Top Banner
LOGIC AND CRITICAL THINKING Lecture 11 – Scientific Reasoning I
43

LOGIC AND CRITICAL THINKING Lecture 11 – Scientific Reasoning I.

Mar 26, 2015

Download

Documents

Gavin Franklin
Welcome message from author
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Page 1: LOGIC AND CRITICAL THINKING Lecture 11 – Scientific Reasoning I.

LOGIC AND CRITICAL THINKINGLecture 11 – Scientific Reasoning I

Page 2: LOGIC AND CRITICAL THINKING Lecture 11 – Scientific Reasoning I.

REVIEW

First, let’s go over homework from last time.

Page 3: LOGIC AND CRITICAL THINKING Lecture 11 – Scientific Reasoning I.

REVIEW

Last time we talked about the relationship between correlation and causation.

If we know that two events co-occur – if they are correlated – that is some evidence that one might cause the other.

But, there are multiple possible explanations of correlations – there are several different relationships the two events might have.

Page 4: LOGIC AND CRITICAL THINKING Lecture 11 – Scientific Reasoning I.

REVIEW

If event A is correlated with event B, any of these are possible: There is no causal relationship between A and B

A causes B B causes A A and B form a causal loop A is a minor cause of B (or B is a minor cause of A)

A and B share a common cause, C. A is an indirect cause of B, via a side-effect C (or B is an indirect cause of A)

Page 5: LOGIC AND CRITICAL THINKING Lecture 11 – Scientific Reasoning I.

REVIEW

We also went over several different fallacies related to poor causal reasoning: The POST HOC fallacy occurs when someone concludes that A causes B without any evidence other than the fact that A comes before B.

Confusing CORRELATION with CAUSATION occurs when someone knows that A and B are correlated, but jumps to the conclusion that A caused B. (usually this involves neglecting a common cause)

REVERSING CAUSAL DIRECTION occurs when someone assumes that A caused B, while neglecting the possibility that B caused A.

Page 6: LOGIC AND CRITICAL THINKING Lecture 11 – Scientific Reasoning I.

REVIEW

We also went over several different fallacies related to poor causal reasoning: The SINGLE CAUSE fallacy occurs when someone takes a complex causal situation and oversimplifies it – looking at only one cause, when really there are many.

The GENETIC fallacy is a little different – it doesn’t make a mistake about what caused an event, but instead it makes a mistake about how important the cause is. The genetic fallacy involves attributing a feature of the cause to the effect – for example, claiming the effect is bad because the cause is bad.

Page 7: LOGIC AND CRITICAL THINKING Lecture 11 – Scientific Reasoning I.

SCIENCE

Today we begin talking about scientific reasoning. The goal for today is to look at some of the big-picture features of scientific method.

We will be focusing today on scientific THEORIES. Next time, we will look in more detail at scientific EXPERIMENTS.

Page 8: LOGIC AND CRITICAL THINKING Lecture 11 – Scientific Reasoning I.

SCIENCE

We’ll start by looking at the basic principle of theory formation and testing in science – the HYPOTHETICO-DEDUCTIVE METHOD.

Then, we’ll look at some of the main features that make scientific theories good or bad.

Page 9: LOGIC AND CRITICAL THINKING Lecture 11 – Scientific Reasoning I.

SCIENCE

We can talk about four main elements that go into the practice of scientific research: The WORLD The THEORY Theoretical PREDICTIONS Experimental DATA

The relationships that these elements have to one another define the goals and methods of science.

Page 10: LOGIC AND CRITICAL THINKING Lecture 11 – Scientific Reasoning I.

SCIENCE

The WORLD is what science is attempting to characterize, describe, and explain.

The objects, properties, events, processes, and so on that are in the world make up the subject matter of science – what science is about.

WORLD

Page 11: LOGIC AND CRITICAL THINKING Lecture 11 – Scientific Reasoning I.

SCIENCE

A scientific THEORY is a set of claims about the world.

This can include statements about facts, laws of nature, etc.

A good theory fits the world – it describes it accurately.

