In Studying Nature, Scientists Make Observations and Form/Test Hypothesis

Misconception of a Controlled Experiment

A common misconception is that the term controlled experiment means that scientists control all features of the experimental environment. But that's impossible in field research and can be very difficult even in highly regulated laboratory experiments. Researchers usually "control" unwanted variables not by eliminating them through environmental regulation, but by canceling out their effects by using control groups.

Coat Coloration: Results

For each environment, the researchers then calculated the percentage of predation events that targeted camouflaged models. The results were clear-cut: Camouflaged models showed much lower predation rates than those lacking camouflage in both the beach habitat (where light mice were less vulnerable) and the inland habitat (where dark mice were less vulnerable). The data thus fit the key prediction of the camouflage hypothesis.

Example of Recording Data

For example, Jane Goodall spent decades recording her observations of chimpanzees behavior during field research in a Tanzanian jungle. In her studies, Goodall also enriched the field of animal behavior with volumes of quantitative data, such as the frequency and duration of specific behaviors for different members of a group of chimpanzees in a variety of situations.

Controlled Experiment

In carrying out an experiment, a researcher often manipulates one factor in a system and observes the effects of this change. The mouse camouflage experiment is an example of a controlled experiment, one that is designed to compare an experimental group (the non-camouflaged mice models, in this case) with a control group (the camouflaged models).

Observations May be Too Puzzling

In other cases, observations remain too puzzling to prompt well-defined questions until further study provides a new context in which one to view those observations. For example, scientists could not unravel the details of how genes encode proteins until after the discovery of the structure of DNA (an event that took place in 1953).

Bounds of Science

Not all hypotheses meet the criteria for science: You wouldn't be able to test the hypothesis that invisible ghosts are fooling with your desk lamp! Because science only deals with natural, testable explanations for natural phenomena, it can neither support nor contradict the invisible ghost hypothesis, nor whether spirits or elves cause storms, rainbows, or illnesses. Such supernatural explanations are simply outside the bounds of science, as are religious matters, which are issues of personal faith. Science and religion are not mutually exclusive or contradictory; they are simply concerned with different issues.

Case Study: Coat Coloration in Mouse Populations

Our case study begins with a set of observations and inductive generalizations. Color patterns of animals may vary widely in nature, sometimes even among members of the same species. What accounts for such variation? The two mice are members of the same species but have different color patterns and reside in different environments.

Biologists Rely on Published Scientific Literature

Our innate curiosity often stimulates us to pose questions about the natural basis for the phenomena we observe in the world. For example, what causes the roots of a plant seedling to grow downward? In fine-tuning their questions, biologists rely heavily on the scientific literature, the published contributions of fellow scientists.

Experiment

The hypothesis is usually a rational accounting for a set of observations, based on the available data and guided by inductive reasoning. A scientific hypothesis must lead to predictions that can be tested by making additional observations or by performing experiments. An experiment is a scientific test, carried out under controlled conditions.

Coat Coloration: Control and Experimental Groups

The researchers built hundreds of models of mice and spray-painted them to resemble either beach or inland mice, so that the models differed only in their color patterns. The researchers placed equal numbers of these model mice randomly in both habitats and left them overnight. The mouse models resembling the native mice in the habitat were the control group (for instance, light-colored mouse models in the beach habitat), while the mouse models with the non-native coloration were the experimental group (for example, darker models in the beach habitat).

Theory Has Different Meaning in Science

"It's just a theory!" Our everyday use of the term theory often implies an untested speculation. But the term theory has a different meaning in science. What is a scientific theory, and how is it different from a hypothesis or from mere speculation?

Deductive Reasoning

A type of logic called deduction is also built into the use of hypotheses in science. While induction entails reasoning from a set of specific observations to reach a general conclusion, deductive reasoning involves logic that flows in the opposite direction, from the general to the specific.

Hypothesis

After carrying out preliminary observations and collecting and analyzing data, scientists begin to form tentative answers to their original questions and to test their hypothetical explanations--that is, their hypotheses. In science, a hypothesis is an explanation, based on observations and assumptions, that leads to a testable prediction. Said another way, a hypothesis is an explanation on trial.

Frequent Hypothesis Testing Leads to a Scientific Consensus

Although a hypothesis can never be proved beyond the shadow of doubt, testing it in various ways can significantly increase our confidence in its validity. Often, rounds of hypothesis formulation and testing lead to a scientific consensus--the shared conclusion of many scientists that a particular hypothesis explains the known data well and stands up to experimental testing.

