Understanding Student Perceptions of Collaboration, Laboratory and Inquiry Use in Introductory Chemistry
Carl Berger, Nancy Kerner, Yohan Lee
University of Michigan
Ann Arbor MI 48109-1259
The purpose of this study was to find if an introductory chemistry collaborative laboratory course had any impact on student perceptions of their collaboration, laboratory experience and inquiry skills and how their perceptions changed as they worked though the course and from one year to others. Four hundred sixty eight students were surveyed three times during the course with a 22-question perception instrument. Three hundred thirty two students were surveyed two years later. Perception scores were analyzed for differences in the perceptions of tasks involving collaboration, laboratory and inquiry. Results indicated significant perception differences both across the tasks and within the three time measures. Student perceptions of inquiry started much lower than those of collaboration or laboratory but gained much more. In a small but significant number of tasks, student perceptions deceased during the first part of the course and rose in the second part. For most of the tasks student perceptions gained little during the first five weeks of the course but made much larger gains in the last eight weeks of the course. Previous student use of specific tasks may have been related to decrease in perception during early weeks of the course. Low initial student perceptions of inquiry tasks may have been related to the level of difficulty of inquiry in beginning college chemistry courses. This was tested by a second sample of 332 students in the same course three years later. It was found that differences in beginning perceptions also might depend on the "home" college of the student taking the course.
Introduction Measuring and using perceptions (self-efficacy) in science courses is not new (Bandura, 1977; Betz & Hackett, 1983: Lent, Brown & Larkin, 1984). This paper describes the measurement of student perceptions in their progress in a collaborative inquiry laboratory class and compares two groups of beginning students, one in 1996 and the other in 1999. The class is a large (approximately 700 students) introductory general chemistry course. Students in the course are primarily freshman with no intention of becoming chemists. The course was redesigned in an attempt to successfully involve students in the development of chemical principles from data and engage them in scientific ways of thinking (Konigsberg Kerner, N., Penner-Hahn, J., Berger, C. & Dershimer, C., 1997). Key course features include an emphasis on processes and general concepts, problem solving in inquiry experiments that do not have a "right answer," and an emphasis on qualitative reasoning. The course seeks to expose students to the thinking and qualitative reasoning processes by which chemists organize and construct principles from data, make predictions, and design experiments. The aim is to provide students with a more accurate picture of the process of science while fostering qualitative reasoning skills.
Part of the course included the use of networked computers to support team based experimentation, collection of qualitative data, and data analysis (using commercial graphing software). Experimental tasks are purposely not divided. Teams replicate tasks but use diverse sets of team reagents. The varying sets of team reagents produce comparable data trends. Students are more likely to conclude that they are observing real trends rather than unique examples when they observe similar data from a variety of reagent sets (Konigsberg Kerner, N. & Penner-Hahn, J., 1994).
The instructional format has been redesigned to support a collaborative learning environment that encourages thinking and active engagement. Procedures include directives and questions to encourage students to think while conducting the experiment. For example, directives ask students to record a hypothesis and "expected observations if your hypothesis is true." Students must record a team consensus on most aspects of experimentation. At the start of every new experiment, teams are assigned specific research questions to answer during lab discussion. Questions focus on organizing, manipulating, or extrapolating from the class "data bank". For example, students determine relationships or non-relationships between the data (e.g., properties) and structure (e.g., ionic radius) or known scientific facts (e.g., electronegativity).
