Problems and Problem Solving in Chemistry Education: Analysing Data, Looking for Patterns and Making Deductions, ed. G. Tsaparlis, The Royal Society of Chemistry, 2021, pp. P005-P008.
Download citation file:
People looking at the field of chemical education from the outside might ask: Why has so much effort been devoted to the topic of problem solving in chemistry? Members of the chemical education community, however, could explain this in terms of either the challenges of teaching/learning chemistry, in general, or by using specific examples from their own experience.
The importance of problem solving in chemistry was addressed in a review we wrote almost 20 years ago.1 We began the review by noting that “… problem solving is what chemists do, regardless of whether they work in the area of synthesis, spectroscopy, theory, analysis, or the characterization of compounds”. Furthermore, we argued that “individuals who were successful in chemistry courses either developed good problem solving skills—more or less on their own—or brought these skills to their chemistry courses.” We also argued that “… we weren't doing as good a job as we could in helping less successful students learn how to build problem solving skills”.
Evidence to support the importance of studies of problem solving in chemistry was found in a study of Ph.D. dissertations carried out as part of a project funded by NSF.2 A total of 555 Ph.D. dissertations that could be considered to fall into the domain of chemical education were included in the analysis. Between 1953 and 2018, a total of 58 dissertations related to problem solving were found. This represented one of the largest categories in the study. Further insight might be provided by looking at the first few decades of the process through which studies that could be called science education research became what we now know as chemical education research. Between 1953 and 1989, a total of 30 Ph.D. dissertations related to chemical education could be found, five of which focused on problem solving.
There is evidence that the literature on research on problem solving in chemistry has evolved over the years.3 Originally, the literature was dominated by studies of high-school students. With time, the number of papers that dealt with research on college-age students increased, but it concentrated on the first-year course. Eventually, this work was extended to organic chemistry and physical chemistry courses. In more recent years, it has begun to include studies of graduate students. This pattern is consistent with the statement from our review of the problem-solving literature that it “runs the gamut from studies of high-school students working on stoichiometry or gas law problems through studies of problem solving by advanced graduate students in chemistry”.
Let's now turn to the way personal experiences have brought practicing chemists into discipline-based educational research on problem solving in chemistry. The author of this foreword was not trained as an “educator”, much less a “science educator” or “chemistry educator”. He began his academic career with a Ph.D. in inorganic and organic chemistry based on the use of 13C FT NMR spectroscopy. He then spent three years as a visiting professor at the University of Illinois at Urbana-Champaign (UIUC), teaching general chemistry and continuing his research as a practicing chemist.
Before each lecture in his general chemistry course at UIUC, he prepared an average of six pages of typewritten lecture notes that were carried into lecture and arranged on the lecture table. He found that students responded well to his lectures in the end-of-the semester instructor evaluations and he felt that he had done a good job explaining the material such as, for example, the concept of molarity.
But he made the mistake of looking at student responses to multiple-choice questions from the hour exams and found, to his surprise, that no more than about 60% of the bright, hard-working science and engineering majors in the course were able to solve “simple” molarity problems, in spite of what seemed to have been well-crafted instruction. Something had gone wrong in the process of teaching/learning and he concluded that the fault was his, not his students’.
While he was at UIUC he asked himself several questions about his NMR research: If he left the field, would anyone miss him? (They did not.) Was there any study that he might do that couldn't be done by someone else? (Apparently not.) So when he came to Purdue, he decided to focus his research on studies that addressed a phenomenon he eventually described in a paper on the constructivist theory of knowledge as follows: “Teaching and learning are not synonymous; we can teach, and teach well, without having the students learn”.4
At Purdue, he was fortunate to be given an office next to J. Dudley Herron, who was an educational researcher. He was also fortunate that colleagues in math and physics education as well as educational psychology were no more than a five- or ten-minute walk from the chemistry building. When a graduate student asked if she could do an M.S. thesis with him, he reflected on the importance of spatial ability in advanced undergraduate and graduate courses in chemistry. But he wasn't teaching one of those courses, so they chose to start by looking at the relationship between spatial ability and performance in the course he was teaching; a first-year chemistry course for science and engineering students.
Based on advice from colleagues familiar with educational research, they designed a study that looked at correlations between the students’ performance on a battery of spatial ability tests and subscores from exams or quizzes that seemed to focus on tasks that were the most likely to involve spatial ability.5 Correlations were, indeed, found of 0.32 (for multiple-choice questions on crystal structure) and 0.35 (for a free-response quiz on crystal structures) at a p-value of p <0.0001. But they also looked at subscores such as multiple-choice questions on stoichiometry from the first hour exam and the total score on a comprehensive final exam that were selected because they expected they were the least likely to involve spatial ability. Much to their surprise, they found correlations of the same magnitude (0.29 and 0.30, respectively) at the same p-value, p <0.0001.
These results were so unexpected that they chose to expand the study the next year to look at two large sections of the course for science and engineering majors as well as a lower-level course for students in agriculture and health science with a combined sample population of just under 2500 students.6 They looked at a total of 46 subscores covering topics that one might predict would involve spatial ability, such as molecular geometry (VSEPR) and crystal structures, as well as those one would expect would not, such as balancing chemical equations, stoichiometry, the gas laws, descriptive chemistry, and so on. This time they found statistically significant correlations between the total spatial score and 42 out of 46 subscores.
Analysis of the data from the second study clearly indicated that the correlation between the spatial ability tests and chemistry content was smallest for subscores that included questions that were routine exercises that could be answered by what they described as “mindless applications” of algorithms and largest for questions that were more likely to be novel problems. Although they eventually extended this work to organic chemistry,7 it became obvious from these studies that what they were really interested in was problem solving, not spatial ability.
At this point, they concluded that although studies based on traditional statistical analysis might tell them something about what was happening in a given content domain, it was not going to provide them with useful information about why this was happening, or what needed to be done to help students develop their problem-solving skills.8 They therefore turned exclusively to qualitative research designs.9 Furthermore, they turned their attention to courses at both the undergraduate and graduate level that built on the introductory course that had been studied previously. They even went so far as to apply these techniques to study the question of whether concepts such as “design” in engineering were related to problem solving.10
The author of this foreword is grateful for the work done by Georgios Tsaparlis and the authors of the individual chapters of this volume for undertaking a monumental task of bringing order to the literature on problem solving in chemistry, a topic that has interested him for almost four decades.
Arthur E. Kelly
Distinguished Professor (Emeritus) of Chemistry, Education and Engineering
Purdue University, USA