Key Points

1. The scientific method is a manner of thinking and working towards more complete knowledge of the world.
2. To be scientific, a statement must, in principle, be falsifiable.
3. Sciences may be classified as experimental, observational, or historical.
4. There are many forms of scientific inference, with different logical foundations and degrees of rigour. to lax.
5. Scientific explanation is linked to causality. A parsimonious explanation is preferred.
6. A scientific statement may be a fact, hypothesis, theory, or law, each with a level of certainty.
7. An important type of scientific reasoning is deductive-inductive.
8. Scientific explanation requires sound logical thinking.

---

What is Science?

To “do science” is to follow a prescribed method to arrive at knowledge. The “scientific method” is not a belief system or religious dogma, but rather a manner of thinking and working towards more complete knowledge of the world. It has been proven to be extremely successful in:

• explaining the world as we observe it;
• predicting what can be further observed, e.g. new observations, new locations, repeat observations, the effect of interventions;
• engineering, i.e. building things that work.

Science is not prescriptive – it can not say what “ought” to be done. It can, however, point out the probable consequences of certain actions, as objectively as possible.

Characteristics of scientific knowledge

• Self-criticism
• Evidence-based
• Theory-based
• Transparency
• No appeal to authority

Types of Sciences

Scientific activity can be classified as experimental, observational, or historical. All three require a separate step of model building.

1. Experimental
2. Observational
3. Historical

Scientific Inference

Science attempts to reach conclusions from premises and observations, using some form of rational argument. The general term for making new statements on the basis of previous statements is inference.

1. Purely logical
2. Deductive-inductive
3. Cause & effect
4. Contributors & impacts
5. Inductive patterns (classification)
6. Case studies
7. Analogy
8. Probabilistic
9. Functional
10. Systems explanations
11. (Expert) Judgement / Wisdom / Intuition
12. Teleological, ‘higher’ purpose, external cause

File:An-approach.png

Figure 1: An approach to scientific knowledge.

Scientific Explanation

To “explain” is to say “why” something happens or is observed – but this is very difficult, if not impossible, to establish. In applied science and engineering we are mostly content with a more limited view of “why”: a coherent statement that allows prediction of the phenomenon in the future, in other situations or at other locations besides the ones already observed.

Process of explanation: “Making an unexpected outcome an expected outcome, of making a curious event seem natural or normal” ; it becomes ‘natural’ once the processes which gives rise to the outcome (given similar conditions) are clear.

The concept of “causality” is also tricky. What appears at first to be the cause must itself have a cause, and so forth. The proximate (immediate) cause may be fairly easy to establish, but the deeper causes requires either more evidence or more speculation. It is probably meaningless to speak of an “ultimate” (last, final) cause.

Ockham’s Razor

“Numquam ponenda est pluralitas sine necessitate”

“Complexity should never be added to an explanation unless necessary”

This means that if several theories equally explain the observed facts, the simplest should be used.

Levels of Certainty

Facts: A fact is something directly observable and measurable.

Hypotheses: A hypothesis is a tentative theory, not yet tested; it is what we believe to be the true explanation or true state of nature, based on previous work or first principles. According to Harvey (1969) a hypothesis should be 'logically consistent controlled speculation' – note that a hypothesis must at least be internally-consistent (’logical’).

Theories: A theory is a conceptual framework which:

• explains existing facts;
• allows predictions; and
• is in principle falsifiable (some experiment or observation could contradict it or force its modification).

A theory is a “highly articulate systems of state- ments of enormous explanatory power”; that is, there is enough evidence behind the theory, and it is expressed in enough detail, to allow many complex predictions. This detail is expressed in a model.

Laws: A law is theory with overwhelming evidence, including the conditions under which it is true. A similar definition is a theory whose falsification, within its context, is almost inconceivable.

The boundaries in the sequence [hypothesis ⇒ theory ⇒ law] are of course fuzzy.

File:Di-approach.png

Figure 2: The deductive-inductive iterative approach to scientific knowledge.

The Deductive-Inductive Scientific Method

The best-known scientific method is known as the “deductive-inductive” approach. It has the following structure:

1. Observe and synthesize general knowledge of the world;
2. Invent a theory to explain the observations ⇒ abduction;
3. Use the theory to make predictions ⇒ deduction;
4. Design experiments or Make more observations to test these pre- dictions;
5. Modify the theory in the light of results ⇒ induction;

Repeat from step 3 until you can’t think of any new predictions that might falsify or modify the theory.

Step 4 is the crucial stage of experimental design: make new observations where they are most likely to contradict what is expected or where an unexpected result would make maximum damage to the the- ory. That is, the maximum information from a new experiment or obser- vation comes either when the outcome is least predictable, or when it so predictable that an unusual result would be devastating.

Since we don’t start from the beginning, the “Observe” and “Theory” steps are based on others’ previous work and our general knowledge.

---

References