We wanted to know how students differed on their perception of twenty-two concepts and processes divided into groups of collaboration, laboratory experience and inquiry. Further we wanted to know how their perceptions changed as they worked though the course. Finally we wanted to know if the perceptions of students beginning chemistry differed over several years. We used the CoLabInq Self-Perception Survey reported on at a previous NARST meeting. (Lee, Y., Konigsberg Kerner, N. & Berger, C., 1998)
CoLabInq Self-Perception Survey (CSPS)
Survey Construction The question items were designed to determine the impact, if any of their collaboration, laboratory experience and inquiry skills. Three domain scales emerged and confirmed that the CoLabInq survey could be divided into three scales representing, respectively, collaboration, laboratory experience and inquiry skills. (Lee, Y., Konigsberg Kerner, N. & Berger, C., 1998) The COLLABORATION scale refers to skills related to working with others in the laboratory and conveying information to peers as in discussion presentations. The LABORATORY scale refers to scientific process skills that primarily relate to "hands-on" lab activities such as collecting and organizing data into graphs and tables. The INQUIRY scale refers to higher-order cognitive skills in scientific process such as predicting the behavior or properties of untested samples from the lab data or designing experiments.
Each of 23 items on the CSPS required a separate response on a five-point Likert scale for KNOWLEDGE (cognitive dimension), EXPERIENCE (behavioral dimension), and CONFIDENCE (affective dimension). The three separate dimensions were designed to assess their perceptions of how much they were learning, how much experience they were gaining, and at what level they rated their confidence in doing tasks required of them in the course. In this paper we will discuss the changes in student perception of the collaboration, laboratory and inquiry domain scales as the dimension analysis comparing knowledge, experience and confidence was included in a previous report (Lee, Y., Konigsberg Kerner, N. & Berger, C., 1998).
This questionnaire has been designed to measure your perception of your knowledge, experience, and confidence on various items.
With each statement are three indicators of your involvement. For each of the questions indicate how you feel about your knowledge, experience, and confidence.
This would mean that I have a great deal of knowledge (response of 5) about changing a flat tire, I have an average amount of experience (response of 3) with changing a flat tire, but I am not confident (response of 1) in my ability to change a flat tire
A similar multi-phasic approach for constructing question items has been used in the past in evaluation studies (Hungerman, Berger, Roderick, & Latz, 1976; Berger & Carlson, 1988; Berger, 1997). The resulting CSPS administered to students had twenty-two items. Students were asked to rate their levels of knowledge (cognitive dimension), experience (behavioral dimension), and confidence (affective dimension) on each item on a scale of 1(low) to 5 (high):
Circle one number in each of the three boxes.
1) Analyze data to reach conclusions.
2) Propose a hypothesis to account for an observation.
3) Design an experiment to test a hypothesis.
4) Use conclusions to make predictions about untested situations or samples.
5) Construct graphs to visualize data.
6) Contribute ideas to group discussion.
7) Obtain information from a chemical reference book.
8) Interpret the relationships represented in graphs.
9) Use my classmates as an information resource.
10) Use the periodic table to predict the properties and reactivities of samples.
11) Identify unknown samples of chemicals.
12) Design and produce an effective presentation.
13) Trust the contribution of my teammates when completing a group project.
14) Work effectively with my teammates while conducting lab experiments.
15) Recognize and characterize properties of chemicals based on observations.
16) Extract and use information from a data table.
17) Speak effectively in front of the class when presenting results.
18) Evaluate the work of others.
19) Use laboratory instruments to make measurements.
20) Collect, record, and organize data.
21) Control variables when designing an experimental procedure.
22) Recognize and characterize reactions based on observations.
Figure 1 The CoLabInq Self-Perception Survey (CSPS)
Initial findings As reported in a previous study, the three domain scales accounted for 58% of the total variance. Most educational studies show components accounting for 20-25% of the total variance. Thus these results are considered very good. Based on the excellent fit of the principal components to the prior determination of the scales, the original items were retained except for one (item 7), this item was dropped from all subsequent analyses. The items in each scale were then subjected to a reliability analysis using Cronbach's coefficient alpha. Values ranged from .82 to .88 indicating that the scales are reliable. Upon administring the perception survey in 1999 the three domain scales accounted for 54% of the variance with similar ranges for Cronbach's coefficient alpha.