Theories are Supported by a Greater Body of Evidence

And third, compared to any one hypothesis, a theory is generally supported by a much greater body of evidence. The theory of natural selection has been supported by a vast quantity of evidence, with more being found every day, and has not been contradicted by any scientific data. Those theories that become widely adopted in science (such as the theory of natural selection and the theory of gravity) explain a great diversity of observations and are supported by a vast accumulation of evidence.

Elements of Inquiry

As in all quests, science includes elements of challenge, adventure, and luck, along with careful planning, reasoning, creativity, patience, and the persistence to overcome setbacks. Such diverse elements of inquiry make science far less structured than most people realize. That said, it is possible to highlight certain characteristics that help to distinguish science from other ways of describing and explaining nature.

Inquiry

At the heart of science is inquiry, a search for information and explanations of natural phenomena. There is no formula for successful scientific inquiry, no single scientific method that researchers must rigidly follow.

Variables

Both the factor that is manipulated and the factor that is subsequently measured are types of experimental variables--a feature or quantity that varies in an experiment.

Scientists Build on Existing Knowledge with Publications

By reading about and understanding past studies, scientists can build on the foundation of existing knowledge, focusing their investigations on observations that are original and on hypotheses that are consistent with previous findings. Identifying publications relevant to a new line of research is now easier than at any point in the past, thanks to indexed and searchable electronic databases.

Inductive Reasoning

Collecting and analyzing observations can lead to important conclusions based on a type of logic called inductive reasoning. Through induction, we derive generalizations from a large number of specific observations. "The sun always rises in the east" is an example. And so is "All organisms are made of cells." Careful observations and data analyses, along with generalizations reached by induction, are fundamental to our understanding of nature.

Theories are Much Broader than Hypotheses

First, a scientific theory is much broader in scope than a hypothesis: "Coat coloration well-matched to their habitat is an adaptation that protects mice from predators." But this is a theory: "Evolutionary adaptations arise by natural selection." This theory proposes that natural selection is the evolutionary mechanism that accounts for an enormous variety of adaptations, of which coat color in mice is but one example.

Example of Deductive Testing

From general premises, we extrapolate to the specific results we should expect if the premises are true. In the scientific process, deductions usually take the form of predictions of results that will be found if a particular hypothesis (premise) is correct. We then test the hypothesis by carrying out experiments or observations to see whether or not the results are as predicted. This deductive testing takes the form of "If...then" logic. In the case of the desk lamp example: If the burnt-out bulb hypothesis is correct, then, the lamp should work if you replace the bulb with a new one.

Independent and Dependent Variables

In our example, the color of the mouse model was the independent variable--the factor being manipulated by the researchers. The dependent variable is the factor being measured that is predicted to be affected by the independent variable; in this case, the researchers measured the predation rate in response to variation in color of the mouse model. Ideally, the experiment and control groups differ in only one independent variable--in the mouse experiment, color.

Scientists will often Modify and Reject Theories

In spite of the body of evidence supporting a widely accepted theory, scientists will sometimes modify or even reject theories when new research produces results that don't fit. For example, biologists once lumped bacteria and archaea together as a kingdom of prokaryotes. When new methods for comparing cells and molecules could be used to test such relationships, the evidence led scientists to reject the theory that bacteria and archaea are members of the same kingdom. If there is "truth" in science, it is at best conditional, based on the weight of available evidence.

Observations Reveal Valuable Information

In the course of their work, biologists make careful observations. In gathering information, they often use tools such as microscopes, precision thermometers, or high-speed cameras that extend their senses or facilitate careful measurement. Observations can reveal valuable information about the natural world. For example, a series of detailed observations have shaped our understanding of cell structure, and another set of observations is currently expanding our databases of genome sequences from diverse species and databases of genes whose expression is altered in various diseases.

Coat Coloration Hypothesis

It was logical therefore, for Francis Bertody Sumner, a naturalist studying populations of these mice in the 1920s, to form the hypothesis that their coloration patterns had evolved as adaptations that camouflage the mice in their native environments, protecting them from predation. As obvious as the camouflage hypothesis may seem, it still required testing. In 2010, biologist Hopi Hoekstra of Harvard University and a group of her students headed to Florida to test the prediction that mice with coloration that did not match their habitat would be preyed on more heavily than the native, well-matched mice.

