Today, we live in a world surrounded by Science. Science is literally everywhere: in the news, in magazines, in help books and school curricula...I thought I knew Science. It wasn't until my first year studying Science at UBC that I realized how little I knew (and still know) about Science. In hindsight, I am surprised that the high school science curriculum at my school did not cover one of the most central concepts of science: how it works. This is known as the scientific method. Knowing how science works is as important as knowing what "facts" Science generates. This post is a basic introduction into the how of Science.
This post is more of a reference post to later posts regarding Science in everyday life. If your interests allow, I encourage you to know more :).
Many believe they know Science. After all, we are surrounded by it. Every other day, news and magazine articles revel over the newest findings..."...new study finds..."..." study finds new way to beat belly fat!"...Science, the new answer to everything, and judging by the sheer number of magazine articles dedicated to it, especially how to achieve the body of a stick-bug.
However, as many people who believe Science is the new gospel, few know how it works. How do scientists generate new findings? Is one study enough to disprove or "prove" something? Is natural selection only a "theory"? These are questions that the layperson who knows science only from magazine articles would likely not have the answer to. And not knowing these answers is the equivalent of being fed new strange food without knowing what it is made of or how it was cooked. We are then vulnerable to whatever claims anyone makes about this new food: without the knowledge to arm ourselves against anyone's claims, we leave ourselves open and vulnerable to believing the wrong people.
I am NOT an expert in this scientific method, but hopefully, this post will help anyone still unknowing about the workings of science arm themselves to baseless claims too common in our world today and engage in one of the most central skills science has to offer: thinking critically. Below is a step by step intro to the Scientific Method (a step by step way of how scientific studies are conducted).
The Scientific Method
We will work through this step by step diagram using a very commonplace, mundane question as our example: "Do cells arise from non-living material spontaneously?"
|
An overview of the Scientific Method |
_____________________________________________
Ask a Question
Any scientific endeavour always starts with a question. Every field of science has its own questions to tackle. In the field of my study, Animal Biology, many are now desperately studying, "How does global warming affect animal species?"
This question will then lead us into coming up with ways to research it to find the answer. And perhaps more everyday questions...
"Why do I get so dizzy when I drink too much?"
"Does eating only grapefruit make someone slimmer?"
"Will this drug work to reduce cancer?"
In 1858, a man by the name of Rudolph Virchow proposed the Cell Theory, a theory that proposed that all cells arise from pre-existing cells, something we take for fact today. However, at the time, this was directly challenging a hypothesis (a proposed explanation) that cells arise spontaneously from non-living materials. So the question that was asked..."Do cells arise from non-living material spontaneously or from pre-existing cells?"
_____________________________________________
Do Background Research
In Science, you will hear a lot about "the literature". What does the literature say about the effects of tobacco on health?
On our homework, "research the literature and cite at least 2 empirical journals to support your findings".
So what is "the literature"?
The literature is basically a compilation of all the scientific findings within a field of Science. These are separated into three types of sources:
Primary sources: consists of original work, published in scientific journals. E.g. if I did a study on the effects of tobacco on the health of mice and published my findings into a journal, this would be considered a primary source, since I am the one who reports the findings.
Secondary sources: if I published my findings about the effects of tobacco on the health of mice, but someone wanted to report the overall effect of tobacco on health, he may report my research article, along with possibly hundreds of other articles related to tobacco on health. This will give us an overview of what the "literature" has to say about tobacco. This process of taking many many studies about a particular topic and summarizing them in an argument, a report etc. is also known as a meta-analysis.
Tertiary sources: gives one a very broad overview of what could be thousands of research articles to give the reader a very broad introduction to the source material. An example of this is an encyclopedia or a textbook.
Note: the definitions above are really really general, and each field will have their unique distinctions of each.
Of the above, the most reliable and informative is the primary source, although the secondary and tertiary sources are good for getting a good idea of the overall view of a topic. Before any study is done, background research must be done.
For example, if I wanted to know whether tobacco is harmful to the human brain, I may consolidate the literature and find out what other studies have found: the effects of tobacco on the rat/guinea pig/monkey brain, what tobacco is made out of, whether the individual components has been found to have an effect of the human brain... all these things, that could help us predict (or make a good guess) whether tobacco is harmful to the human brain.
