Topic 1

THE SCIENTIFIC METHOD

1
Learning objectives
• Define science and explain what the scientific method entails and why it is important
• Distinguish the types of science, and types of reasoning, and outline both of their roles
in the scientific method
• Differentiate between hypothesis vs. prediction vs. theory
• Explain why science proceeds via rejecting, not proving, hypotheses
• Summarize the characteristics that distinguish science from non-science
• Explain confounding variables and the role of controls in addressing them
• Explain the concept of inferential strength and extrapolation, and
how these relate to observational vs. manipulative studies
• Outline the four requirements for science to result in
knowledge acquisition
• Demonstrate concepts from above via the case study on the evolution
of human skin colour

2
 Definitions
• Science: the intellectual and practical activity encompassing the systematic
study of the structure and behaviour of the physical and natural world through
observation and experimentation

• Biology: the science of life

• Scientific method: an approach to knowledge acquisition that

seeks to ensure that our understanding is not shaped by
personal beliefs or biases, but rather is based on evidence
(i.e. data acquired through observation and experimentation)

3
Two types of science

• Descriptive: seeks to characterize ‘patterns’ (i.e. to describe

the physical and/or natural world)

• Hypothesis-testing: concerned with testing one or more causal

explanations for an existing pattern (i.e. to explain observations
of the physical and/or natural world)

4
Both types of science are important to the scientific method

• descriptive science provides the grist for the hypothesis-testing

science mill (i.e., it provides patterns and may suggest possible
explanations)

• hypothesis-testing science interprets patterns and, in doing so,

provides direction as to where to look for other patterns

Descriptive
science

Hypothesis-testing
science

5
 Example: eutrophication in freshwater lakes
1) Descriptive study reveals a pattern: 3) An experiment manipulating [K] finds no
effect, rejecting this hypothesis.
Primary production

4) Characterize more patterns; Is primary

production correlated with something else?
Is [K] correlated with this?

Potassium [K] concentration

2) Hypothesis: [K] is a limiting nutrient such that

increasing availability fosters increased algal growth.
6
 The scientific method
Descriptive
science
Biological
hypothesis
Deduction
Descriptive Induction Hypothesis-testing
science science
Predictions

Study

Inference
Data
Conclusions
(support or reject (statistical (patterns)
biological hypothesis) hypothesis testing)

7
 Induction
• Specific observations (patterns) are synthesized to produce a general statement or
conclusion (reasoning from the particular to the general)
• Even if all the axioms are true, the conclusion is not necessarily true.
• Inductive reasoning is often the source of biological hypotheses, but is ideally not
used to test them

This bird is a swan &

it is white.
And this bird is a swan
& it is white.

∴ All swans are white

8
 The scientific method Descriptive
science
(biological)
hypothesis
Deduction
Descriptive Induction Hypothesis-testing
science science
Predictions

Study

Inference
Data
Conclusions
(support or reject (statistical (patterns)
biological hypothesis) hypothesis testing)

9
 Deduction in hypothesis-testing science
• Hypothesis: a causal explanation for a given pattern

• Prediction: a statement of what will be observed under specified conditions (i.e.,

those of the study we’re going to do) if the hypothesis is true.

• A prediction only exists within the context of a hypothesis and particular study

• The scientific method uses deduction to derive predictions

and hence to test hypotheses

10
 Deduction

• A form of reasoning from one or more general statements (premises) to a

logical conclusion
• There is no uncertainty: if the premises are true then the conclusion necessarily
follows
• It can be represented by a syllogism:

– Premise 1: All birds have feathers.

– Premise 2: All robins are birds.
– Deduction: Therefore, all robins have feathers.

11
 Deduction

• With respect to testing a scientific hypothesis, predictions must follow

deductively from hypotheses

• Syllogisms can also be presented as if…then statements

– If hypothesis X is true,
– and a study of type Y is performed,
– then result Z will be observed.

12
 What makes a “scientific” hypothesis?
• According to Sir Karl Popper, in addition to being causal all
scientific hypotheses must also be refutable, at least in
principle
• A refutable hypothesis is one for which there are possible
outcomes that are inconsistent with it
• I.E., it can be falsified (in theory)
• The ‘hypothesis’ that the fossil record of life on earth was
created by god is not falsifiable. There is no observation that
could refute the existence of a supernatural being. It is
therefore not a scientific hypothesis.

