The Scientific Method
Writing a Scientific Paper
Research Topic Ideas
The Scientific Method
The Scientific Method: An Overview
Although inquiry-based learning does not require that students actually carry out
experiments, allowing your students to design and carry out experiments is an excellent
way to begin to train them to think like scientists. Using the insects that you
are rearing in your classroom in simple experiments provides an ideal opportunity
to introduce students to scientific experimentation. If you have another preferred
format for teaching students to do experiments, it will be easy to incorporate the
insects into that format. If not, this section should help. For all experiments,
we suggest the following format.
We wrote about the scientific method in our
2005 newsletter. Read more about hypotheses and
is it that you would like to learn? Students will come up with excellent questions
as they observe their insects, and you may be able to design experiments to answer
their questions. If this happens in your classroom, take advantage of the opportunity
and follow your students’ lead. However, young students may not come up with questions
that are readily amenable to classroom experimentation. In this case, you may want
to suggest the questions provided in this section or those of your own choosing.
After students have had time to observe and become familiar with an organism, they
can move into the experimental process. This strategy helps students generate ideas
for experiments by guiding their thought process from the general (what do I want
to learn about?) to the specific (what variable will I change and how?). As students
do this, they will learn the concepts of independent
and dependent (or response)
variables. A characteristic whose value may change,
vary, or respond when manipulated experimentally is called a dependent variable.
Conversely, something that affects the characteristic of interest is called an independent
variable. The dependent variable is the one students will study. These concepts
will be used again as students' design experiments. If you would rather not use
these scientific terms with your students, you can decide to use other terms to
explain the same concepts.
The common definition of a hypothesis is a prediction
or an educated guess about what might happen in an experiment. However, a
more accurate definition is that a hypothesis simply describes one possible outcome
to an experiment, or one possible answer to your question. Hypotheses are statements
that can be tested through experimentation or observation; they can be disproved
or supported by evidence that you collect. Scientists regularly use
multiple hypotheses when conducting experiments. Using multiple hypotheses
teaches children that there are several possible outcomes to any experiment, and
these hypotheses will either be supported or not supported by the data. One hypothesis
that should always be considered is the null hypothesis.
This acknowledges that there might not be an effect of the independent variable
on the dependent variable.
Encouraging students to generate multiple hypotheses before they conduct an experiment
is not only a better representation of how scientists work, it is also a way to
help students avoid feeling like they were "wrong" if the experiment doesn’t turn
out the way they expected it to. In general, hypotheses clarify the question being
addressed in an experiment, help direct the design of the experiment, and help the
experimenters (students) maintain their objectivity.
After generating hypotheses, students are ready to design an experiment to test
their hypotheses. This is a time to pay close and careful attention to all the details—both
before starting the experiment and during the period of data collection. The steps
described below will help students think of as many details as possible when planning
their experiments. There are two important scientific concepts to discuss with your
students before they design their experiments: 1) replication or sample size and
2) constant conditions. In addition, students that are doing experimental studies
will need to consider a third concept: the control treatment.
Replication, or Sample Size
The best way to explain
replication is to use an example. Let’s say a scientist is interested
in how many children there are in different families. Her hypothesis is that families
who live in cities have fewer children than families who live on farms. If she only
counts the children in one city family and one farm family, random chance will affect
her results. She might happen to pick two very large families, two very small families,
a large farm family and small city family, or any other combination. If, instead,
she counts the children in 100 farm families and 100 city families, she will get
a much better picture of the average number of children in farm vs. city families.
Scientists have to use a large enough sample size to accurately test a hypothesis,
while taking into account things like cost, availability of experimental subjects,
and time. With younger students, however, the scientific "truth" may not be our
highest priority. We want students to understand the process and enjoy themselves.
If more replicates become tedious, then keep it simple. The concept of how to go
about answering a question may be more important than the veracity of the results
in some situations. You can always discuss how additional replicates could affect
your results and conclusions.
The second scientific concept to consider is the importance
of holding everything but the independent variable constant. For example,
if you want to study temperature effects on monarch growth, the larvae must be the
same age, kept in the same size and type of cage under the same light conditions,
and given the same type and amount of food and water. This is an essential part
of an experiment. Likewise, the scientist studying family size wouldn’t want to
study farm families where the parents were 40-50 years old, and city families where
the parents were 25-35 years old.
Again, we will use an example to illustrate this concept: Let’s say that you want
to test the hypothesis that loud music causes high mortality in earthworms. Your
students may suggest keeping a group of earthworms in a room with loud music going
constantly. However, if all the earthworms die, you won’t know if the music killed
them, or if there was something wrong with the earthworms in the first place. In
this experiment, you need a control. Place half of
the group in a “normal” environment (control group), and half in with the loud music
(experimental group). If more die in the room with the loud music, your hypothesis
is supported. Again, it is important to hold constant conditions between the two
groups. The conditions the earthworms are exposed to in the two groups should be
as similar as possible. Controls are rarely necessary in observational studies,
when you’re studying naturally-occurring variation. They are also not always needed
in experimental studies, but if you are subjecting living organisms to extreme conditions
of any kind, you should always include a control group that is not exposed to these
The fact that not all experiments need controls is illustrated by the following
example. If you want to find out if the earthworm density is the same under pine
trees as deciduous trees, you can simply measure the density in the two soils and
compare them. There is no need to have a separate control treatment in which earthworms
aren’t under either pine trees or deciduous trees. However, you do want to hold
other conditions constant. For example, you wouldn’t want to sample the earthworms
under the pine trees just after a rainy day unless you also sampled the deciduous
tree worms on the same day.
Describing and understanding the results of an experiment are critical aspects of
science. There are three parts to this lesson: making a data table, making graphs,
and analyzing data with simple statistical tests. You can choose any or all of these
parts, depending on your instructional goals. If you have access to computers, you
can use a spreadsheet program such as Microsoft Excel for all three parts. However,
students should also practice making tables and graphs "by hand."
Once students learn how to make organized data tables and graphs, they should use
this knowledge to present the results of their insect, plant or schoolyard studies.
They will need help as they do this, but going through this lesson will give them
After conducting an experiment and analyzing the results, students should come to
some conclusion as to what their results told them about the answer to their question.
Sometimes this conclusion will be quite easy to put into words: for example, Caterpillars
prefer milkweed to apple leaves. However, students may think that their results
may need to be qualified: for example, Caterpillars seem to prefer milkweed to
apple leaves, but all of the caterpillars we tested had only been fed milkweed leaves
before we used them in the experiment. This may have affected our results.
In any case, a conclusion should reflect what students have learned by doing the
The RERUN Method
The RERUN method is one method of writing conclusions.
RERUN is a short paragraph used to summarize the results from a scientific study.
The RERUN paragraph should be a minimum of five well-written, complete sentences.
RERUN is an acronym for five types of information that a conclusion should include:
- R = Recall: Describe what you did briefly.
- E = Explain: Explain the purpose of the study.
- R = Results: State the results, including which hypothesis was
supported by the study.
- U = Uncertainty: Describe uncertainties that exist, if any.
- N = New: Write two new things you learned.