Creation of an online stoichiometry course that melds scenario based leaning with virtual labs and problem-solving tutors

Course features and rationale

This section discusses some of the course features along with the rationale for these features. Taken together, these rationales capture our assumptions regarding the nature of the learner in these environments.

Topics covered
The target audience is students who have had high school chemistry but who need a review of stoichiometry. We assume a basic familiarity with the meaning of chemical formulas, such as H₂O, and reactions, such as that for the burning of hydrogen in air: 2 H₂ + O₂ 2 H₂O, but the students may have forgotten most of what they learned about stoichiometry as a tool.

The effort began with an analysis of the topics and concepts that are typically covered in a high school course and are measured by ACS and AP exams. This list of objectives was used as a guide to the development. The course begins with foundation topics such as dimensional analysis, significant figures, the mole, and molecular weight (molar mass) followed by solution, composition and reaction stoichiometry. The most advanced topics are empirical formula, limiting reagents, titrations, and mixture composition. The syllabus can be easily customized to offer any subset of these topics.

One area in which we intentionally deviate from the typical sequence is by covering solution stoichiometry early in the course. This allows us to include solution-phase reactions while covering topics such as limiting reagents. Our goal here is to better prepare students for topics such as equilibrium and acid-base chemistry, where these concepts show up almost exclusively in solution. We also include titration, since it provides a nice example of a quantitative analysis technique and serves as an authentic application of the limiting reagent concept.

Scenario
Since the course was to be a linear sequence, we felt the strong need, as discussed earlier, to set the course in a real-world application that serves to contextualize the tools of stoichiometry. The scenario we chose is arsenic contamination in the groundwater of Bangladesh. The emphasis shifts from casting this as a human tragedy that chemists can help alleviate to the challenges facing modern analytical chemistry.

One goal of the scenario is to help make the uses of stoichiometry in the domain more explicit. In a previous work we compared what we teach in chemistry to the activities of the domain. [2] Our analysis of what chemists actually do identified three core behaviors: analysis, synthesis, and explanation. Of these, current instruction occurs almost exclusively in the explanation category, and our goal was to choose scenarios that bring this more into balance with the domain. Since stoichiometry is perhaps most central to analysis, we chose an important application of analytical chemistry. Synthesis is addressed by the portion of the course that discusses the attempt to convert local materials into powders that absorb arsenic. Some attempt to include explanation is made by setting the empirical formula practice in the context of analyzing ground samples to determine the form of arsenic that is present, but this is rather weak since we do not connect back to an explanation of how arsenic got into the groundwater of Bangladesh.

In addition to the motivational advantages, the arsenic scenario may provide cognitive advantages. In particular, by using the scenario to highlight the utility of the stoichiometry tools, we may be providing a memorable location to which students can attach their knowledge. Some examples of how this particular scenario may serve such a role include:

• The world health organization limit is phrased in micrograms of arsenic (As) per liter, but As exists in water as oxides. Composition stoichiometry allows one to extract the mass of As from the mass of oxide.


• The scenario brings out the distinction between quantitative and qualitative analysis by providing examples of both in the scenario context.


• The scenario discusses empirical formula as a type of qualitative analysis by using it to provide information on the form As has in the groundwater.

An additional layer of benefits can be envisioned for scenarios that enable one to more easily navigate the problem space. This requires that the scenario have aspects or characters that map to specific aspects of the problem solving process. A potential location where this occurs in the course is when students are asked to determine the amount of arsenic that can be absorbed by a powder made from locally available materials. The proportional reasoning required by this problem may be aided by being attached to the powder in the scenario, invoking the intuition that twice as much powder will absorb twice as much arsenic. In the latter part of the course, the emphasis of the scenario switches to the difficulties that arise when detecting small a mounts of material. The scenario may at this point aid the problem space by focusing attention on the relative magnitude of the numbers involved.

Video explanations
In addition to videos that convey the scenario, most of the explanations are done through videos with voice over narration. A principle goal and review criteria for these videos was that the explanations include not only the how, but also the why of the procedures. An attempt was made to go beyond explaining each step in the problem solving process, and include both the bigger picture of the problem solving strategy and the motivation for each step in the problem solving process. As an explicit example of emphasizing the "why", consider the video on the mole, [Understanding the mole] which attempts to explain the utility of the mole concept and give instances when it is useful.