Administration The CoLabInq survey was administered three times during the Fall 1996 semester: at the initial class meeting, the fifth (mid-semester) week of class, and at the final class meeting (week thirteen). A total of 1283 CoLabInq surveys were collected over the semester (n1=460, n2=337, n3=486). For Winter semester 1999 the CoLabInq survey was administered once at the initial class meeting. (n=332).
CoLabInq Survey Results
Differences in Perceptions by Scale There was a significant increase in the levels of the responses to all the items in the CoLabInq survey as the course progressed in Fall semester of 1996 (F=1011.9, p<. 0001). All of the scale dimension combinations achieved their highest mean response at the end of the semester:
Figure 2. Student perceptions of Collaboration, Laboratory and Inquiry
Both the COLLABORATION and LABORATORY scales demonstrated rather modest gains when compared to INQUIRY which made sizable leaps with each administration. This difference accounted for a significant scale-week interaction (F=90.6, p<. 0001).
The pattern of changes was not uniform over time. Not only were there smaller differences in means between weeks 1 and 5 for LABORATORY and COLLABORATION, what changes there were for LABORATORY were negative. While the overall effect was small, there were clear differences among the items that indicated that students perceptions of "constructing graphs", "using lab instruments" and "collecting and organizing data" actually decreased during the first five weeks of class The effect on the COLLABORATION scale was similar to that on the LABORATORY and items indicating "contributing ideas to group discussion" and "using classmates as an information resource" decreased for the first five weeks. Most interesting was the response to "working with a team" where perceptions of cooperation decreased significantly during the first five weeks and did not return to initial starting levels of the class by the thirteenth week.
To make sure the data we were gathering was not isolated incident, we compared the pattern of perception at the beginning of the term for Fall term 1996 to Winter term 1999. The results are shown in figure 3.
While the pattern of perception was very parallel, there were differences in perception between the two years. We realized that the proportion of students from the most representative colleges; Engineering and Literature, Science and the Arts (LSA), might not have been equal for the two terms. Traditionally, more engineering students enroll during Fall term than Winter. We prepared a contingency table for the years and schools with the results shown in table 1.
Figure 3. Comparison of Fall 1996 with Winter 1999 student perceptions
Literature, Science and the Arts
Table 1. Distribution of students from two colleges for two years
A Fisher exact of 0.0150 indicates that there is a stronger proportion of engineering students in Fall than in Winter. We next examined the graph comparing engineering students with those from LSA.
Figure 4. A comparison of student perceptions for two colleges at the university.
Some interesting pattern differences emerged. Notice that the LSA student reported equal perception on Collaboration but lower perceptions on Laboratory and Inquiry items as shown in figure 4. Students in LSA may perceive themselves as less able in the areas of Laboratory and Inquiry reflecting (perhaps) background, career path or learning preference.
Discussion and Conclusion
Students' perceptions of their INQUIRY competencies are lower than either LABORATORY or COLLABORATION skills. While all the perceptions increase over time the perceptions of Inquiry competencies remained lower that the others even at the end of the course The fact that the largest increases are shown in the INQUIRY domain scale may indicate that students believe the course is fostering higher-order cognitive skills. In a traditional laboratory the student is "expected to produce a verification of something that he already knows, and so ends up trained to ask what a result is supposed to be, not what it in fact is" (Pickering, 1980). In contrast, the approach of this collaborative inquiry course is to require students to formulate conclusions using the team generated and computer summarized "class data bank". In a traditional laboratory, students are concerned that their experiments "work correctly", i.e. produce the desired outcome or outcomes. In contrast, the course under study is both data-driven and laboratory inquiry-centric and thus students are encouraged to reflect on the implications and relevance of the class data. Thus questions concerning data analysis, making predictions, designing experiments, and drawing conclusions all showed marked improvement over the semester.