WORLD THEORY

Fits

Page 12: LOGIC AND CRITICAL THINKING Lecture 11 – Scientific Reasoning I.

SCIENCE

A theory in science only describes a small portion of the world – maybe only a single phenomenon or type of phenomenon.

But it attempts to give a full explanation of that phenomenon, leaving nothing ‘mysterious’ or unexplained.

WORLD THEORYFits

Page 13: LOGIC AND CRITICAL THINKING Lecture 11 – Scientific Reasoning I.

SCIENCE

Examples of theories in science: the theory of evolution, the theory of gravity, the theory of special relativity.

Notice that each of these is more than just a single statement or hypothesis – each is a complicated set of laws and principles.

WORLD THEORYFits

Page 14: LOGIC AND CRITICAL THINKING Lecture 11 – Scientific Reasoning I.

SCIENCE

Theories in science generate PREDICTIONS about what will happen under certain specified conditions.

Predictions can be about the past, present, or future.

WORLD THEORY

Fits

PREDICTION

Generates

Page 15: LOGIC AND CRITICAL THINKING Lecture 11 – Scientific Reasoning I.

SCIENCE

Examples of predictions: the theory of evolution predicts the existence of transitional life-forms; the theory of gravity predicts that if I drop something it will fall.

WORLD THEORY

Fits

PREDICTION

Generates

Page 16: LOGIC AND CRITICAL THINKING Lecture 11 – Scientific Reasoning I.

SCIENCE

Predictions are the tools we use to test our theories; if the prediction comes out true, this is good for the theory. We do this through looking at DATA.

WORLD THEORY

Fits

PREDICTION

Generates

DATA

Generates Confirms

/disconfirms

Page 17: LOGIC AND CRITICAL THINKING Lecture 11 – Scientific Reasoning I.

SCIENCE

Data, of course, is generated by the world – we usually elicit it by running EXPERIMENTS.

Once we gather data, we can analyze it to see if it either confirms or disconfirms our predictions.

WORLD THEORY

Fits

PREDICTION

Generates

DATA

Generates Confirms

/disconfirms

Page 18: LOGIC AND CRITICAL THINKING Lecture 11 – Scientific Reasoning I.

SCIENCE

Let’s look at the process of testing a theory in more detail. The process we use to test scientific theories is called the HYPOTHETICO-DEDUCTIVE METHOD (Lau has the name slightly wrong).

It consists of four basic steps: 1) Identify hypothesis 2) Generate predictions 3) Gather data via experiment 4) Determine whether hypothesis is confirmed or disconfirmed.

Page 19: LOGIC AND CRITICAL THINKING Lecture 11 – Scientific Reasoning I.

SCIENCE

Let’s go over each step in detail. STEP ONE: Identify the hypothesis to be tested. A hypothesis is a proposed explanation for something we’ve observed.

Generally, this hypothesis is a component of a theory which we are trying to test. But it can also be a proposed explanation that isn’t part of a bigger theory.

Example: One hypothesis that might arise from the theory of evolution is the claim that animal camouflage is an adaptation to a particular environment that developed over many generations.

Page 20: LOGIC AND CRITICAL THINKING Lecture 11 – Scientific Reasoning I.

SCIENCE

Let’s go over each step in detail. STEP TWO: Generate predictions from the hypothesis. In this step, we want to come up with something we can check in the world to see whether our hypothesis is correct.

Example: If our hypothesis is that animal camouflage is an evolved adaptation to a particular environment, we might predict that a creature with camouflage that was put in a different environment would eventually evolve new coloring.

Page 21: LOGIC AND CRITICAL THINKING Lecture 11 – Scientific Reasoning I.

SCIENCE

Let’s go over each step in detail. STEP THREE: Gather data via an experiment. In this stage, we set up a situation under which our prediction will be tested.

Example: To test our prediction about animal camouflage, we might introduce a certain kind of short-lived animal – perhaps a moth – to a new, different-colored environment and observe whether its coloration changes after a large number of generations.

Page 22: LOGIC AND CRITICAL THINKING Lecture 11 – Scientific Reasoning I.

