Some members of the science education community have placed much faith in the investigative formula referred to as “the scientific method”—making observations, defining the problem, constructing hypotheses, experimenting, analyzing results, and drawing conclusions. Most scientists agree, however, this does not represent the way contemporary science works.

First, within the parameters of the scientific method, questions are often based on what is interesting or doable, but in real science, questions emerge from tentative models of how some part of the natural world works. A model represents the interrelationships between observable world features of a phenomenon (like a balloon expanding) and unobservable features (like the increasing number of air molecules colliding with the inside skin of the balloon). This kind of causal model, when used in classrooms, is typically an illustration, updated and changed as new investigations provide evidence supporting or rebutting relationships. But when a testable question is the only criteria for an investigation, then school science can become uninformed and “contentless.” Random solubility activities are one example. Data from these experiments are analyzed to determine only how outcomes are related to conditions (for example, small crystals of sugar dissolve faster in water than large sugar crystals), but the underlying “why” explanations (how molecular motion helps break the chemical bonds of sugar) are left unaddressed.

The second flaw relates to the first: Because the activity has no provisions for students to develop an initial model to inform their questions, no discussion can occur at the end of the inquiry about how new evidence fits or contradicts the model. Scientific argument does not just seek to demonstrate relationships between variables or differences between experimental groups, but also to use these findings to convince others that some processes—at the unobservable level—are at least partially responsible for the outcomes seen in data.

The third problem with the scientific method is it often promotes direct comparison between a control group and a manipulated experimental group as the only method of investigating the world. However, in science fields such as geology, field biology, molecular biology, natural history, and astronomy, controlled experiments are all but impossible; yet they all use systematic collection of data and coordination of evidence to propose explanations. These explanations are often in the form of models. Our collective reliance on oversimplified formulas for inquiry learning has given rise to some classroom practices that need to be reconsidered.

• Investigating arbitrary questions. Authentic science inquiry does not involve questions such as “Will my bean plants grow faster listening to rock music or classical music?” Such questions, although testable, have little to do with developing any coherent understanding of underlying causes. The questions are not grounded in any proposed model, and the results do not help us understand any natural processes.
• Investigations outside the bounds of the natural world. School science includes the broad domains of physics, biology, Earth and space sciences, and chemistry. It does not investigate questions of human behavior, such as “How many students prefer pizza vs. tacos for lunch?” or “Does extra sensory perception really exist?” While these can be motivational “hooks” for students, they are essentially “contentless” inquiries.
• Cookbook investigations. Some activities are so rigidly scripted that students do not have to employ any reasoning skills: All they have to do is follow instructions. Students can, in fact, earn passing grades in these activities without comprehending the meaning of the work. Such “confirmatory exercises” have a legitimate role when students have no previous inquiry experiences to draw upon, but a steady diet of these will soon cause students’ enthusiasm for science to wither away.
• Substituting isolated process skills for complete inquiries. The research on learning offers little evidence that process skills (observing, classifying, measuring, predicting, hypothesizing, inferring, and so on), learned in isolation from a real investigation, help students understand the purpose of these skills. Inquiry investigations should instead be treated as a coordinated set of activities and taught as a whole. Inquiry should be kept complex, but the teacher should scaffold students’ efforts as needed. The idea that any inquiry can be done in a day shortcuts students’ opportunities to reason about science—to discuss evidence, compare explanatory models, and identify other sources of information they need to be more confident about their explanations.

In my work with students, I’ve found it takes at least a full class period for them to respond to the question “What evidence do we now have that supports our explanatory model, and how strong is it?” Though students need support to have this conversation, they are the ones doing the intellectual work, and they get better at it as the year progresses.

While many science teachers do instill a sense of excitement and curiosity about the natural world in their students, my point is that, even for young learners, science should be about evidence, causal explanation, and the testing of models—however basic these models might be. Those interested in learning more about authentic forms of inquiry should read the chapter entitled “What Is Inquiry? A Framework for Thinking About Authentic Scientific Practice in the Classroom” in Science as Inquiry in the Secondary Setting from NSTA Press.

Mark Windschitl, a former secondary science teacher, is associate professor of science education at the University of Washington. He has done multiple studies on how early career science teachers develop and how inquiry is implemented in secondary classrooms.