Quantitative Data

Quantitative data are generally expressed as numerical measurements and often organized into tables and graphs. Scientists analyze their data using a type of mathematics called statistics to test whether their results are significant or merely due to random fluctuations.

Qualitative Data

Recorded observations are called data. Put another way, data are items of information on which scientific inquiry is based. The term data implies numbers to many people. But some data are qualitative, often in the form of recorded descriptions rather than numerical measurements.

Science

Science is a way of knowing--an approach to understanding the natural world. It developed out of our curiosity about ourselves, other life-forms, our planet, and the universe. The word science is derived from the latin verb meaning "to know." Striving to understand seems to be one of our most basic urges.

A Hypothesis Must be Testable

Scientific inquiry is a powerful way to learn about nature, but there are limitations to the kinds of questions it can answer. A scientific hypothesis must be testable; there must be some observation or experiment that could reveal if such an idea is likely to be true or false. The hypothesis that a burnt-out bulb is the sole reason the lamp doesn't work would not be supported if replacing the bulb with a new one didn't fix the lamp.

Repetition is Key

Scientists use a process of inquiry that includes making observations, forming logical, testable explanations (hypothesis), and testing them. The process is necessarily repetitive: In testing a hypothesis, more observations may inspire revision of the original hypothesis or formation of a new one, thus leading to further testing. In this way, scientists circle closer and closer to their best estimation of the laws governing nature.

Theories are General Enough to Create New Hypotheses

Second, a theory is general enough to spin off many new, testable hypotheses. For example, the theory of natural selection motivated two researchers at Princeton University, Peter and Rosemary Grant, to test the specific hypothesis that the beaks of Galapagos finches evolve in response to changes in the types of available food.

Hypotheses Can Never Be Proven

Second, we can never prove that a hypothesis is true. The burnt-out bulb hypothesis is the most likely explanation, but testing supports that hypothesis not by proving that it is correct, but rather by failing to prove it incorrect. For example, even if replacing the bulb fixed the desk lamp, it might have been because there was a temporary power outage that just happened to end while the bulb was being changed.

Coat Coloration Matches Mouse Environment

The beach mouse lives along the Florida seashore, a habitat of brilliant white sand dunes with sparse clumps of beach grass. The inland mouse lives on darker, more fertile soil farther inland. The natural predators of these mice, including hawks, owls, foxes, and coyotes, are all visual hunters (they use their eyes to look for prey).

Scientific Inquires are Rarely Rigid

The desk lamp example traces an idealized process of inquiry sometimes called the scientific method. However, very few scientific inquires adhere rigidly to the sequence of steps that are typically used to describe this approach. For example, a scientist may start to design an experiment, but then backtrack after realizing that more preliminary observations are necessary.

Coat Coloration: Signs of Predation Events

The following morning, the team counted and recorded signs of predation events, which ranged from bites and gouge marks on some models to the outright disappearance of others. Judging by the shape of the predators' bites and tracks surrounding the experimental sites, the predators appeared to be split fairly evenly between mammals (such as foxes and coyotes) and birds (such as owls, herons, and hawks).

Simple Example of Scientific Process

We all make observations and develop questions and hypotheses in solving everyday problems. Let's say, for example, that your desk lamp is plugged in and turned on but the bulb isn't lit. That's an observation. The question is obvious: Why doesn't the lamp work? Two reasonable hypotheses based on your experience are that 1. the bulb is not screwed in properly or 2. the bulb is burnt out. Each of these alternative hypotheses leads to predictions you can test with experiments. For example, the improperly screwed-in bulb hypothesis predicts that carefully re-installing the bulb will fix the problem

Additional Hypotheses are Always Possible

We can use the desk lamp example to illustrate how other key points about the use of hypotheses in science. First, one can always devise additional hypotheses to explain a set of observations. For instance, another hypothesis to explain our nonworking desk lamp is that the electrical socket is broken. Although you could design an experiment to test this hypothesis, you can never test all possible hypotheses.

Importance of a Control Group

Without the control group, the researchers would not have been able to rule out other factors as causes of the more frequent attacks on the non-camouflaged mice--such as different numbers of predators or different temperatures in the different test areas. The clever experimental design left coloration as the only factor that could account for the low predation rate on models camouflaged with respect to the surrounding environment.