The scientific method is a framework of techniques and questions that scientists use to investigate phenomena with the aim of making scientific discoveries simple and reproducible. It's been loosely observed by experimenters going as far back as the 4th century BC, but the first properly formalized scientific method was coined during the European Renaissance. Here individuals at the forefront of science like Francis Bacon, Galileo, and Isaac Newton started putting into routine practice the rules that we use to carry out experiments today.
Typically, the first step of the scientific method is to formulate a question, usually after observation of a phenomenon. For example, say you have been raising caterpillars and have noticed that some take longer than others to get to pupation. And you wonder, do the caterpillars develop at different rates depending on the temperature?
This is where the second part of the scientific method comes in, the hypothesis. A hypothesis is an uncertain explanation as to why we observe what we observe, and there are two main types. The first is the experimental or alternative hypothesis, and it implies that there will be a relationship between the variables being investigated, the temperature and caterpillar development, in this case. So, our experimental hypothesis could be that the caterpillars will take longer to go from egg to pupation if they're raised at colder temperatures. Crucially, a good hypothesis will be testable. For our caterpillars, we can change the temperature, and record the time it takes for them to go from egg to pupa, and falsifiable. So, if it takes around the same time for the caterpillars to develop no matter what the temperature, then we can accept that the hypothesis was likely false. The second type of hypothesis is the null hypothesis. This typically speculates that there won't be any observed significant change or difference during the experiment. In our caterpillar example, we would state that the caterpillars will develop at the same rate in each temperature condition.
Once we have our hypotheses, the third step of the scientific method covers experimentation and data collection. In a typical experiment, there will be two types of variables. The independent variable is something directly manipulated by the experimenter. So, with our caterpillars, we are altering the independent variable when we change the temperature. The dependent variable, also known as the response variable, should be affected by the state of the independent variable. So, when we expose our caterpillars to different temperatures, then the response, the dependent variable, is the rate at which they develop.
There are also two main types of data that could be collected to support or falsify the hypotheses. The first is qualitative data, which typically refers to descriptive observations made with the senses, seeing, touching, hearing, smelling, or even tasting. In our experiment, we might record that the caterpillars seem to move around and eat a lot in the normal temperature condition, compared to the cooler one. In contrast to qualitative data, quantitative data can be measured and written down as numbers. So, when we count the number of hours it takes the caterpillar from hatching to finally pupating, this gives us a definite figure. Where possible, it's almost important to have a control condition in any experiment where we manipulate the independent variables. In our caterpillar experiment, we can grow the caterpillars at a set standard room temperature of 21 degrees as a control, because this demonstrates what happens when the caterpillars develop under normal conditions in comparison to experimental settings.
In observational experiments, a control may not be needed or even possible. For example, imagine our caterpillars are now grown up butterflies, feeding on nectar in a flower garden. In our experimental hypothesis, we suggest that they prefer to feed from the big pink flowers, while our null hypothesis suggests they have no preference and will visit the flowers at random. In this case, simply observing and recording the number of times the butterflies visit each flower type will provide enough data to confirm or reject our hypotheses without needing manipulation of any variables or the need for a control.
Once the data have been collected, the next step is to figure out what it all means. Scientists will compare the predictions of their two hypotheses to figure out if they can reject the null hypothesis. This can be done by comparing the values of the dependent variable in the control versus the experimental conditions. If they are not equal, the null hypothesis can be rejected. If the data collected supports a hypothesis, like the caterpillars did take significantly more hours to go from egg to pupa when kept at the cooler climate, then this gives the experimental hypothesis more credibility, but critically it does not indicate that the hypothesis is definitely true, because future experiments may reveal new information.
The final part of the scientific method is where we draw conclusions, and discuss what our findings might mean. Here, scientists might refer to other experiments or other literature to put their findings into context, and come up with explanations of why the results showed what they did. For example, the conclusion could be that the caterpillars like to grow at temperatures closest to their natural habitat. This may, in turn, spiral new questions, like do other species pupate at different rates at different temperatures, too? This may inspire new experiments, which we can test using, you guessed it, the scientific method.