SCIENCE

Let’s go over each step in detail. STEP FOUR: Determine whether the data confirms or disconfirms the hypothesis. If the prediction was right, the hypothesis is confirmed; if it is wrong, the hypothesis is disconfirmed.

Sometimes this step will require analyzing the data very closely – in particular, it’s important to try to rule out the possibility that the data fit the prediction only by coincidence.

Example: Imagine we notice that most moths in the 50th generation were darker in color than moths of the first generation, and they now match the local colors. This would confirm our hypothesis.

Page 23: LOGIC AND CRITICAL THINKING Lecture 11 – Scientific Reasoning I.

SCIENCE

There are a few very important things to notice about the hypothetico-deductive method.

First, in order for the method to work, our hypothesis must be able to generate predictions that we can test.

Since the hypothetico-deductive method forms the foundation of science, this means that if a hypothesis can’t doesn’t generate predictions we can use the method to test, then it is not a scientific hypothesis.

Page 24: LOGIC AND CRITICAL THINKING Lecture 11 – Scientific Reasoning I.

SCIENCE

Imagine someone tells you that they have the following hypothesis: invisible, undetectable ghosts are influencing world events.

The problem with this hypothesis is that there is no way to test it – the ghosts are undetectable, and we can’t set up an experiment to show that they are there. So the hypothesis cannot be said to be a scientific one.

Page 25: LOGIC AND CRITICAL THINKING Lecture 11 – Scientific Reasoning I.

SCIENCE

Second: even though it contains ‘deductive’ in its name, the hypothetico-deductive method is an inductive method. If an experiment supports your hypothesis, this provides good evidence for the theory you are testing. But it does not PROVE the theory.

Page 26: LOGIC AND CRITICAL THINKING Lecture 11 – Scientific Reasoning I.

SCIENCE

There are always alternate possibilities that might explain why your experiment had the results it did.

Think of the moth case – maybe the moths are getting darker because of some sort of disease that is spreading through their population, rather than due to adaptation.

Page 27: LOGIC AND CRITICAL THINKING Lecture 11 – Scientific Reasoning I.

SCIENCE

We said that an experiment can CONFIRM a hypothesis, but this can be misleading.

The word ‘confirm’ usually means something very close to ‘prove’ or ‘make certain’. However, in science, it just means that the probability that the hypothesis is true has increased.

Page 28: LOGIC AND CRITICAL THINKING Lecture 11 – Scientific Reasoning I.

SCIENCE

Third: Disconfirming a theory does not mean it is false.

Again, ‘confirm’ and ‘disconfirm’ have special meanings in science. When a theory is disconfirmed, this only means that the probability that it is true has decreased.

It may still turn out true – particularly if there was something wrong with the experiment!

Page 29: LOGIC AND CRITICAL THINKING Lecture 11 – Scientific Reasoning I.

SCIENCE

So now we have a basic idea of the fundamental method of science.

Next, let’s look at some of the features we look for in a good scientific theory.

Page 30: LOGIC AND CRITICAL THINKING Lecture 11 – Scientific Reasoning I.

SCIENCE

Ideally, running experiments will give us some idea about whether a theory is likely to be true or false. But this isn’t the only thing we look for in a good theory.

In fact, sometimes multiple theories are CONSISTENT with all the experimental data. In this case, we may need to use other criteria to decide which is the better theory.

Page 31: LOGIC AND CRITICAL THINKING Lecture 11 – Scientific Reasoning I.

SCIENCE

A good theory should: Have PREDICTIVE POWER Explain the relevant phenomenon in terms of underlying causal MECHANISMS

Be FRUITFUL Be SIMPLE Be COHERENT

Let’s look at each of these in turn.

Page 32: LOGIC AND CRITICAL THINKING Lecture 11 – Scientific Reasoning I.

SCIENCE

A good theory has a lot of PREDICTIVE POWER. We’ve already mentioned that if a theory or hypothesis doesn’t generate predictions, it is not truly scientific at all.

But the number and usefulness of the predictions matter, too. A theory that only results in a few predictions is hard to confirm/disconfirm. A theory that generates a lot of predictions is more desirable.

