The Scientific Method

Clickable Index of the Scientific Method
Observation
Question
Hypothesis
Prediction
Experiment
Analysis
Decision
Some Examples

As an early part of our study of a science, we need to all be "on the same page" when it comes to a few terms. We can define science as a methodical approach to the acquisition of knowledge. This important word distinguishes how a scientist works from how other people learn about the world. Science is an approach that is methodical, and that approach helps acquire knowledge. Science is not the knowledge gained through the approach. Knowledge can be gained through a variety of ways, but science acquires knowledge methodically.

The method of science is a pathway that involves several steps. Scientists themselves might organize the pathway in slightly different ways, but scientists would agree that what is presented here is one format of the scientific method.

The scientific method is based upon evidence rather than belief. This distinguishes science from faith. A scientist is suitably skeptical of anything but good evidence. That is not to say that scientists lack faith...it is just that faith for them operates in a different sphere of their lives. In scientific work there is little room for faith; in life there is plenty of room for both.

Observation

The scientific method, then, is founded upon direct observation of the world around us. A scientist looks critically and attempts to avoid all sources of bias in this observation. But more than looking, a scientist measures to quantify the observations; this helps in avoiding bias. Which of these lines is longer?

In fact, neither is longer if you measure them, though human bias might generate belief that the one on the right is longer than the one on the left. The arrowheads on the lines "trick" the human integrating system, so an accurate ruler is required to avoid bias.

The system of measurements used in this observation part of the scientific method is the metric system. Now many people in the US get all upset when this system is mentioned as a replacement for the English system currently in common use. But I am here to tell you that there are two important reasons to use the metric system.

First, the metric system is universal. All scientists agree on what constitutes the measures taken within this system. All scientists worldwide use this exact same system. This means that scientists can compare results they obtain with results obtained by any other scientist without conversions or resulting errors. The lack of metric system use in the US is really an aberration; the rest of the world uses the metric system. America is more English than England in sticking to the out-moded English measures. The government, realizing our position as "odd man out" tried in the 1970s to convert the US to metric. It later gave up on our citizens as "too stupid" to understand it and repealed the act.

This of course was wrong because the government went about conversion the wrong way. Humans learn measures by measuring not by converting. So the dual labeling and memorization of conversion factors was a total failure. Metric-only labeling would have guided people to learn the new system by sensing rather than converting. You will use this method in laboratory to learn the metric system.

Second, the metric system is simple! If you were not riled up by my previous remarks, this one should make your blood pressure rise. How could a system be simple if the entire American population failed to learn it?

The metric system has only one basic unit in each category of measures. This basic unit is converted to smaller or larger units by using powers-of-ten multipliers.

Measurements might be made
in these basic units:
  • Length - meter
  • Volume - liter
  • Weight - gram
  • Temp - degrees Celsius
  coupled to a set of modifying prefixes
that work for all basic units:
  • kilo = 1000
  • centi = 1/100
  • milli = 1/1000
  • several others (less important to us here)

So, once you know what a meter is and learn the prefixes, you can measure from here to wherever quite easily and precisely. Small items might be measured in millimeters (about the thickness of a dime); larger items might be measured in centimeters (about the width of one of your fingernails; huge items might be measured in meters (about, well, you will learn it in lab--your hands so far apart); map distances would be measured in kilometers.

Compare if you will with English units. In typography we use points and picas (how long is that?), for small items we would use fractions of inches (now would that be tenths, eighths, quarters, or what?), for larger items we use inches, huge items we measure in feet (now how many inches in a foot?), longer distances require yards, rods, or miles (Gosh, how are all of those related? Let's see...12 inches in a foot...are there 12 feet in a yard? Nope, just 3. OK, how about rods, 12 yards in a rod? Nope. 3 yards in a rod? Nope. How about that mile? 12 yards? 3x12 yards? 12 rods? 12x3 rods? None of the above). I guess that English system is not really so simple or easy to remember. In fact most Americans do not have a clue about most of the English system of measure...they are just VERY familiar with a few of those measures.

Now let us move on to volume. The metric unit is a liter. Now this was the only success story of the metric conversion act. The soda/pop bottling industry pushed very hard in advertising its 1, 2, and 3- liter bottles and they noted that the liter was just a bit MORE than a quart. Americans thought from this that they were getting some extra for nothing and so learned quickly about the liter. Americans have a real tactile sense of what a liter is now. The good news is that we can measure small volumes in milliliters, moderate volumes in liters, and huge volumes in kiloliters...and the prefixes mean the same multiplier as in the length case! So to switch thinking from length to volume you only have to learn ONE basic unit and nothing more!