The scientific method is used to solve problems and explain phenomena. The development of the scientific method coincided with changes in philosophy underpinning scientific discovery, radically transforming the views of society about nature. During the European Renaissance, individuals such as Francis Bacon, Galileo, and Isaac Newton formalized the concept of the scientific method and put it into practice. Although the scientific method has been revised since its early conceptions, much of the framework and philosophy remains in practice today.
Prior to investigation, a scientist must define the question to be addressed. This crucial first step in the scientific process involves observing some natural phenomena of interest. This observation should then lead to a number of questions about the phenomena. This stage frequently requires background research necessary to understand the subject matter and past work on similar ideas. Reviewing and evaluating previous research allows scientists to refine their questions to more accurately address gaps in scientific knowledge. Defining a research question and understanding relevant prior research will influence how the scientific method is applied, making it an important first step in the research process.
An everyday example: You are trying to get to school or work and your car won’t start. The thought process that most people go through in that situation clearly mirrors the official scientific method (after you are finished getting upset). First, you make an observation: my car won’t start! The question that follows: why isn’t it working?
The next step is making a hypothesis, based on prior knowledge. A hypothesis is an “uncertain explanation” or an unproven conjecture that seeks to explain some phenomenon based on knowledge obtained while executing subsequent experiments or observations. Generally, scientists develop multiple hypotheses to address their questions and test them systematically.
All hypotheses must meet certain criteria for the scientific process to work. First, a hypothesis must be testable and falsifiable. This aspect of the hypothesis is critical and of much greater importance than the hypothesis being correct. A testable hypothesis is one that generates testable predictions, addressed through observations or experiments. A falsifiable hypothesis is one that, through observation of conflicting outcomes, can be proven wrong. This allows investigators to gain more confidence over time, not by accumulating evidence showing that a hypothesis is correct, but rather by showing that situations that could establish its falsity do not occur.
Hypotheses come in two forms: null hypotheses and alternative hypotheses. The null hypothesis is tested against the alternative hypothesis and reflects that there will be no observed change in the experiment. The alternative hypothesis is generally the one described in the previous two paragraphs, also referred to as the experimental hypothesis. The alternative hypothesis is the predicted outcome of the experiment. If the null hypothesis is rejected, then this builds evidence for the alternative hypothesis.
An everyday example: Maybe it is freezing outside and therefore it is fairly likely that your car battery is dead. Maybe you know you were low on gas the night before and therefore it is likely that the tank is empty.
Either way, the next step is to make more observations or to conduct experiments leading to conclusions. Following the formulation of hypotheses, scientists plan and conduct experiments to test their hypotheses. These experiments provide data that will either support or falsify the hypothesis. Data can be collected from quantitative or qualitative observations. Qualitative information refers to observations that can be made simply using one's senses, be that through sight, sound, taste, smell, or touch. In contrast, quantitative observations are ones in which precise measurements of some type are used to investigate one's hypothesis.
An experiment is a procedure designed to determine whether observations of the real world agree with or refute the derived predictions in the hypothesis. If evidence from an experiment supports a hypothesis, that gives the hypothesis more credibility. This does not indicate that the hypothesis is true, as future experiments may reveal new information about the original hypothesis. Experimental design is another critical step in the scientific method and can have a great effect on the results and conclusions one draws from an experiment. Careful thought and time should be devoted to experimental design and minimizing possible errors. The experiment should be designed so that every variable or factor that could influence the outcome of the experiment be under control of the researcher. Two types of variables are used to describe the conditions in an experiment: the independent and the dependent, or response, variable. The independent variable is directly manipulated or controlled by the scientist and is generally what one predicts will affect the dependent variable. The dependent, or response, variable thus depends on the value of the independent variable. Experiments are generally designed so that one specific factor is manipulated in the experiment in order to illuminate cause and effect relationships.
An everyday example: Does the car still have all of its parts? Is this the right key? What does the gas gauge say? Does a jump start help?
Another important aspect in experimental design is the role of the control treatment, which represents a non-manipulated treatment condition. The control treatment is kept in the same conditions as the experimental treatment, but the experimental manipulation is not applied to the control. For example, if a researcher were testing the effects of soil salinity on plant growth, the soil in the control treatment would have no added salt. The control provides a baseline of “normal” conditions with which to compare the experimental treatments.
