- [Voiceover] Now run the experiment, and observe what happens. When you run the experiment here comes your truth or a falsehood. You'll either uphold your prediction saying it was true. It worked, or you realize that your prediction was incorrect. That either the experiment had a flaw. Something didn't work, or simply, your hypothesis was not upheld. Your hypothesis was not valid. But to make this work, the experiment has to be very carefully planned. In controlled experiments, you account for every variable except for one. If we go back to the example of taking a drug, an Ibuprofen, Tylenol, whatever you're taking. Pick one drug, and one drug's important. Not many, just one. Then okay, you take the drug. Do you take it with water? Do you take it with juice? Is it with a full stomach, an empty stomach? Those are variables to think about. Do you take it before you go to sleep, when you wake up in the morning? Other variables are also age. Variables can be gender. Variables can be how long the individual had a fever for. Variables can be how high the fever is. So all these are variables that you need to try to account for to try to keep it as uniform as possible and even then, you still wanna think about using a placebo or a false treatment, because the mind is a very powerful tool. A lot of times if someone takes a drug that drug might not actually help them, but because they think it should help, it will help them. It's called the placebo effect. So what happens a lot of times in experiments, in controlled experiments, one group of individuals or one group of items is given this false treatment. Probably one of the easiest ways to think about this is a sugar pill. In clinical trials, a group of individuals is still given a pill, but their pill is just pure sugar. There's no drug in it. Nothing can actually, physically, chemically change what's going on, so this placebo, it's looking at does the person get better just by thinking they should get better? So it's your placebo effect. Well once that's done, once you've run your experiment, if it held up the hypothesis, if it was true, great. You're good to go, but that very rarely happens in the first time. Usually what happens is you modify the hypothesis. You go back and repeat. You make up a new experiment and that new prediction, and you go through and test everything. You keep repeating until your results are held up as a truth. Until your hypothesis and prediction are not shown to be false. This might be repeated once or twice. You might repeat this dozens of times. If you're thinking hold on. You said earlier that we use the scientific method to figure what clothes to put on. I knew what clothes to put on the very first time, and I didn't have to repeat any process. True, but was that the first time you ever decided to put clothes on and go outside? No, I bet not. Probably when you were a child and your parents first allowed you to pick your own clothes, I don't know if every single time you picked the right way. Maybe you really, really, really wanted to wear shorts in the middle of winter. Well what happens then? You walk outside. You realize you're cold. You either do one of two things. You deal with it because you don't wanna show that you're cold, could be stubborn. Or you go back in the house realizing you're cold. You failed the experiment because you are cold, so you go back and change the clothing to put warmer clothes on, go back out. Now you're warm. Your prediction of warm clothes keeps your body warm is now upheld. So even with the most basic idea of what clothes to wear, you have at one point in your life, run the experiment a couple times in a row to figure out what to wear. So if that prediction is false, the hypothesis must be modified, so change it until you get one that is upheld. When that prediction is true only one small part has been tested. Going back to that drug trial again I set up. Even if you did test it with everyone at the same exact temperature fever, with the same amount of time being ill, with the same illness, because even different illnesses could cause problems here. You might say okay, hey this drug worked on this particular illness. Is that further testing? Sure. Are there illnesses, other temperature fevers, higher and lower, different age groups? So even if it's true, you still have to keep testing other variables. A hypothesis cannot be proven true. It can only be supported 'cause no matter what, you can never say a drug will reduce fever 100%. It can be proven to be true because all of a sudden sure, yeah I saw that, no problem, the fever went down. In this particular instance. So you support your hypothesis of what you thought. Now go back, change it up a little bit and test it again with a slightly different variable. So you kind of think of it as this big round robin here. You start with number one observe and generalize. No problem. Then you form a hypothesis, make a prediction, experiment and observe it and then get to the modify and repeat as necessary. Odds are the very first time is probably not gonna be supported, so you modify the prediction, you modify the experiment, test again. Well this one might have been upheld, but good chance it probably wasn't, so you go in again. Go through and test the prediction, test and experiment again and keep going over and over until your hypothesis is upheld. Each time you retest it's a better chance the hypothesis will be supported. One of the biggest things, one of the key things whenever you're sampling with living organisms, you want a larger number of individuals. A larger number of organisms. 'Cause if you test just two people well, there's a good chance two of them will be positive, but the more individuals you have the better chance of having a better representation of the entire population. What you do is you take that grouping of subjects, and you split it into at least two groups. These two groups should be random. Do not pick based on age, or gender, or anything. It should be basic assigning a number to each person then randomly picking those numbers out and put 'em in two groups. And you treat each of these groups identically. I mean perfectly equally. The only difference is there's one variable. The experimental group gets the treatment. The control group, the one you're testing against, gets a placebo. It's important now because it shows you, it takes the mental idea out of the picture. Because you can see what the mental idea of getting a treatment is, but not really getting it. Placebo is nothing's happening. You're giving a false treatment. In this case they're testing blood pressure. Did the treatment group, the experimental group, end up with a lower blood pressure than the control group? If the answer is yes, then you support a hypothesis. But if it's no that means that probably, just the sheer mental idea of getting a treatment was just as effective as the actual treatment. Which means that treatment isn't really that effective at this point. It needs a little more modification.