To finish this little story, let us think about English volumes. Well for small volumes there are drops, drams, and fluid ounces. For larger volumes we have cups, pints, quarts, and gallons. For huge volumes we have two different sizes of barrel. So what are the relationships between these? What is a dram? Is it 3 or 12 drops or 3x12 drops. Nope. So how many ounces in a cup? Not 3 or 12, but 8! How many cups in a pint? Not 3, 12, or 8, but 2! There are 2 pints in a quart (wow at least two consecutive measures with the same multiplier!) but there are 4 quarts in a gallon and 55 gallons in one of the two barrel sizes. What? 55? What a mess. The English system of measures is extremely complex as you can see. Worse, it is not even universal (a barrel might not be a barrel). A good example is the gallon. In the US a gallon is one volume but in Canada it is more! And Americans do not even know much about it! The Canadians were smart enough to abandon their gallon in favor of the liter!

I will relate one personal story before we leave this topic. I was on a three-week science exchange program in the People's Republic of China. We delegates loved eating Chinese food...for the first week or so. But after that, the greasy sauces and the duck-head and fish-head staring at us from the platters were getting to us. Some of us had stopped eating very much. One morning a British delegate sat down to breakfast next to me and remarked "I'm in terrible shape--I've dropped a stone!" I was very sympathetic. "Oh, I'm so sorry! I hope it wasn't too painful!" My grandfather had told me about excruciating pain when he passed his kidney stones. The delegate looked at me strangely but continued "By the end of this trip I shall have lost so much weight I'll be able to fly home without the plane." This seemed not to follow from the kidney stone idea I had in my American mind. I must have looked puzzled (I was!). He went on to explain to me that British scales are often calibrated with an official weight measure called a stone. This is an English unit (equal to 14 lbs) that Americans know nothing about. He was merely remarking about his weight-loss! OUCH!


Question

The second step in the scientific method is to formulate a question. Scientists have to be curious. Humans are naturally curious, visit with a three-year-old sometime and you will see what I mean. Unfortunately parents and school teachers get tired of answering questions (we all need patience and want it right now!) and so the natural curiosity of children it kicked right out of them. In some schools, for example, the children are forced to sit quietly in neat rows of desks with their idle hands folded neatly on the desktop. While this might be some form of discipline and important lessons in conformity are needed, this style of classroom is antithetical to science. An effective science classroom is filled with hands-on activity and lots of questions. It involves a productive noise!

In this course, please let go of your cultured inhibitions somewhat. Be curious, ask questions! There is one truly foolish question...the one you internalize and never answer! The uninvestigated question is usually common to many people in the room and if all of them stifle themselves with fear or whatever, no one learns from it. Ask that question! We will work together toward an answer.

You want your question to be answerable. Science can answer many questions, but there are some which cannot be answered by science. An example might be: why am I here? The word why implies purpose, and begs an answer from a creator. This question cannot be answered by science as we cannot test a creator for humans by the means available to science. This question is one that can only be answered by faith. So you might find some questions are not very appropriate for science, but some rewording might make a question answerable. For example if you change "why am I here" to "how did I get here" it becomes answerable by science. Why humans are on the planet is a matter of faith, but how we evolved has been studied and much is already known about it. We will not go into that here. But before I leave this idea I want to pose one more question. "Which came first, the chicken or the egg?" Now many people think this is a really difficult question to answer, or even claim that it has no answer. They are wrong. Biology can give you a very clear-cut answer to this question. The egg is a very ancient biological entity. Even the lowest of animals...yes even PLANTS have eggs. The dinosaurs had eggs. Chickens, on the other hand, are a distinct species of birds. Birds evolved from dinosaur-like ancestors long after there were eggs, and chickens are not even an ancient kind of bird. Thus the egg came LONG, LONG before there was a chicken of any kind. Rephrase the question slightly..."which came first, the chicken egg or the chicken"...and your question is at least a bit more difficult.


Hypothesis

The next part of our scientific method is to form a hypothesis. This is merely an educated guess as to the answer for the question. You examine the literature on the subject; scientists need libraries, reading is critical to scientific performance! You gather as much book knowledge as you can on the subject to begin to arrive at an answer to your question. This tentative answer...this best educated guess...is your hypothesis.

Please notice that hypotheses do not always have to be correct. In fact most of science is spent trying to determine the validity of a hypothesis, yet this effort is NOT likely to give a single perfect answer. So, in formulating your hypothesis, you should not worry too much that you have come up with the best or the only possible hypothesis. The rest of the scientific method will test your hypothesis. What will be important is your decision at the end of the method.

The one aspect of your hypothesis is important, though. It really must be rejectable. There must be a way to test the possible answer to try to make it fail. If you design an untestable hypothesis, then science cannot be used to help you decide if it is right or not. For the moment, let us say that your question is "Is God awake?" and you have made the hypothesis "God is awake." There is no way to test the slumbering state of God scientifically. Switch the word God to, for argument, Ross Koning, and the hypothesis is testable. Sleepy yet?