Page 33: LOGIC AND CRITICAL THINKING Lecture 11 – Scientific Reasoning I.

SCIENCE

A good theory also has EXPLANATORY POWER – it’s capable of generating explanations of the phenomena we’re interested in.

A theory can gain explanatory power by proposing CAUSAL MECHANISMS for the phenomena we’re interested in.

Page 34: LOGIC AND CRITICAL THINKING Lecture 11 – Scientific Reasoning I.

SCIENCE

An example of a theory that doesn’t explain via causal mechanisms: Caffeine cures headaches.

This theory is testable and generates predictions (like if I drink coffee, my headache will go away), but it does not explain the relationship – it doesn’t say why caffeine causes headaches to go away.

Page 35: LOGIC AND CRITICAL THINKING Lecture 11 – Scientific Reasoning I.

SCIENCE

A similar theory that DOES explain via causal mechanisms: Caffeine causes the blood vessels in your brain to contract, decreasing blood flow and therefore pressure in the brain. This can help cure headaches.

This theory will make similar predictions to the first, but it is much more explanatory. It tells us WHY caffeine helps – it explains the mechanism.

Page 36: LOGIC AND CRITICAL THINKING Lecture 11 – Scientific Reasoning I.

SCIENCE

Another desirable feature in a theory is FRUITFULNESS. This feature has to do with how many applications the theory has outside its direct area.

Can it lead us to new predictions, or new theories? Can it help make connections between theories we already have?

Page 37: LOGIC AND CRITICAL THINKING Lecture 11 – Scientific Reasoning I.

SCIENCE

An example: The theory of evolution is a very fruitful theory. It explains how complex animals came to be, but it also helps explain things outside its direct area.

For example, the theory of evolution can explain why it is so difficult to make a vaccine for colds – cold viruses mutate and evolve very quickly.

It also explains why antibiotics become less effective over time.

Page 38: LOGIC AND CRITICAL THINKING Lecture 11 – Scientific Reasoning I.

SCIENCE

Another feature we can use to decide between two theories is SIMPLICITY.

Simplicity isn’t as important as predictive power/accuracy, or explanatory power. But it can help decide between two theories that both seem to be very accurate and powerful.

Page 39: LOGIC AND CRITICAL THINKING Lecture 11 – Scientific Reasoning I.

SCIENCE

Example: “Caffeine reduces headaches by constricting blood vessels” is a simpler theory than “Caffeine reduces headaches by triggering the release a not-yet-discovered chemical which acts as a painkiller, and has the side effect of constricting blood vessels”.

Page 40: LOGIC AND CRITICAL THINKING Lecture 11 – Scientific Reasoning I.

SCIENCE

Both theories might be consistent with the data; but, the second introduces a new element – the unknown chemical. This isn’t strictly needed in order to explain why caffeine helps with headaches – there is no reason to assume it exists.

All the ‘undiscovered chemical’ hypothesis does is complicate the theory.

Page 41: LOGIC AND CRITICAL THINKING Lecture 11 – Scientific Reasoning I.

SCIENCE

Finally, a good theory should have the feature of COHERENCE. This means two things – first, the theory shouldn’t internally contradict itself. It is very bad if a theory generates conflicting predictions, for example.

But it also means that the theory should fit well with our other well-established scientific theories.

Page 42: LOGIC AND CRITICAL THINKING Lecture 11 – Scientific Reasoning I.

SCIENCE

It would be a very bad thing if my new theory of how black holes work contradicted the predictions of the theory of gravity.

Of course, sometimes good theories contradict well-established theories – the theory that the earth goes around the sun is an example.

But most of the time, a theory that contradicts current scientific understanding is likely to be wrong.

Page 43: LOGIC AND CRITICAL THINKING Lecture 11 – Scientific Reasoning I.

NEXT TIME

We’ll finish with an inclass exercise today.

The third homework is due in one week – it will be posted on Moodle later tonight.

Next time, we’ll talk about the details of designing a good experiment, and of interpreting experimental data.