The fact that the COLLABORATION domain scale shows only a positive trend over time indicates that students perceive that the course is fostering these skills. In a typical traditional lab, individual reports are collected. In the collaborative inquiry lab students prepare a written and oral team report. Team members are required to share observations, data or insights, and record a team consensus. When entering data or observations into the computer, the team consensus is recorded. Oral presentations of group findings are both peer and instructor evaluated. The perception survey results indicate that such strategies are fostering collaboration skills. The fact that gains in COLLABORATION skills are more modest than INQUIRY skills over time could be due to an initial underestimate of the difficulty in doing group work. This is particularly supported by the difficulty students had with "working with a team" where their original perceptions were high and fell significantly during the first five weeks not even to return to initial levels by the thirteenth week.
The LABORATORY scale describes items, which primarily relate to "hands-on" lab activities and lower order cognitive skill demands (as opposed to the higher order demands of the INQUIRY scale items). The positive increase in ratings over time indicates that students perceive that the course methodology is also fostering these skills. As noted above, the items of "constructing graphs", "using lab instruments" and "collecting and organizing data" actually went down during the first five weeks of the course. It may be that upon starting the course the students regarded these as skills that they had already mastered. These perceptions are consistent over time but appear to depend on the "home" college of the student. Most of the students enrolled in the course have completed high school chemistry or some other science courses that have very beginning skills in these areas. After performing some experiments in a conceptually difficult and rigorous environment, perhaps their perceptions of their competency in these skills were challenged and therefore fifth week ratings went down. By the end of the semester their overall ratings increased due to use of these skills over time.
As noted above, both the LABORATORY and COLLABORATION domain scales demonstrated rather modest gains when compared to INQUIRY which made sizable leaps with each administration. INQUIRY domain skills such as using data to predict the behavior or properties of untested samples are systematically incorporated into all course experiments. It is worth noting that the domain items initially rated the lowest by students were the ones that dealt with INQUIRY skills. However, even more interesting is the observation that these same items showed dramatic improvement. It therefore stands to reason that students in the course are being exposed to undeveloped higher level cognitive skills. More importantly, if students’ perceptions are accurate, they also are improving in these cognitive skills over the course of the semester.
Bandura, A. (1977). Self-efficacy: Toward a unifying theory of behavioral change. Psychology Review, 84, 191-215.
Betz, N. E. & Hackett, G. (1983). The relationship of mathematics self-efficacy expectations to the selection of science-based college majors. Journal of Vocational Behavior, 23, 329-345. 2 2 Konigsberg Kerner, N., & Dershimer, C. (1992). Collaborative learning and networked computers: An experiment in introductory general chemistry. Abstracts of papers, 12th International Conference on Chemical Education, Bangkok, Thailand.
Konigsberg Kerner, N., & Penner-Hahn, J. (1994). Cooperative learning in an introductory inquiry lab. Abstracts of Papers, Chem. Ed.#205, 208th ACS, Washington, D.C.
Konigsberg Kerner, N., Penner-Hahn, J., Berger. C., Black, B. and Dershimer, C. (1996), Computer Assisted Collaborative Inquiry. Chemistry: Expanding the Boundaries, Proceedings of the 14th International Conference on Chemical Education, Royal Australian Chemical Institute, 205-206.
Konigsberg Kerner, N., Penner-Hahn, J., Berger, C., & Dershimer, C. (1997). Computer assisted collaborative inquiry in the laboratory. In J. A. Chambers (Ed.), Selected papers from the Eighth National Conference on College Teaching and Learning, (pp. 131-147). Jacksonville, FL: Florida Community College at Jacksonville.
Lee, Y., Kerner, N., & Berger, C. (1998). Student perceptions of their progress using computer-assisted collaborative inquiry in the laboratory, National Association for Research in Science Teaching. San Diego
Lent, R. W., Brown, S. D., & Larkin, K. C. (1984). Relation of self-efficacy expectations to academic achievement and persistence. Journal of Counseling Psychology, 31, 356-362
Pickering, M. (1980). Chronicle of Higher Education, 19, 80