Prediction

The prediction is a formal way to put a hypothesis to a test. If you have carefully designed your hypothesis to be sure it is falsifiable, then you know precisely what to predict. The prediction has three parts:

  1. If my hypothesis is true...
  2. Then _____ should happen
  3. When _____ is manipulated

The manipulation is what you knew would likely falsify your hypothesis.

  1. If Ross Koning is sleeping...
  2. Then his breathing will remain slow and even
  3. When I brush his cheek with a feather

The blanks in the generic format for the prediction above, represent what are called two variables. The first blank above is for the dependent variable and the second blank above is for the independent variable. The independent variable is the one that you manipulate and the dependent variable is the response that you measure. So in our example, the independent variable is the feather brushing the cheek...and the dependent variable is the slow and even state of breathing.

This part of the scientific method is the key to testing the hypothesis. If this prediction holds then you will not be able to reject your hypothesis. If this prediction does not hold, then you will reject your hypothesis.

Rejecting the hypothesis is usually the desired outcome as we shall see...


Experiment

This is the actual hands-on part of the project. Here you carry out your manipulation and compare the results with results from a control setting. Our sample project gets tough here. We can find Ross sleeping daily and we can try the feather trick on several occasions when we are sure he is sleeping (how do we know? that's the point of the project, no?). We can stroke him when we know he is awake (there ARE symptoms for that!). We can measure ventilation (inspiration + expiration) rates easily. We can compare those results with what we observe during the actual test.

We really have to know how deeply Ross does sleep, however. Some people will waken even with the slightest touch...others sleep through alarms, smoke detector alarms, thunderstorms, and ignore the touches of their significant-other. We have to know what Ross' sensitivities are before we proceed. We might need to change our feather for a pine cone or maybe a wood rasp!

We cannot go on to a decision with just one observation. We desperately need to carry out several touches to decide. Of course if the touches are too close together in time, we might be rousing the sleeper, so we have to plan carefully.

To be an experiment, we must compare the results of some manipulation with the results of an unmanipulated (control) situation. Not everything we do in science compares a treatment with a control, but most useful information is derived from experimental science.


Analysis

How do we compare the results? As good scientists we will try to repeat (replicate) our experimental treatments several times to avoid chance error. But once we repeat, we may get a mixture of "positive" results and "negative" results. How will we know which results are typical or correct?

There are many sources for error. Ross might not be paying attention but nevertheless is awake. We might have touched him so lightly that he would not respond even if awake. Certain parts of the body are more sensitive to touch than others. A sleeping Ross might awaken for other reasons just at the time we touch him. So there are chances for false positive results and false negative results.

Statistical analysis is designed to help us answer our question by assessing results to minimize false positives and false negatives. I will not go here into lots of details about hypothesis testing with statistics, but I will say that all statistics can do is provide you with a measure of how probable your answer is. Statistics does not give perfect answers, but it gives you an estimate of how wrong your decision might be.

In most statistical procedures in biology, we will allow 5% error to occur before we start changing our minds (to minimize both kinds of possible error). This means that our project can fail 1 time in 20 repeats and still be considered viable (1/20 = 0.05 = 5%). This much error is accepted as "due to chance alone." In a court room we might use the terms "reasonable doubt" here.


Decision

Here we use our estimate of error and the allowance for error (5%?) and make up our minds. We have two options: "reject the hypothesis" or "cannot reject the hypothesis" and only these two options! If the chance we are wrong is more than 5% (it failed more than 1 time in 20 tries) then we usually reject the hypothesis as flawed. If the chance we are wrong is less than 5% (it failed less than once in 20 tries) then we cannot reject the hypothesis.

Please notice that we do not prove hypotheses! Proof exists when the chance for error is 0. There is always some chance for error (no matter how small it is!) and this existence of chance error means we cannot prove anything in true, honest, science.

The words "scientific proof" therefore constitute an oxymoron (think: "Little Giant"). Advertisers are either scientifically-challenged or consider the American population incredibly gullible. This oxymoron abounds on US television advertising. Viewers should question the validity of any claims in advertising that includes such oxymorons. How good can the science behind the advertisement be if they do not know this critical and elementary point of science? The credibility of such ads should be exceedingly low!

Speaking of credibility, should a scientist worry when the hypothesis is rejected? Certainly not! The scientist generally has several possible hypotheses in mind that relate to the question at hand. Rejecting one hypothesis eliminates one of the hypotheses and thereby brings the scientist one step closer to the truth. In fact, the scientist is usually disappointed when the hypothesis cannot be rejected because one of the possibilities has NOT been eliminated and so little progress has been made. The same handful of hypotheses are still "in the running."