_____________________________________________
Construct Hypothesis
This step happens after you have done all necessary background research. In the context of a new study a hypothesis is a prediction. That is, it predicts the results from a study I am about to do or perhaps even studies similar to mine.
In another context, it can be an explanation for something observed. For example, before Darwin's ideas became the theory of natural selection, his general explanations for animal diversity could be considered a hypothesis.
One requirement of a hypothesis is that it MUST be testable. By that, we mean that there must be some way to test the hypothesis to falsify or support it. When the test (experiment) is done, we can conclude that the hypothesis' alternatives are not true, but that does not prove our hypothesis is true. After all, we never know whether another study is just waiting around the corner to falsify our hypothesis.
If a hypothesis is not testable, then there is no way to falsify it! That would be like someone who enters a game where he might not win, but he'll never lose. "I am right, because you can't prove I'm wrong!" That just isn't fair.
So can we test if cells arise spontaneously OR from other cells? Sure we can! All we need is what Louis Pasteur used, a straight and swan necked flask, and some nutrient broth to grow cells in. Plus, we need a hypothesis...
In this case, Louis Pasteur didn't test either the hypothesis or the cell theory. He was only trying to discern which is false. Therefore, he didn't really have a prediction. But never mind that, let's move on!
|
Straight and swan necked flask |
_____________________________________________
Test with an Experiment
There are a few key things about experiments that must be done...
- There must be an Experimental and a Control condition
- The experimental and control condition must differ by only one condition
- They must differ only from the independent variable
- The dependent variable is then measured
Experimental and Control Condition
Let's say you were in a lawn... you wanted to test if a fertilizer is going to make it greener. You spray the fertilizer over the entire lawn. 4 weeks later, your lawn looks like Heaven made a visit. You think to yourself, "That ridiculous price I paid for this bottle of fertilizer was worth it after all!
But the thing is, you don't know if it was really the fertilizer! If you had sprayed the fertilizer in late March or early April, there's a good chance your fertilizer wasn't the one that did the magic, it was summer. Or perhaps, it just happened to rain more over the four weeks than it did the entire year. Or better yet, maybe the weather was beautiful and dogs who were being walked decided your lawn looked spectacularly inviting for their business.
The point is, you have no idea whether it was the fertilizer or not.
So what SHOULD you have done? How about spray the fertilizer on one side of the lawn and leave the other untouched? Assuming rain or summer doesn't prefer one side of the lawn over the other, the two sides are relatively equal. Therefore, the only significant difference between the two sides should be only the fertilizer.
Experimental Condition: the condition that receives the "treatment" (in this case the side of the lawn that receives the fertilizer)
Control Condition: The condition that does not receive the "treatment"
Why is this important?
Because only one side receives the condition and the other doesn't, the one that doesn't provides a base condition for comparing the two groups. Therefore, any difference between the two groups can be attributed to the "treatment"
Independent and Dependent Variable
I'll keep this short... the independent variable is whatever is manipulated by the experimenter (you) to make the difference between the experimental and control group. However, the dependent variable is what is then measured to evaluate the difference between the two groups (probably the "healthiness" of the lawn).
Pasteur's Experiment
Pasteur's goal was to determine whether exposure to pre-existing cells would or would not cause cells to arise.
Therefore...
Independent variable: exposure to cells
Dependent variable: whether the inside of the flask will have cells in it
In order to control for the exposure of cells, he used two flasks (see above image). One of the flasks would be exposed to the atmosphere, where many cells exist. The other is a swan-necked flask, which traps cells in the swan neck so that the broth will NOT be exposed to pre-existing cells.
Experimental Condition: straight necked flask
Control Condition: swan-necked flask
Note: the "treatment" in this case is the pre-existing cells
_____________________________________________
Analyze Results
Once you are actually done the experiment, you can then analyze your results.
For Pasteur's experiment, the results were this...
Straight necked flask: cells grew in broth
From this, we can conclude that both the cell theory and spontaneous generation hypothesis is supported, since we cannot discern between whether it was the pre-existing cells or the non-living material (the broth). However...
Swan necked flask: cells did not grow in broth
This conditions serves to falsify the spontaneous generation hypothesis since only the broth but no pre-existing cells could get into the flask.
We can then draw our conclusion: that our data supports the cell theory and does not support the spontaneous generation hypothesis.
