• There appears to be no correlation between the number of college-level physical science courses and the ability to correctly answer basic questions related to thermal equilibrium (1).
• Addressing students' alternate conceptions can do more than improve their understanding of concepts; it can also improve their problem-solving ability (2).
• The results suggest that problem solving can be a powerful means to increase the effectiveness of instruction, provided that the students are expected to question their decisions and understand the underlying structure of the algorithms they apply (3).

These three statements, each taken from the conclusions section of the paper referenced, lead collectively to a rather negative view of college and university curricula and teaching. (The first statement was reinforced by a recent conversation from which I learned that students who had just completed physical chemistry did not score significantly higher than those who had just completed general chemistry on a test that emphasized understanding of concepts (4).) It appears that students can take courses from which they gain little or no real understanding of physical or chemical systems. However, tools exist that could help us prevent this from happening. Why aren't we enthusiastically using those tools and energetically working to develop more?

Of course some people are doing just that, and some of them started a long time ago. One of those, Gordon Barrow, passed away on January 8 of this year at the age of 78. Barrow pioneered the use of programmed instruction in chemistry, first with written materials and subsequently with computer simulations that generated realistic data for students to analyze. He wrote many textbooks that emphasized the need for understanding of chemical and molecular-scale phenomena in addition to the ability to correctly answer numerical questions. As Executive Director of the NSF-sponsored Advisory Council on College Chemistry, Barrow oversaw publication of teaching aids and resource papers and organized conferences and workshops in support of better teaching of chemistry. His work in all of these areas was predicated on his belief that the reason that students had difficulty with chemistry was the way the subject was taught.

I used Barrow's computer-based materials when I taught physical chemistry, and they certainly helped students to think about real problems and develop good understanding of physical chemistry concepts. Barrow's careful observations of students must have revealed the same kinds of conclusions as the chemical education research studies listed above. He acted on those conclusions and successfully addressed some of the difficulties he saw. But many others have taught similar courses and similar students, failed to see the problems that Barrow saw, and therefore did not create the same kinds of effective learning environments he did. Why?

One reason is that it takes careful observation and thorough analysis to detect whether students are really learning chemistry. Tests made up of numerical problems are not sufficient and may even mislead us into a false sense of success. We need better means of testing whether students understand, and we need to induce more people to use them. The Chemistry Concept Inventory described by Mulford and Robinson (2, 5) is one example, as are the questions reported by Jasien and Oberem (1). It is sobering to realize that among students who had taken one or more semesters of physical science only one in seven could correctly answer a conceptual heating-curve question (see ref 1, question 3). Since the question contains five multiple-choice responses, the statistics of pure guessing would predict that one answer in five would be correct. Thus many students must hold conceptions about temperature and heat capacity that differ from the scientific consensus. Teachers of physical science scored only 25%–barely above the statistical result—so apparently many of the teachers also have conceptual lacunae.

Conceptual questions seem simple, even if they are not. I have often been criticized for dumbing down my course by including conceptual material, when in fact such questions are often more difficult for students, not easier. If you don't believe this, try incorporating into your exams some of the questions from the papers I have cited, or write similar conceptual questions of your own. You may be surprised at the results, and discrepant events are excellent motivators for change.

Literature Cited

1. Jasien, Paul G.; Oberem, Graham E. J. Chem. Educ. 2002, 79, 889–895.
2. Mulford, Douglas R.; Robinson, William R. J. Chem. Educ. 2002, 79, 739–744.
3. Ardoc, Dilek J. Chem. Educ. 2002, 79, 510–513.
4. Barnhard, Katherine I. Personal communication.
5. The Chemistry Concepts Inventory (accessed May 2002).