• Why must hypotheses be refutable?

13
 Science proceeds by falsifying hypotheses
• Popper argued that science best proceeds by eliminating hypotheses, not
proving them, because you cannot prove a hypothesis

• “When you have eliminated the impossible, Watson, whatever remains –

however improbable – is the truth.”

Sherlock Holmes, Hypotheses

The Sign of Four

Pattern we want
to explain

14
A logical fallacy: you can’t prove a hypothesis

If H then P Humidity is high this morning because it rained last night (H).
P observed If it rained last night, the garden will be wet (P).
The garden is wet (i.e. P is observed).
∴ H true
Therefore, the high humidity is because it rained last night.

This argument is invalid because the conclusion can be

incorrect even if P follows deductively from H and P is
observed.

Why? Because H is not the only potential cause of P.

So observing P SUPPORTS, but does not prove, H.

15
But you can disprove a hypothesis

If H then P Humidity is high this morning because it rained last night (H).
P NOT observed If it rained last night, the garden will be wet (P).
The garden is NOT wet (i.e. P is not observed).
∴ H false
Therefore, the high humidity is NOT the result of rain last
night.
This argument is valid because the prediction follows
deductively from the hypothesis.

So a failure to observe P falsifies the hypothesis.

16
 Example: why the bathroom light
doesn’t work

Hypotheses

Power What we want

off to house to explain

Bulb burnt
out Light switch is on,
but there’s no light
Short in circuit

17
 Hypotheses and predictions
• Hypothesis: a statement about the cause of some pattern
• Prediction: the pattern one will see in the results of a
particular study if the hypothesis is true
• Inference:
– if predicted pattern is observed, hypothesis is
supported (but not proven)
– if predicted pattern is not observed, the hypothesis is
rejected (falsified).

18
 The scientific method Descriptive
science
(biological)
hypothesis
Deduction
Descriptive Induction Hypothesis-testing
science science
Predictions

Study

Inference
Data
Conclusions
(support or reject (statistical (patterns)
biological hypothesis) hypothesis testing)

19
Type of study
Separate from the type of science, there are two types of study:

1) Observational - researcher observes/measures/characterizes, but

does not alter the system

2) Manipulative (aka an ‘experiment’) - the researcher

changes something and compares what happens to a
control (i.e. unmanipulated) treatment, or one or more
other treatments with different values of the manipulated
variable

The type of study is independent of the type of science.

Don’t confuse them!
20
Examples of the two types of study
2) Manipulative
(experiment)
1) Observational study
Primary production

Phosphorous concentration

21
All combinations exist for types of science
and study
Type of science
Type of
study Descriptive Hypothesis-testing
When do hummingbirds Measure the correlation between
arrive in the spring? chlorophyll content and phosphorus
Observational Where are the areas of across many lakes to test the
highest biodiversity? hypothesis that P is limiting

Fun science. What will Classic experiment to test

happen when I…? predictions of a hypothesis.
Manipulative Treatments are compared to each
other or to a control.

22
 Observational vs. manipulative studies:
why do we care?
Inferential strength is a measure of how strongly the results support the
conclusions.

All else equal (caution, it never is), manipulative studies have greater
inferential strength than observational studies

Why? Because manipulative studies better control for

confounding factors.

This is why you often hear: “correlation doesn’t imply causation”

23
 Confounding factors
A separate, often unknown, factor that may be responsible
for the observed pattern. Statistically, a third variable that is
correlated with the independent variable and which may be
causing the association between the dependent and
independent variables.

Lake primary productivity

(dependent variable)

Potassium conc. Phosphorous conc.

(independent variable) (confounding variable)

24
Consider a manipulative experiment
• The independent variable (e.g. potassium conc.) is
actively changed by the researcher, so confounding
differences are FAR less likely

• When a potential confound exists, it can be addressed via

appropriate controls

• A control is an experimental procedure or treatment level

designed to minimize the effects of confounding variables.

25
 Example: effects of an oncolytic virus on
tumour growth in mice
• Biological question: can a tumour-killing virus effectively reduce
tumour growth rate in vivo?

