Scientists often design and conduct investigations to develop and test theories that answer questions about how the world works. Students need opportunities to engage in empirical investigations, to test explanatory theories of the world and gather data to support or refute their predictions. Engaging in empirical inves tigations is a multifaceted activity that includes a set of procedural and conceptual activities, such as asking questions; generating hypotheses; designing experiments; collecting and recording data; analyzing, representing, and interpreting results; coordinating theory and evidence; and formulating and revising theories or models (Duschl et al., 2007). Below, we elaborate on a few of these subpractices included in the practice of engaging in empirical investigations:

  • Asking questions is an important component of science literacy, since an essential aspect of science is to develop explanations or models that address questions about phenomena in the natural world. We emphasize that the questions scientists attempt to account for occur in the natural world, not the human-made world (the latter is the domain of engineering or a technological field such as computer science). The type of question that a student can ask can indicate the level of sophistication in that student’s thinking (Yarden, Brill, & Falk, 2001). Cuccio-Schirripa and Steiner (2000) suggested, ‘‘Questioning is one of the … processing skills which is structurally embedded in … critical thinking, creative thinking, and problem solving’’ (p. 210). Thus, questioning may contribute in important ways to the development of a scientific theory, since it is one potential bridge between theoretical ideas and the empirical investigation that includes data collection, interpretation, and evaluation of evidence needed to test or revise ideas (a key feature of critical thinking and problem solving). Accordingly, it is important to help students ask well-formulated questions that can be answered through empirical investigation.

  • Generating a hypothesis involves proposing a rational and tentative explanation of an observed phenomenon that has not yet been proven. A hypothesis is often produced to identify the relationship between a dependent variable and an independent variable. Many hypotheses are stated in the form of “If a particular independent variable is changed there is also a change in a certain dependent variable.” Testing the hypothesis developed is the goal of the designed experiment. Generating hypotheses helps students plan their data collection by considering variables and expected observations involved in their inquiry.

  • Designing experiments involves identifying causal variables and generating interpretable observations that will serve as evidence – evidence that can be related to a hypothesis about the questions proposed by students . One important aspect of design is to isolate variables so as to rule out competing hypotheses. A well-designed investigation often includes controlled variables that allow valid inferences and narrows the number of possible experiments to consider (Klahr, 2000). Another important aspect of designing investigations is selecting appropriate measures or tools to examine the phenomena of interest (e.g., determining whether the total mass or the density of a material might be a better measure of a property to identify a material’s identity, or determining whether height or mass is a better measure of growth). Students may also need to understand certain aspects of measurement error and the need for accurate and precise measurements that might require appropriate or specific tools (e.g., a balance is a more accurate measure of the comparative masses of two different objects than relative heft in a person’s hand). There might also be a need to appreciate sample size (i.e., whether there are enough observations to make a generalizable inference) and the role of randomness in selecting a sample (i.e., the potential for bias because the sample is unlikely to be a true representation of the population).

  • Collecting and recording data is also an important component of engaging in empirical investigation, since access to interpretable, useful, valid, and reliable data is essential to making effective evidence-based explanations or arguments. A datum is an observation or measurement of a natural system or of a designed and constructed experimental situation recorded for subsequent analysis. Research has shown that students are unlikely to check whether a current hypothesis is inconsistent with experimental results (Dunbar & Klahr, 1989). The access to collected data may help students realize inconsistencies between what is predicted and what is observed. Students must identify the best ways to record data accurately and reliably.

  • Analyzing and interpreting data often follows after data collection. This is when students make sense of data and connect the data as evidence to support the hypothesis. This process is often related to students’ skills in using mathematical and computational tools. The National Research Council (NRC) framework (NRC, 2012) states that mathematical tools (e.g., formulas, mathematical rules underlying a simulation’s programming) enable students to express ideas in a precise form, and computational tools (e.g., simulations, animations) enable students to visually represent data and scientific phenomena. Both types of tools allow students to explore and identify patterns in data and their observations. It is also recognized that students’ abilities to engage in mathematical thinking are related to their ability to identify patterns in data and represent data in appropriate ways. In the domain of science, mathematical thinking skill is also related to the ability to effectively design or select a measure for a given variable, particularly in simulations. For example, students need to apply their mathematical thinking skills to consider what values should be set for the variables of the amount of water and the amount of salt in a simulation that helps understand the concept of density.

Computational tools, such as computer simulations, provide cost-effective ways of doing experiments. Students can use simulations to observe scientific phenomena that cannot be observed easily in real time (e.g., volcano eruption, protein synthesis, and spread of disease). The enabled visualization of scientific phenomena and interactive opportunities provided by computer simulation has been associated with gains in conceptual understanding and enhances epistemological understanding about the nature of science as exploring and discovering knowledge through multiple trials of experimentation and evidence-based reasoning (Wilensky & Reisman, 2006; Zacharia, 2007. From the aspect of assessment, computer simulations also allow testing of both conceptual understanding and the ability to engage in the central practices of science that are not well tested in traditional assessments (Quellmalz & Pellegrino, 2009).

One major challenge with the practice of engaging in empirical investigation is that often there is little classroom instruction focused on designing and carrying out a high quality investigation and analyzing and interpreting data collected from the investigation. Therefore it is necessary to scaffold students as they attempt to complete assessment tasks that target their ability to engage in empirical investigations. Another challenge is that it takes time to engage students in a full cycle of empirical investigation, a particular problem for short assessments. One possible approach is to target pieces of the practice of engaging in empirical investigation such as asking questions, designing an investigation, or analyzing data in one task. Alternatively, we may also use surrogate methods to measure individual practices (e.g., using multiple choice items asking students to choose from a list of potential procedures). In addition, we might design extended tasks that will be completed across multiple class periods so that students have time to engage in the various aspects of empirical investigation. Furthermore, we may ask students to conduct an experiment in a simulation-based setting where they can take less time than in a traditional real-time lab.

Back to Science Practices