So, rejecting one's hypotheses does not make for a bad scientist... indeed as long as the justifiable decision is made, the scientist is performing correctly.



Some Examples

Here I give you a sketchy outline of several cycles through the scientific method in an attempt to arrive at the truth in an everyday situation.

The situation is this:

You arrive home late at night, walk up to your house door, unlock the door, reach in to the light switch just inside the front door. The light does not come on! Now what?

As a normal human being, you will go through a mental and physical process of hypothesis testing. The steps happen very rapidly in your mind and, prior to this, you may not have had names for the various steps. Nevertheless, I hope you will recognize what your brain is doing as you stand there in the darkness. You are already a scientist as you will see, you just did not know it!

Observation: Night, Come Home, Switch On, No Light....we are "IN THE DARK"

Question: Power Out?

Hypothesis: Power IS out!

Prediction: If power is out, then light is out at all neighbors, when I look

Experiment: Manipulation: switch on, no light.
Control?=neighbor's lights, street lights

Analysis: If ANY house with light, prediction fails, decision = reject?
error? Coleman Lantern, Generator
If all houses dark, prediction holds, decision = not reject?
error? All out to dinner, movies, fireworks?

Decision: above, note chance error = no definitive answer...NO PROOF!


Do you want to wait in dark for all neighbors to come home?


Question: Switch broken?

Hypothesis: Switch IS broken

Prediction: If hypothesis is true, then there should be a brief flash, when I flip switch several times

Experiment: flip switch...several times (brain is doing a statistical test and knows when to stop the replications!)...not even a flash

Analysis: Too broken to even flash? needs a stronger test, replace switch?
Do switches fail frequently?
Maybe another hypothesis is more likely?
abort...no electrical work in the dark!


Question: Light bulb burned out?

Hypothesis: Light bulb IS burned out!

Prediction: If hypothesis true, then light will come on, when I install new bulb

Experiment: Grope in closet in dark or rob another lamp...install bulb...LIGHT!

Analysis: Original bulb would not light, new bulb does light

Decision: Cannot reject hypothesis!
error: NO PROOF...more testing needed!


Prediction: If hypothesis true, then bulb will tinkle, when I shake it

Experiment: Shake it, control is new bulb

Analysis: No tinkling!

Decision: Reject hypothesis! NOW WE ARE CONFUSED! More testing!
error: wire not broken or wire broken in only one place (no tinkle!)


Question: Bulb loose in socket?

Hypothesis: Bulb was loose

Prediction: If hypothesis true, then light will come on, when I re-install it

Experiment: Tighten it...It lights!

Analysis: Can we be sure that it was loose?

Decision: Cannot reject.....NO PROOF!
Error: maybe power just came back on...
switch is weirdly intermittent... or.....


PAST EVENTS CANNOT BE TESTED!!


Hypothesis: Genie of the lamp was originally displeased with us...
after all the cord stroking, bulb changing, switch flipping,
Genie is now happy with us so it lights?

Hypothesis: Genie of lamp not listening for requests
(Genie asleep or, worse, dead)
(Thank goodness we did that CPR)

We cannot test these last hypotheses because Genies cannot be manipulated scientifically. Worse, whatever happened to cause the initial failure, occurred in the past and we cannot go back in time to run tests. So we cannot eliminate the Genie in the Lamp ideas. But the evidence leads us to the ultimate theory: The Bulb Was Loose In the Socket!

A theory in science is an idea that has been tested thoroughly, and despite extensive testing, cannot be rejected. It is as close to the truth as we can get while still admitting that we cannot eliminate the rest of the possible hypotheses (Genies and such).

EVOLUTION is a theory exactly like this. It is an event that happened in the past, so we cannot know for certain precisely how it happened. Thus there is room for error (however slight) and alternatives (even if highly bogus), and so we cannot prove that evolution occurred. But this theory has been tested from many points of view (not just fossils!) and never has been found to fail to explain what we see in the biological world. Because of this extensive testing and lack of failure, it is as close to fact as we can come in science. We thus give it the special name of theory.

Unfortunately, in modern vernacular, the word theory has a completely opposite meaning. Our news media reports that the theory of evolution is speculative (the vernacular meaning of the word) and that scientists have some doubts (less than 5% but > 0% error) about it. Now you know that our level of doubt based on extensive testing is vanishingly small. As honest scientists we cannot say PROOF, but the theory of evolution is as close to proven as any idea in human thinking. That includes such ideas as "we are here." In Stephen J. Gould's words, evolution is a fact. There is essentially no doubt about evolution as the mechanism of creation of life and species on our planet.


Some Personal Observations

 

This page © Ross E. Koning 1994.

 

 

Go to the Course Schedule Page.

 

Go to the Plant Physiology Information Homepage.

 

Send comments and bug reports to Ross Koning at koningre∂gmail⋅com.