• Procedure: inject a virus suspension into spontaneous tumour mouse

model and track tumour growth

• Design question: what are the appropriate controls?

– A second set of mice that don’t receive the virus injection?
– Something better?

26
 All else is often not equal: extrapolation
• Studies, especially manipulative experiments, are almost
always conducted on ‘model' systems of smaller scale

• Drawing inferences from results of studies on model

systems requires that we assume that the model system
behaves similarly to the actual system of interest

• This is called extrapolation, and the more that is necessary

the lower the inferential strength

• Observational studies often involve far less extrapolation

than manipulative experiments

27
 Common types of extrapolation
Extrapolation is common and sometimes extreme:

• Interspecies (very common in biomedical studies – e.g. rates as

models for humans)
• From experimental indicators (that which we measure or estimate)
to system properties of real interest (e.g. from expression levels to
protein levels, from species richness to “biodiversity”, etc.)
• Spatial and temporal scales
• In vitro to in vivo

28 28
 The scientific method Descriptive
science
(biological)
hypothesis
Deduction
Descriptive Induction Hypothesis-testing
science science
Predictions

Study

Inference
Data
Conclusions
(support or reject (statistical (patterns)
biological hypothesis) hypothesis testing)

29
 Statistical hypothesis testing
• In almost every study, we want to know if a pattern in the results is
real (i.e. is it the result of chance – i.e. random sampling variation –
or is it a repeatable, biological phenomenon?)
• This is the field of statistical hypothesis testing
• Don’t confuse this with scientific hypothesis testing. Descriptive
science often includes statistic hypothesis testing (again to
determine if the patterns seen are real)
• It’s very important, but is poorly named. It should be called
something other than hypothesis testing (e.g., statistical
inference)

30
Summary – the scientific method
• Falsifiable hypotheses are derived (often
inductively) from patterns arising from
observation and experimentation

• Deductive predictions are tested via

observational and/or manipulative studies
with appropriate controls

• Statistical inference is used to determine whether

predicted patterns are observed

• Inference is made to support or reject hypothesis based

31 on the evidence
 Recall…
• Science: the intellectual and practical activity encompassing the systematic
study of the structure and behaviour of the physical and natural world through
observation and experimentation

• Biology: the science of life

• Scientific method: an approach to knowledge acquisition that

seeks to ensure that our understanding is not shaped by
personal beliefs or biases, but rather is based on evidence
(i.e. data acquired through observation and experimentation)

32
 Science and knowledge acquisition
Knowledge acquisition requires researchers to be:
• Rational: i.e. employ the scientific method
• Skeptical of hypotheses and evidence:
– seek to repeatedly and carefully scrutinize patterns (i.e. Are they real?)
and hypotheses (Are they reasonable? Consistent with data?)
– be willing to reject or modify hypothesis based on the evidence
• Objective: unbiased by preconceived notions, beliefs,
ideologies, experiences, etc.)
• Methodologically materialistic: restrict assumptions and
explanations to the material world (i.e. the supernatural is not
considered)

33
 Science vs. Pseudo-science
• Studies that seek only to confirm beliefs are not science (Popper called them
pseudo-science)

• Consider the hypothesis the world is flat

– One seeking to confirm this hypothesis could find (apparent) evidence in support of it,
and if this fits with your preconceived notion there may be little incentive to do an
exhaustive search for additional evidence that might refute it
– But one seeking to disprove this hypothesis would only need
show that one deductive prediction it makes is false to reject it
– Observations interpreted as evidence of a flat earth are also
consistent with a spherical earth (i.e. they don’t reject a spherical
earth)

34
 Hypothesis vs. theory
• A hypothesis that has survived many attempts at falsification is
referred to as a theory (e.g. the theory of evolution)

• A scientific theory is an explanation of some aspect of the natural

or physical world that has been repeatedly tested via the scientific
method. It has withstood this rigorous scrutiny such that it
constitutes accepted scientific knowledge

• This is very different that the everyday usage of theory to

mean ‘speculation’

35
 Case study: skin colour evolution in humans
• Casual observation and more formal descriptive studies show geographic
variation in skin colour:

