printed from

Experiences with Chemistry ConcepTests

Arthur B. Ellis, Ann Cappellari, Julie K. Lorenz, and Dale E. Moore
University of Wisconsin-Madison

George C. Lisensky
Beloit College

Introduction. ConcepTests are a method of in-class instruction developed by Eric Mazur for introductory physics courses. Mazur has published information on the method and recently established a World Wide Web site of physics questions (1). The intent of this article is to give chemistry teachers some ideas for implementing the method based on its use in classes at the University of Wisconsin-Madison, expanding on a brief description that has been published (2). Principal virtues of ConcepTests are their ease of use and versatility, and a number of examples are included herein.

The basic idea behind ConcepTests is peer instruction: Students who have just grasped a concept may be able to explain it to a classmate more effectively than the instructor for at least three possible reasons: 1. The insight is fresher. If the instructor learned many of the fundamental chemistry concepts some time ago, it is easy to forget whatever obstacles there were to mastering the ideas initially; in cases for which the concept was obvious, it can be difficult to have an appreciation for why it may not be obvious to someone else. 2. Classmates speak a more common jargon that facilitates communication. Years of sanding explanations can leave a teacher disconnected with regard to how to cast an appropriate explanation for his or her students. 3. The teacher is the authority figure who hands out the grades. This represents a barrier for some students in terms of establishing a suitable level of comfort at which they could learn directly from the instructor.

Besides engaging students in helping to communicate the course concepts, the method gives the teacher feedback in real time on the appropriateness of the pace of the course. As will be described below, it is easy to tell when most of the class has or has not mastered a concept, and lecture explanations and speed can be adjusted on-line to reach the majority of the class.

From a teacher's perspective, ConcepTests are a wonderful pedagogical tool because they can be used with virtually any kind of content and require very little time investment and little or no resources: A teacher simply identifies a number of key ideas that form the basis for the lecture and "packages" as many as desired as ConcepTests. As Mazur has shown, it is possible to collect a considerable amount of enlightening statistical information from student responses, particularly with modern touchpad technology, but a show of hands is sufficient to obtain good qualitative insights into the level of student understanding.

From a student's perspective, ConcepTests provide a clear indication of the key concepts that a teacher values and some immediate feedback as to whether he or she understands an idea or is going to need to invest additional effort in mastering it. Students have expressed appreciation for the fact that the time provided in lecture to conduct the ConcepTest provides an opportunity to digest information and to look at ideas from several points of view.

To establish a culture of cooperation and make the use of ConcepTests a natural part of the course, it has proved advantageous to grade the course on an absolute scale rather than on a curve. That is, students are told at the outset that a certain number of total points in the class guarantees a particular letter grade (2, 3). Students are willing to help one another, since they recognize it will not jeopardize their grade.

Implementation. There can be particular concern with the initial use of ConcepTests if it is the first time "the class is turned over to the class." ConcepTests discussed here have been used in large-lecture introductory general chemistry classes of nonmajors (although a few students subsequently became majors) of typically 200 to 350 students, and in recitation sections of about two dozen. Introducing ConcepTests on the first day of class works well, although there is no reason why its introduction could not occur at any time during the course. A typical introduction has been to pick an engaging topic, pose a question, and provide two to four choices to stimulate discussion. Of course, it is important to ensure beforehand that students are seated near enough to one another to be able to converse.

As an example, a topic covered in the first lecture of general chemistry concerns our ability to image atoms using a scanning tunneling microscope (STM). A simple demonstration piques curiosity: two wires are brought together and when they are close enough to touch, an electrical circuit is completed and a bulb is lighted (4). This raises the question of how close the wires need to be for electrons to flow as electricity. After explaining that a wire tip can be prepared that terminates in a single atom and that it can be scanned over the surface of the other wire in atomic-scale increments, the picture shown below is sketched on the blackboard (or an overhead transparency can be prepared). The class is told that they will be asked in a moment for a show of hands as a response to possible answers to the following question: "Which curve describes how the electrical current will vary as the tip-to-surface distance changes, A or B?" After giving the class a few moments for reflection, the show of hands is requested: "How many of you think curve A is the appropriate relationship? How many of you think it is curve B?" Usually some hands are observed for each answer. There are also students who don't commit to either answer. At this point the fun begins: "I'd like you to turn to your neighbor, introduce yourself, and then convince him or her that your answer is correct." There may be a moment of stunned silence, as though the class is thinking, "You mean, we can talk in class?" Then, typically, loud discussion ensues.

The intensity level of class discussion that follows often determines how much time to allot for it. After a suitable period of time has passed, typically marked by a lull in the discussion level, the instructor interrupts and asks for another vote. If by show of hands most students have converged on the correct answer, the instructor can briefly affirm why it is correct and move on. If the class has converged on the wrong answer or not many hands are raised, this is a signal that the class is not following, and the teacher has some choices. One choice is to provide an additional clue, if the question lends itself to that, and repeat the process, with or without a discussion period; or, the instructor can try to explain why another answer is more appropriate. The value of the ConcepTest, as noted above, is that the pace of the course is adjusted on-line as class mastery is assessed in real time.