Operation Lawn Experiment
We look at our lawn after spraying one side but not the other. The side that was sprayed was indeed greener!
Can we then conclude that the fertilizer did indeed have an effect? Not quite. In these experiments, where we are concerned with not whether we HAVE greenness or not, but the DEGREE of greenness... there is always the off-chance that something affected our results. Maybe the dog did indeed poop on one side but not the other. In this case, one study is not enough, we must REPLICATE, or repeat the experiment several times.
In this case, we use something called the standard deviation. Let's say there was a way to calculate how healthy our lawn was, perhaps by the length of the grass.
I won't go into detail... through a set of statistical steps, including calculating the average length of each set (treatment and control), then calculating variance (for both treatment and control)...that is about how much the data VARIED from the average
|
where u is the average and N the number of scores (repeats) |
We can then calculate the 95% confidence interval. The 95% confidence interval, once calculated, allows us to decide with a 5% chance of error (which is pretty darn small), whether the differences between two or more samples (replicates of an experiment) are due to random chance/sampling error/human error.
So when scientists say that there is a significant difference between treatment and control, it typically means they are at least 95% confident that it was due to their treatment, not by chance (like dog pooping on lawn).
Interesting Side-note: this language difference between Science and everyday language has been used by the media to draw the conclusion that global warming is disclaimed. A scientist once said, "the data is almost significant, but not quite". He was referring to the trend of increasing temperatures and how likely it was attributable to global warming. What he was actually saying was that the data was, instead of the conventional 95% + confident, was instead 92% confident. The next day, the news were filled: "Scientist disclaims global warming".
_____________________________________________
Hypothesis SUPPORTED/NOT SUPPORTED
Remember, even if our hypothesis is supported, we must repeat our experiment. An experiment may have to be repeated dozens and dozens of times before the results are published.
If it is NOT supported, then we remodel our hypothesis and start over. In other words, we change our original hypothesis (guess again) and try again.
Either way, the results can be published.
_____________________________________________
Publish/announce Results
Then you publish your results. Whether your study was used to support/disclaim an existing/newly made hypothesis, the results are there.
If a hypothesis has garnered enough empirical support (in other words, its stood up to a number of challenges), it can become a principle of the field... in other words, it can become a theory.
Theory: "an explanation for a broad class of phenomenon or observation which has been verified to some degree (i.e. experimentally)".
Creationists have sometimes used this to dispel the Theory of Natural Selection (in other words, evolution). I have nothing against creationists as they are entitled to their beliefs, however, if I were in a debate with someone who happens to be a creationist, I would warn him/her that they're argument is not valid.
Although in everyday language, a theory is synonymous with a "guesstimate". E.g. to paraphrase from the Pink Panther..."I have a theory about who stole the Klopman Diamond, but its only a theory. However, in Science, a theory is far more than a guess. It is probably one of the most powerful statements and the closest something can get to "fact" in Science. It is something that has not only withstood tonnes of challenges from individual studies (receiving lots of support), it is also, by definition, an explanation for and a predictor of a broad class of phenomenon or observations. For example, the Theory of Natural Selection is one of the founding principles of Biology and dominate almost everything biology. To make the point more obvious, here are a couple of well known theories:
- Theory of Natural Selection
- Explains animal diversity by evolution, adaptation of animals to different habitats, distribution of characteristics (such as beak length) in a population of birds in different areas etc...
- Predicts that if we were to leave two populations of the same species in different enough habitats, that their characteristics, over a certain period of time will differ; predicts what would happen to the population dynamics (beak length across the species, weight etc.) of finches with certain changes in the environment (hotter weather, drought etc...)
- Theory of Gravity
- Explains why objects fall to the ground, why the earth orbits the sun (Theory of Heliocentrism) and why walking under apartments can be potentially dangerous
- Predicts how far we must shoot a rocket into space in order for it to orbit the earth, not crash into it or be shot out into space; predicts how much support is needed for a building to not crash to the ground
- Theory of Heliocentrism
- Explains why we have seasons, explains why the stars in the sky move the way they do
- Predicts what day will be the best day to get a tan.
The above theories, especially the last two, are typically taken as "fact" rather than "theories". In Science, however, a theory is probably as close to "fact" as it can get.
I hope I'll have the time to explain real world applications to the above that I have written, including why we should be VERY cautious about what the media has to say about "Science", how to really decipher fad science etc.