Experimental design should also incorporate replicates of each treatment. Repeatability of experimental results is an important part of the scientific method that ensures the validity and accuracy of data. It is quite difficult to control all aspects of an experiment so there is inherent variation in results that cannot be controlled for even under the most carefully designed and controlled experiments. Having replicates enables an investigator to estimate this inherent variation in results. Precise recording and measurement of data is also of great importance for ensuring the accuracy of results and the conclusions one draws from the results.
The next step in the scientific method involves determining what the results from the experiment mean. Scientists compare the predictions of their null hypothesis to that of their alternative hypothesis to determine if they are able to reject the null hypothesis. Rejecting the null hypothesis means that there is a significant probability that values of the dependent variable in the control versus experimental treatments are not equal to each other. If significant differences exist, then one can reject the null hypothesis and accept the alternative hypothesis. Conversely, the investigator may fail to reject the null hypothesis, meaning the treatment has no effect on the results. Before scientists can make any claims about their null hypothesis from their experimental data or observations, statistical tests are required to ensure the validity of the data and further interpretation of the data. Statistical tests allow researchers to determine if there are genuine differences between the control and experimental treatments. From there, they can create figures and tables to illustrate their findings.
The last portion of the scientific method involves providing explanations of the results and the conclusions that can be logically drawn from the results. Generally, this step of the scientific process also requires one to revisit scientific literature and compare their results with other experiments or observations on related topics. This allows researchers to put their experiment in a more general context and elaborate on the significance of particular results. Additionally, it allows them to explain how their work fits into a larger context in their discipline.
The scientific process does not stop here! The scientific process works through time as knowledge on topics in science accumulate and drive our understanding of particular mechanisms or processes explaining natural phenomena. If we fail to reject our null hypothesis, then it becomes necessary to revisit the initial stages of the scientific method and try to reformulate our questions and understand why an anticipated result was not attained.
The only difference between the use of this method in every-day life and in the laboratory is that scientists carefully document their work, from observation to hypothesis to experiment, and finally conclusions and peer review. In addition, unlike problem solving outside the lab, the scientific method in the lab includes controlled conditions and variables.
Let’s investigate the scientific method using an example from the lab. It is known that plant growth is affected by microbes, such as bacteria and fungi, living in their soil. It is possible to figure out what microbes have which effects by potting plants in completely sterile soil, then adding in microbes one at a time, or in different combinations and measuring the growth of the plant. Now let’s fit this into the terms used to describe the scientific method:
Observation and Question: There are microbes in the soil…do these affect plant growth?
HYPOTHESES:
Experimental: One particular microbe of interest will cause the plants to grow more slowly.
Null: The presence or absence of microbes will have no effect on plant growth
Experiment: set up groups of plants in 1) sterile soil, 2) soil with the microbe added in, and 3) natural soil. Measure the growth of the plants over time, using a ruler.
Conclusion: if the plants in group 2 grow more slowly than the other two, the hypothesis is supported. This needs to be backed up with statistical analysis from many plants to be considered significant. An experiment like this is not legitimate with just one plant per group.
Group 1 is a control which shows the plants can grow in the sterile soil. Group 3 is a control that shows the plants can grow under normal conditions. Group 2 is the experimental group. It would be possible to add different amounts of the microbe, or different microbes, to introduce more variables. The main point is that the researcher has something to which to compare the experimental group- the control group. If the experiment included only group 2 and the researcher determined that the plants “looked sick,” that would be a matter of opinion. The only way to make that observation scientific is to have healthy plants to measure. The type or amount of microbe used is the independent variable, because the researcher has control over it. The size of the plant at the end of the experiment is the dependent or response variable because it is the result.
Ultimately, work like this is published in scientific journals so that other researchers can read about the methods used and conclusions drawn. Publications like this are subject to peer-review, which means that an article won’t be published in a journal until other researchers have checked it out and agree it is well-done. As a community of scientists, general concepts are developed based on observed patterns in the experiments that individual scientists conduct. This results in the development of a scientific theory. This term means that there is a consensus among researchers that a particular concept or process exists. It is important to note that the word theory does not mean the same thing as hypothesis. Once scientists label a concept with this term, it is considered to be true, considering all currently available data. Of course, if a large body of experimentation demonstrates information to the contrary, theories can be modified.