Some variations on the delivery of the ConcepTest are noteworthy. In some instances, a large majority of the class immediately chooses the correct answer (this is often the case with simple counting-type questions, such as whether Na+ has 10, 11, or 12 electrons), in which case the discussion period can be skipped and a quick explanation provided as to why 10 is the correct answer. With more challenging questions like whether the absorption of light by a solution is linear or logarithmic (this is coupled to a demonstration; see sample questions below), there is an opportunity to walk around the lecture room and listen to some of the discussion, which can provide the instructor with some insight into possible misconceptions. Some concepts lend themselves to a series of related choices and the initial show of hands can be eliminated: For example, after reviewing a rule-of-thumb for predicting when to expect extended structures and when to expect discrete molecules based on chemical formula (5), students are asked to indicate the appropriate structure for each participant in the reaction, after giving them a little time to discuss this with their neighbor. The instructor then requests shows of hands, asking in turn whether Na, Cl2, and NaCl is a discrete molecule or extended solid.

An example that has been used in the first lecture for the second semester course of our two-semester introductory sequence, by way of review, is to ask students to indicate whether a relatively simple equation is balanced ("How many think it is? How many think it isn't?")

How many ConcepTests are desirable? During a 50-minute lecture period, typically three to six have been used, depending upon how many are tied to demonstrations (see examples below). Some colleagues use a single ConcepTest. The class quickly becomes accustomed to this approach. It has been found to boost attendance considerably in the large lecture presentation, for which attendance has not been mandatory.

Instructor Attitude. When giving a ConcepTest, it is helpful to be prepared for anything in the way of class response. Hearing students animatedly discussing chemical concepts in the lecture hall can be an exhilarating experience, and it is gratifying when the class converges on the correct answer. Some occasional acknowledgment of their success may be well received by the class. On the other hand, convergence on a wrong answer or obvious confusion is best served by a patient explanation or an attempt to use an alternative approach to communicate the concept.

An experience worth relating occurred with a demonstration involving a concentration gradient, which was used as a vehicle for introducing concentration cells and related electrochemical concepts. A pair of aqueous cupric ion solutions, one concentrated and dark blue and the other dilute and light blue, were separated by a removable plastic barrier, as sketched below. The prediction asked of the general chemistry class in a ConcepTest was what would happen when the barrier was removed: Would the two solutions, now in direct contact, retain their colors; would the dark solution become even darker and the light one even lighter; or would the two become indistinguishable in color? The initial vote was in favor of the dark solution becoming darker and the light one lighter! The instructor may occasionally be thunderstruck by predictions such as this and can use them to advantage to obtain the class' attention. In this case, the barrier was simply removed and the color of the entire solution shown to became uniform. This demonstration engaged the class in discussions of diffusion and its directionality and of the spontaneous direction of current flow in an electrochemical concentration cell.

A similar result occurred during a demonstration involving the reaction of Ni and Al powder, pressed into the shape of a bar, to make, quantitatively, NiAl alloy in a thermite-like reaction (6; and see below). Magnetic properties of the bar before and after the reaction were shown to be completely different. The ConcepTest question was whether an elemental analysis conducted before and after the reaction would yield the same or different results. About half the class initially thought the analyses would be different! The ConcepTest drove home the true significance of an elemental analysis through peer instruction.

About the Collection. The examples of ConcepTests given here have been used in a two-semester general chemistry sequence. Preceding each are key words indicating the topic(s) associated with the question. Instructors will recognize many of the questions as relating to standard material. Our experience has been that framing ideas with ConcepTests has engaged the class in making predictions and constructing interpretations. This often puts a very different spin on the way these concepts and demonstrations are perceived by the class. In some cases material is taken from "Teaching General Chemistry: A Materials Science Companion," and is so indicated by "Companion." Answers are bold-faced. A short index of the key words has been included to facilitate finding questions for a given topic. The authors would welcome comments and additional suggestions for questions applicable to courses throughout the chemistry curriculum (

(1) Mazur, Eric Peer Instruction: A User's Manual; Prentice Hall: Upper Saddle River, NJ, 199, pg 253. Also, World Wide Web server (URL:
(2) Ellis, A. B. Chemtech; 1995, 25, 15-21.
(3) Herschbach, D. R. J. Chem. Educ. 1993, 70, 391.
(4) Ellis, A. B.; Geselbracht, M. J.; Johnson, B. J.; Lisensky, G. C.; Robinson, W. R. Teaching General Chemistry: A Materials Science Companion; American Chemical Society: Washington, DC, 1993, pgs. 15-16.
(5) Ibid. pgs. 49-50.
(6) Ibid. pg. 332.

Acknowledgments. We thank Sheila Tobias and Eric Mazur for making us aware of the ConcepTest methodology. We thank Larry Dahl, Clark Landis, John Moore, Fleming Crim and David Phillips for helpful comments. The construction of the Chemistry ConcepTests web site has been generously supported by the National Science Foundation through grants awarded to the ChemLinks (NSF DUE-9455918) and New Traditions (NSF DUE-9455928) curriculum reform projects.