As discussed in more detail below, creation of videos is considerably more demanding than producing the equivalent content in a text-only manner. Our hope is that the video accrues a number of advantages that lower the barrier to understanding the material. In particular, videos may allow students to keep their visual attention on a chemical or algebraic representation while hearing various aspects of this representation described. Such an approach may be especially helpful, since the various numbers and symbols in a chemical reaction are loaded with meanings regarding microscopic constructs. For instance, in CH₄ + 2 O₂ CO₂ + 2 H₂O the subscripts convey molecular structure while the inline numbers convey rules for combining molecules to make new molecules. As we identified earlier, this is a location of cognitive complexity because the notation does not directly support the differing meanings of the numerals. Video demonstrations with voiceovers allow students to hear such distinctions through inflection and gesture while focusing attention on the representation itself.

The issue of the relative benefits of video versus text is complicated by the possibility that the ideal modality may depend strongly on the where the student is in the learning process. It is possible that students who are learning a new set of complex material may benefit from video due to the arguments given above. However, students who are reviewing material and refining their knowledge may not need more than the text; the auditory information, while not harmful may not be helpful to such students. Some recent work by Kurt VanLehn supports this idea.[3]

Tutors
A video explanation is typically followed by a simple tutor that serves as a bridge between the expository presentation and autonomous practice of a concept's procedures. These tutors pose a series of questions that walk students through a problem solving process.

Example of a flash tutor

Two forms of scaffolding are provided, feedback and hints. Feedback on an incorrect response causes the entered response to turn red and invokes a message that helps guide the student to the correct response. A correct response turns green and gives reinforcement through a brief explanation of why the response is correct. Hints, about 3 to 6 per response in a tutor, can be invoked at any time by clicking on the hint button. The first hint typically reminds students of the goal of that particular problem solving step (relating to the why of the procedure), with follow-on hints giving more detail on how the step may be carried out. The last hint is a "bottom out" hint, meaning it gives an explanation that includes the answer.

Empirical formula tutor

The goal of these tutors is to provide a simple form of fading in the problem solving support. At one extreme, students can drill down to the bottom-out hint, in which case the tutor serves as a fully worked example. At the other extreme, students may answer each question without reading hints, taking advantage of only the feedback mechanisms. Studies on similar tutors [4] suggest that effective use of such help systems varies depending on the student's prior knowledge and metacognitive skills. Ideally, students would use the hints frequently when they first encounter a topic and then have their use fade as they shift to a practice mode. Analysis of student behavior with the tutors through examination of log files, for instance, may help us determine the extent to which beginning college students can make effective use of these help systems

The course also includes three more involved tutors that help students with the more complex calculations (empirical formula, limiting reagents, and mixture composition). The parameterization is meant to provide variation between instances that is comparable to that found in different end-of-chapter textbook problems. These tutors, especially that on empirical formula, provide a more fine-grained support for the problem solving with the interface capturing essentially every step. [snapshot and link] The interface needed to enable monitoring at this level of granularity has the disadvantage of providing strong cues to the problem solving, making it especially important that this scaffolding fade appropriately. Our current approach to fading is to first ask students for the answer to the problem and then fall back to the heavily scaffolded interface if they request help or fail after two attempts [1].

Virtual labs (V-Labs)
When appropriate, a course section ends with a V-lab that allows students to apply their knowledge in an environment that is more authentic than paper and pencil problem solving. Example tasks include determining the As content in a sample of well water, carrying out dilutions, and designing and performing titrations. The system provides hints that remind students of the goal of the activity and give some general advice on how to solve the problem. Feedback on the student response to the problem is also provided. In addition to correctness, the system analyzes the responses for anticipated student errors and provides appropriate feedback if such an error is discovered. The problems are parameterized by, for instance, randomly generating unknowns. After three failed attempts to solve a problem, the system gives the student the correct answer and then requires the student to reload the problem such that new random parameters are assigned.

Virtual Lab density activity

The V-labs are meant to help students attach the mathematical procedures they are learning in the course to authentic laboratory chemistry. Our student observations suggest that such attachments are not automatic and instead require explicit practice. For instance, after performing a dilution calculation prompted by the question, "What volume of 5.7 M glucose is needed to create a 100.0 mL of a 0.32M glucose solution?", it can take a student some time to figure out how to carry out the dilution in the virtual lab. This is not a user interface issue. Rather, the student appears to have translated the initial chemistry goal into an algebraic problem solving goal and once the algebraic goal is met, it can take some effort to remap this back onto the chemistry goal. It is tempting to dismiss this phenomenon as students failing to "transfer" what they learned, but a closer or more fine-grained task analysis shows that students have been asked to practice one thing and then are assumed to be able to do something else that is quite different. Practice in the V-Lab is designed to ameliorate that situation.

More advanced virtual lab problems ask students to design and carry out their own experiments. In such cases, students must connect the procedures and concepts they learned in the course to laboratory manipulations. Our hope is that such connections will make it more likely that the procedures and concepts are retained and invoked as needed in future learning.