collect a good sample
• Random samples should, in theory be more
representative than a non-random sample
• Randomness is the
___________________________________
How to collect a good sample
Random samples have two properties
1. Each individual sampled has the same chance of being
sampled as any other individual
2. Selection of individuals is independent of other
individuals in the sample
• i.e. previous selections do not influence the current selection
Samples of convenience
A __________________is a sample based on
individuals that are easily available to the researcher
• Presents a lot of opportunity for bias
• E.g. only the cats who were brought to the vet were
measured. Those that survived with no injuries or died
were unlikely to see the vet
Types of data and variables
• Data is composed of variables, which may take
several forms
• Data types have a huge impact on how the data can
be analyzed
Types of data and variables
• ___________ variables – describe membership into
a category or group
• Alive or dead
• Big, medium, large
• Red, green, blue
Types of data and variables
• Categorical variables can be either:
• Ordinal – have an order
• e.g. life stage, pain scale (like when you're at the ER)
• Nominal – have no order
• e.g. sex chromosome genotype (XX or XY)
Types of data and variables
• Numeric or Numerical – quantitative and have
magnitude, they are numbers
• Counts
• Mass (g)
• Lengths (mm)
• Percentages
Types of data and variables
• ___________________ data can take the form of
any real number within some range
• 1, 1.358, 6.2, 4.5
Types of data and variables
• Discrete numerical data are real number but are
indivisible units (you cant have a fractional part)
• Number of amino acids in a protein
• Number of bones in the hand
Types of data in R
• In general, computing (such as R) has its own
definitions of data type which are somewhat
separate from that of statistics
• Don’t confuse them

http://vitalflux.com/learn-r-different-data-types-code-examples/
Explanatory and response
variables
• In statistics, a lot of effort is spend relating one
variable to another
• The explanatory (aka independent) variable
predicts of affect the response (aka dependent)
variable
Explanatory and response
variables
Frequency and probability
distributions
• Probability
distributions show how
often each value of the
variable occurs in the in
the whole populations
• Frequency distributions
show how often each
value of the variable
occurs in the sample
• Like the relationship
between the parameter
and the estimate
Types of studies
• _____________ studies – the
researcher assigns different
treatment groups or values of an
explanatory variable randomly to
the individuals units of study
• e.g. clinical trials
• Can determine cause-and-effect
(sort of)
Types of studies
Types of studies
• Observational studies – the research has no control
over which units fall into which groups