The Scientific Method

The scientific method places a higher significance on uncovering errors than it does on finding new truths.  Finding truth is the ultimate goal and error-correction is the means to achieve this goal.  To guard against personal biases (scientists are humans after all) error-checking is assigned to scientists other than those who make an initial discovery.  Scientist A claims a new discovery and so scientists B, C, and D (each working independently) try to replicate the discovery and then to disprove it.  As strange as it might sound, science moves forward by a sustained barrage of attempts to prove wrong every claim of truth a scientist makes.  In fact, another name for the scientific method is falsificationism.  How does it work?

The Scientific Method

1.      A person makes observations and formulates a question.
 

2.      Further observations are made.

3.       Potential answers to the question are generated. Each potential answer is called a hypothesis.  Ideally, all possible answers to the question are considered.  In practice, it is not generally possible to know if all possible answers have been considered, since the fact that you can't think of any more possible answers does not mean there are no more. In any case, no possible answers are ignored and all possible answers are considered.

4.       Each hypothesis is tested.  The idea is to review the possible answers and reject those that are certainly false.  As we reject false hypotheses we move closer to the truth.

Hypotheses are tested with a careful rubric.
 

a.       Consider the consequences if the first Hypothesis (H1) is true. If H1 is true, then...

b.      Make a prediction (P1) based on the truth of the hypothesis.  If H1 is true, then P1 is true.

Let's say that my guidebook classifies the path to the waterfall as 'moderate' and the glossary notes that a 'moderate' hike is always less than 10 miles long.  If this first path I found (H1) is the path to the waterfall, I know that it will lead to the waterfall in less than 10 miles (P1). 

If this is the path (H1), then I predict that it will lead to the waterfall in 10 miles or less (P1).    

c.       Test the prediction.  Perform an observation or experiment to determine if the prediction turns out to be true.
 

d.       If the prediction is false, you can reject the hypothesis as false.

Suppose I follow the trail for 12 miles.  What can I conclude?  If I know the waterfall is 10 miles or less from the clearing, and I follow this trail 12 miles, I know for sure that this is not the right trail. 

If H1 is true, then P1 will be true.
P1 is not true                                  
Therefore, H1 is not true.

This is a valid deductive argument form. What this means is that as long as we are sure the first two claims are true, we can be 100% certain that the third claim is true. In other words, if the first two claims are true, it is impossible that the third claim is false.  Try it out. Insert ANY variables into the first two claims to make them true and you'll see the third claim is always true.

e.        If the prediction is true, the hypothesis is still alive --  it might be true.

Suppose I follow the trail and, after 8 miles, I find the waterfall.  My prediction is confirmed, but am I certain that this was the trail described in the guidebook?

If H1 is true, then P1 will be true.
P1 is true                                       
Therefore, H1 is true.

This is an invalid argument form.  There is no necessary guarantee that the truth of the first two claims will guarantee the truth of the third claim.

Maybe there is more than one trail to the waterfall!  I found a trail from the clearing to the waterfall but not the one listed in the book.  

4.      If the hypothesis is falsified, reject it and consider other options.
 

5.      If the hypothesis is not falsified keep it as a live option.  Make new predictions and test those.
 

6.      Continue to make observations and to consider new hypotheses.

A successful scientific theory is one that has survived all attempts at falsification - that is, it has not been proven wrong.  Another mark of a successful theory is that it has high predictive power.  Recall how hypotheses are tested. We use a hypothesis to make a prediction and we test this prediction.  If the hypothesis was inaccurate in its prediction, we reject it. If it is accurate, we keep it alive.  Successful hypotheses make accurate predictions and, further, don't make inaccurate ones.  The ability to make accurate predictions is useful.  Successful scientific theories are useful.

Predictive power is extremely important.  Humans want to be able to predict what will happen in the future in order to change the future.  What do I mean?  Consider the science of meteorology -- the study of weather and weather systems.  Meteorologists collect information about (observe) weather and formulate hypotheses that answer questions.  A question like, "what causes hurricanes," is an extremely important question. If we can answer it, we might be able to:

·         Predict hurricanes before they happen.

·         Predict where hurricanes will go after they form.

·         Predict how strong a hurricane will become.

There is no doubt that we want the ability to make these sorts of predictions.  Take a moment to reflect on the practical importance of being able to make accurate predictions. Consider not only meteorology but medicine, architecture, computers, aviation, etc.  Honestly it is difficult to overstate the impact science has had on the nature of human existence over the past few hundred years. 

Accurate predictions are, necessarily, based on hypotheses - theories describing some aspect of the nature of reality (how things work).   The scientific method is a specialized way of investigating the physical universe that is extremely effective in discovering useful hypotheses that make accurate predictions and fail to make inaccurate ones. 

 The Limits of Science

Science is powerful and successful at what it sets out to accomplish. If you value predictive power, you must value the scientific method.  It is worth highlighting a few of the rules of science -- rules we must agree to follow.

·         Scientific evidence is empirical evidence -- things that are seen, touched, smelled, tasted, heard or otherwise measured.

·         Scientific evidence must be reproducible and verifiable.  Results that can't be reproduced by other scientists are discounted.

·          A scientific hypothesis must be able to generate a testable prediction (If H1, then P1). 

·         A scientific hypothesis must be falsifiable. (If Not P1, then Not H1)

Asking whether these rules are fair is the wrong question.  To get the results science promises, these rules are necessary. Note that this is a matter of value -- a subjective issue.  If you don't value predictive power, then you might not need to value science or its specialized method. However, it is probably fair to say that most of us do put considerable value on predictive power. If so, we need to protect the integrity of the scientific method. 

This being said, it is important to recognize that science has limitations.  The rules of the scientific method which provide its power also limit its scope.  If the universe contains things that can't be measured by empirical methods, science will never be able to investigate or understand them.  Science can only ever offer physical explanations to questions of "how" and "why" and these might not be the only type of answers that people care about.