Академический Документы
Профессиональный Документы
Культура Документы
Describes the:
Researchable question itself
Why it's important (i.e., the rationale and significance of your
research)
Propositions that are known or assumed to be true (i.e., axioms
and assumptions)
Propositions that will be tested (i.e., hypotheses or postulates)
Goals and specific objectives of your research activities
Methods you will use to test hypotheses and achieve objectives
Expected results and scope of inference
Hypotheses
General definitions
A new idea
A statement to be tested - an 'educated guess' that needs more
study to be confirmed or disproved
A proposition that explains some phenomenon
Think it through!
Typical formats
Each hypothesis or objective often has its own set of methods
OBJECTIVES My objectives are to:
Objective 1
Objective 2
Approach
The strategy connecting hypotheses to conclusions
Observational, experimental, modeling?
Connection
between methods
and conclusions
must be clear why are you doing
these things?
Design
Randomization, replication, etc
How do you know replication is sufficient?
Statistical approaches
Observations (axioms)
Hypothesis
deduction
Expected results
Prediction
Reject hypoth.
(deduction)
False
Accept hypoth.
(induction)
Test
True
Scope of inference
Closely linked
The conditions to which the conclusions from the research will apply:
Scientific Scope of inference
Biological
Geographical
Temporal
Statistical Scope of Inference
Limitations
Scope of inference limitations
Describe constraints - i.e., resource, time constraints
Evaluation
Are the materials and methods adequate to test the
hypotheses and achieve the objectives?
Is the scope of inference defined, realistic, and adequate?
Are issues of representation, replication, and randomization
appropriate to the proposal and if so, are they addressed?
Is it clear how conclusions will be drawn?
Is the proposed study doable and repeatable?
Are the pitfalls and limitations understood?
Are the experiments novel or creative?
Draw conclusions
Replication
Before we accept the existence of an
effect, the effect must be observable
in replicates that represent the range
of variation* over which inference is to
be made.
-Hurlbert (1983)
*The
scope of inference!
Replication
is the repetition of independent
applications of a treatment or
protocol
Doug-Fir
a bed?
or a section of a bed?
or a or a seedling?
Herb
Herb
water
Severe
Medium
Low
Pine
To confuse
To mingle so that the elements cannot be
distinguished
Confounding: an example
Interest in whether bats forage more along streams then
within forest stands.
In August, sample nighttime foraging activity of
bats along streams in the coastal range.
In October, sample nighttime foraging activity of
bats in forest stands in McDonald Dunn Forest
(near Corvallis).
(Note that in the literature it says that nighttime
foraging activity of bats increases with increasing
nighttime temperature)
Confounding: an example
In August, sample nighttime foraging activity of
bats along streams in the coastal range.
In October, sample nighttime foraging activity of
bats in forest stands in McDonald Dunn Forest
(near Corvallis)
To what should you attribute a difference in foraging activity?
Forest type
Nighttime temperature
Other seasonal effects (e.g. day length, seasonally
available food, day or night light levels)
Randomization
What we do:
Randomly select pieces of material to sample.
randomly select
Randomly assign a piece of material to a protocol.
randomly assign
Order items or protocols randomly.
randomly order
Physically place items randomly.
randomly placed
Why do we randomize?
Randomization is somewhat analogous to insurance, in that
it is a precaution against disturbances that may or may not
occur, and that may or may not be serious if they do occur.
Cochran and Cox 1957
Randomization
What do we mean by randomization?
Mixed up the order?
Cant repeat a selection or an assignment?
See no pattern in a selection or an assignment?
Cant explain how we did a selection or
assignment?
Randomization
Each replicate unit has a known chance of
being assigned to a treatment.
Or
Each sample has a known chance of being
sampled
The process is definable and repeatable.
Randomization ensures that the effects we estimate are
reasonably believed to be true for the whole set were
interested in, not just for the subset.
Inferences
Observational studies can only report associations
between responses and groups
Because you dont know and cant be sure that
something unknown is responsible for the difference
you see between your groups
Controlled designed experiments allow you to draw
cause and effect conclusions
Because in theory, all other effects known to affect
the response have been controlled
Note: natural resource studies are commonly a mix of
observational and design studies. It is not easy to have an
natural resource study that can make cause and effect
conclusions!