Biasutti (1941) By en:User:Cburnett, CC BY-SA 3.0,

https://commons.wikimedia.org/w/index.php?curid=2866948

36 Modified from “The Evolution of Human Skin Color” by A. Prud’homme-Généreux. 2011.

National Center for Case Study Teaching in Science, Univ. at Buffalo, State Univ. of New York
 A biological hypothesis
• Humans and chimpanzees shared a common ancestors 6-7 Mya
• Chimpanzees are light-skinned but covered by dark hair
• Evidence suggests that early humans left the cover of trees for the open
savannah where there is little shade
• Humans lost much of their body hair (possibly because of selection to
facilitate evaporative cooling to dissipate heat)
• UV light causes DNA mutations
• Melanin, a pigment produced by skin cells, absorbs UV light,
shielding cells from UV-induced DNA damage
• Hypothesis: variation in human skin colour evolved from
selection for increased melanin in areas of high UV
exposure because this reduces UV-induced DNA
damage (i.e. skin cancer)
37
 Global UV Index
Testing the hypothesis:
an observational study

Lighter
skin

National Oceanic and Atmospheric Administration.

Retrieved 18 Oct 2009

Darker
skin

Barsh (2003) from Relethford (1997)

38
Problem
• Skin cancer generally arises late in life, long after
reproduction, and is usually not fatal

• Jablonski & Chaplin (2000) argued that selection for

increased melanin resulting from decreased cancer risk
will therefore be very weak Prof. Nina Jablonski

• …and that the cancer-protecting function of melanin is

unlikely to be the primary selective agent favouring
increased melanin (i.e. it may have contributed weakly or
not at all)

39
 Folate
• Folate (folic acid) is an essential nutrient for DNA synthesis and is
especially important during pregnancy when DNA replication rates are
very high in the fetus
Blood folate in people
exposed (‘Patients’) or
• Folate deficiency causes anemia in not (‘Normals’) to UV
light for 9h/d for 3
mothers, serious neural defects in the months.
developing fetus, and increases risk of
Correction to what I say:
miscarriage melanin reduces the loss
of FOLATE due to UV-
• Melanin protects against UV-induced induced degradation.
breakdown (i.e. photolysis) of folate
in the skin

40 Branda & Eaton (1978) Science

New hypothesis
• Hypothesis: humans evolved increased melanin (and hence darker skin)
in areas of high UV exposure because this protected them from UV-
induced degradation of folate

• This can explain the evolution of darker skin in humans following hair
loss, but it CANNOT, on its own, explain the evolution of light skin (i.e.
there is no advantage of light skin, so darker skin should
eventually evolve everywhere)

41
Vitamin D3
• UVB is critical for the synthesis of vitamin D3 which starts in the skin
• D3 is needed for calcium absorption and hence bone growth; deficiencies
can lead to immobilization, developmental deformities, and death
• In northern latitudes, dark skin can cause D3 deficiency

• Hypothesis: selection in more extreme latitudes favours lighter skin to

increase vitamin D production

Vitamin D3

42
 Jablonski & Chaplin (2000)

Insufficient UVB to synthesis in

D3 in light, moderate and dark skin

Sufficient UVB to synthesis Insufficient UVB to synthesis

D3 in even dark skin in D3 in moderate to dark skin

43
Further support
• Female D3 requirements are higher than males during pregnancy and
while breast feeding. And across human populations, females
consistently have slightly lighter skin colour than males.

• D3 can also be obtained through certain foods including fisher liver oil.
Indigenous populations at extreme latitudes have darker skin
pigmentation but historically also had diets rich in such
foods.

44
Overall
• Skin colour has evolved in human populations in response to environmental
differences

• A trade-off exists between selection for darker skin to reduce folate

photolysis and selection for paler skin to facilitate vitamin D3 synthesis

• Effects of melanin in reducing skin cancer probably

contributed little to the evolution of current differences

45
Topic 1: Additional resources

• Excellent Crash Course video about K. Popper and the scientific

method: https://www.youtube.com/watch?v=-X8Xfl0JdTQ&t=397s

• Evolution of human skin colour: Howard Hughes Medical institution video

(with transcripts): https://www.biointeractive.org/classroom-
resources/biology-skin-color

• Biology 2e textbook discusses science and the scientific method

in section 1.1 (pp. 7-13), but they do quite a poor job of it
unfortunately.

46