Science: An Elementary Teacher’s Guide/Hypothesis testing, data collection, analysis, and publication

The Scientific Method
The "Scientific Method" is a group of ideas and techniques for solving problems, making discoveries, and proving old ideas false. It can be very difficult to use the scientific method fully, but even applying it only partially can be helpful in school, the lab, and in day to day situations. We use parts of the scientific method intuitively--we come up with ideas for why something is not working, and we test various solutions; we see something unexpected and we come up with possible explanations. To the extent that we think logically and critically we are using the methods of science. The full application of the scientific method is complex and involves more than following a step-by-step process like you would a recipe. Science is especially useful when answers are not obvious or there are multiple possible explanations--science is the method by which we can come up with the most reliable answers to questions. Sometimes specialized equipment, such as microscopes, telescopes, or spectrophotometers, may be needed to test hypotheses or gather data so it may seem that science can only be done by scientists. However, learning science as a method of inquiry can help students learn to think more clearly and critically and to carefully consider competing ideas based on data.

Curiosity is fundamental to science--nobody will do the work of finding an answer if they are not curious about the subject. Children are naturally curious and full of questions--educators can help cultivate this curiosity by not simply providing an answer but instead teaching how to discover answers through research and experimentation. Working through the steps together is more helpful than simply teaching the steps. It is important to keep in mind that science can be difficult and messy. Not just messy in the literal sense, but also messy in the sense that your experiment may not work as planned and your data may not have a clear pattern. Even though you set out to answer a question, you may be unable to answer it fully, and sometimes not at all. On TV it seems like science is a collection of proven facts--in reality science is a method to discover truth while trying to not fool ourselves. It is full of uncertainty and is ever-changing.

Here are the basic steps of the Scientific Method, but keep in mind that it rarely flows smoothly from step to step, and it is rarely done by a single person in a short time.


 * 1) Identify the problem- Ask questions about your subject/problem that you are observing: How, What, When, Who, Which, Why, or Where?
 * 2) Consider current information- Do background research on your subject. Use sources such as databases, internet sources, books, etc.
 * 3) Form a hypothesis- A hypothesis is an educated guess about your subject. It is a trial to conclude your question about your subject that          can be tested.
 * 4) Do experiments- Test your hypothesis with an experiment. This will prove whether your hypothesis was accurate or not.
 * 5) Analyze the new information- Once your experiment is completed analyze your new information to see if your hypothesis was right or not.
 * 6) Repeat steps 3 through 5, if necessary- If necessary to repeat construct a new hypothesis and follow the process again.
 * 7) State a conclusion- When ready, state your conclusion and record your results.

Some key principles of scientific thinking include empiricism, which means using only information that can be measured and verified, rationalism, which means thinking logically (this is harder than it sounds--we are not born as logical creatures, but instead naturally use emotional thinking or wishful thinking and can easily fool ourselves or be fooled by others), and skepticism, which involves always questioning ideas and beliefs to see whether or not believing in them is really justified.

Forming a good hypothesis
This can be very tricky! A hypothesis is often defined as "an educated guess," and most people are content to stop there, assuming that their educated guess is correct. A scientist is skeptical of their own ideas, so the "best guess" cannot be accepted without evidence. For a hypothesis to be useful in the scientific method, you must be able to test it--and to test it, it must be falsifiable. In other words, once you have a hypothesis, you should be able to make certain predictions based upon your hypothesis--"if this hypothesis is correct, then 'x' will happen when we do this experiment." You then perform the experiment, and if your prediction was wrong, your hypothesis was wrong (or you did a poor job collecting your data). Sometimes you cannot perform an experiment (there is no way to move the stars around, for example), so your "test" may simply involve collection of more data. If you cannot collect data that would disprove your idea, then it is not actually a hypothesis. This is why science can only function in the realm of what is observable--if you say an invisible, undetectable space creature is causing the wind to blow, then we cannot disprove your idea; hence it is not a hypothesis. For this reason science cannot operate in the realm of the supernatural.

If you have conducted your experiment and the data support your hypothesis, you have not proven that your hypothesis is correct--there may still be other possible explanations (you have tested only the explanation that you thought was the best). You have "failed to disprove" your hypothesis, or you could say your hypothesis is supported by all available data. You or others may think of new ways to challenge your hypothesis. It is only after withstanding many, many challenges and being supported by large amounts of experimental data that a hypothesis reaches the rank of "theory." In everyday language people say "I have a theory about this," but in scientific language all they have is an idea (maybe a hypothesis, if it is testable). Some of the well-supported theories of science include gravity and evolution (both are well supported by thousands of studies, yet questions still remain and new experiments continue to refine our understanding of these phenomena).

Data collection
Again, this can be tricky. Your test results can only be accurate if your data collection is accurate. To get accurate data may require a tool or instrument as well as the ability to properly use it. If you were asked to measure the height of a plant with a ruler, that seems very straight forward. However, if you ask 10 people to all measure the height of the same plant, using the same ruler, it is unlikely you will get the same answer 10 times. This would be a good thing to do in the classroom, because it introduces the ideas of uncertainty and variance and degree of accuracy.

Experimental design
There are many possible variables that can effect the outcome of an experiment, and you want to control as many of these variables as possible. For example, lets imagine you were growing bean seeds as a class and you wondered about how the amount of water affects the rate of growth. What other things, besides water, might affect the growth of your seeds? Some possible ideas might include the amount of sunlight, the temperature, whether or not there were bugs that liked to eat bean plants, what type of bean seed you were planting, how compact the soil is, how deep you plant your seed, etc. To get the most accurate data possible, you want to control as many of the other variables as possible by keeping them the same between your groups of seeds. You might explain the importance of starting with seeds from the same pack, and maybe seeds that are similar in size to each other. All the soil should be mixed up together and the same amount of soil used in each cup. The low, medium, and high watering groups should all be kept in the same general area so they get the same amount of sunlight and the same temperature. And when the plants are watered the amount of water should be carefully measured so each plant in a group receives the same treatment.

Data analysis
Sometimes the data are so clear that the answer is obvious. Perhaps with the bean growing experiment the average height of plants in each group could be drawn as a bar chart, and maybe there will be a clear difference between the groups. Usually the differences are not too clear, so scientists rely on statistical analysis and probability theory to estimate the strength of an effect. Most statistical tests are not appropriate for elementary school, but you can introduce the idea of averages, and maybe talk about how small differences are not as good of evidence as large differences. Presentation of results as charts and graphs can certainly be started early and can help children think logically about the relationship between variables.



Process Skills
One of the most important and pervasive goals of schooling is to teach students to think. All school subjects should share in accomplishing this overall goal. Science contributes its unique skills, with its emphasis on hypothesizing, manipulating the physical world and reasoning from data.

Observing: using the senses to get information about an object or event.

Inferring: Interpreting or explaining one or more observations, often on the basis of prior experience or perceptions.

Measuring: using both standard and nonstandard measures or estimates to describe the dimensions of an object or event.

Communicating: using words or graphic symbols to describe an action, object or event.

Classifying: grouping or ordering objects or events into categories based on properties or criteria.

Predicting: Forecasting future events or conditions, based on patterns recognized in past observations.