Читать книгу Geology For Dummies - Alecia M. Spooner - Страница 31

Now the real fun begins: experimenting to determine if one of your hypotheses is correct. Scientists use their hypotheses to develop predictions that can be tested. Based on the observations about the color of dog fur, a prediction could be this: “I predict that all dogs have either brown fur or white fur with black spots.” The prediction is a restatement of my hypothesis, based on my observations.

To determine if my prediction is correct I need to collect more information. I will make new observations, but this time I will manipulate the situation and observe the outcome. In other words, this time my observations will be the result of an experiment.

In science, the experimental design, or the way you go about collecting the new information, is very important. An experimental design describes the parameters of your experiment: how many samples you will take (how many observations you will make) and how you will choose those samples. These decisions are partly determined by the question you are asking and partly determined by the nature of the observations you are collecting.

In most cases it’s impossible to observe every single instance of the physical world that you are exploring. Therefore, you must take a sample that can represent the rest. For example, I can’t look at every dog in the world to see what color their fur is, so I may decide that looking at 100 dogs will provide me with enough observations to determine if my prediction is correct. Those 100 are a sample of the worldwide population of dogs. If I choose those 100 dogs wisely, they may be a very accurate representation of the worldwide dog population. The best sample size is different for each experiment; it all depends on the question being asked.

In earth science, experiments are often natural or observational. This means that scientists go out into the field and observe events that have already happened, such as the formation of rocks, rock layers, or features of the landscape. Scientists make these observations without changing any aspect of the event or its result.

Geologists also use another kind of experiment called a manipulative experiment. A manipulative experiment is done in a laboratory, where the scientist can manipulate or change certain factors in order to test which factors are most important in creating the observed outcome. In this case, multiple experiments can be done, each one testing the importance of a different factor (or variable), with the goal of zeroing in on the one (or ones) that explain the observed outcome.

Most importantly, a scientific experiment, whether it is a natural or manipulated experiment, must be repeatable. This means that the scientists must clearly describe the steps they have taken so that another scientist can repeat the same experiment and see if she too, gets the same result.