Did the dark moths have an advantage in the dark forests? If so, the change in the moths was a result of natural selection. Natural selection was proposed by Charles Darwin to explain how new species evolve. All types of living things have small differences between the individuals in the species. If one of those differences allows the individual to live longer, they will likely have more offspring. As that trait is passed on, the population starts to look more like the successful individual.
Over time, the species changes. In , J. Tutt suggested that the peppered moths were an example of natural selection. He recognized that the camouflage of the light moth no longer worked in the dark forest. Dark moths live longer in a dark forest, so they had more time to breed. All living things respond to natural selection. Over other species of moth were observed to darken over time in polluted forests. What do you think would happen if the simulation continued for another four generations?
In Part A, you were instructed not to preferentially select one color bean or another. However, did you ever find yourself looking for one color over another? How did a different type of environment affect natural selection of the prey population? What happened when the prey were living in the environment made up of the darker wild rice compared to the white rice environment? What would happen if a non-visual predator moved in, such as one that hunts at night using smell?
Would certain types of beans still have an advantage? In this simulation, white or black beans were different variations within the prey population. What would be necessary to consider them separate species? How realistic is the model for simulating predators and prey in Part A?
Explain your reasoning. In Part B, what affected predator efficiency? Where are any variations in predator efficiency between group members when everyone was one predator type all chopsticks or all forceps? Do you think predator type affects prey selection?
Why or why not? In Part B, why were you asked to replicate your predator trials? How is this different from the multiple trials in Part A? Two researchers conducted the observations. After practicing on two class sections and determining our final behaviors in each category see above , the two of us observed the same two class sections simultaneously one in each treatment , ensuring that we were observing the same students at the same time by nodding or other non-verbal communication.
Since the two treatments necessarily included different behaviors Table 1 , we calculated agreement separately for each treatment. The Affect category showed the most consistent inter-rater reliability across the two treatments.
After those jointly observed sections that we used to calculate inter-rater reliability, we individually observed 10 lab sections 5 of each treatment over the course of 4 days, balancing across time of day.
These are the observations we include in our analysis. We recorded the behavior of each student in the section about 5 times average of 4. Because of the different nature of the lab activities, group size differed between treatments: students in the physical sections were divided into two groups, with an average of For analysis of the observation data, we collapsed some of the subcategories in each behavioral category: for the Motor category, we lumped all of the activity-related behaviors; for the Gaze and Motor categories, we lumped all off-task activities.
This analysis includes all observations so the sample is number of observations physical: , virtual: , not number of students. Because we intentionally did not know the identity of the students we observed, we cannot compare observations of students in the classroom with their performance on the assessments. In both treatments, student performance on the natural selection assessment improved after completing their lab activity.
In the physical treatment, the average score on the post-test increased to Similarly, in the virtual treatment, the average score on the post-test was greater For both treatments, those changes represent medium effect sizes for the pre-test to post-test comparison, and small effect sizes for the pre-test to delayed post-test comparison, suggesting that both lab exercises were effective at teaching natural selection, and that the effect persisted for several weeks.
To compare student performance between treatments, we calculated normalized change scores as a measure of student learning gains and conducted a repeated measures ANOVA. Normalized change is the fraction of potential improvement achieved by each student; 1 indicates improvement to a perfect score and 0 means no change from the earlier score.
The normalized change was essentially the same for both treatments Fig. To evaluate how much students retained their knowledge of natural selection, we compared the normalized changed score between the pre-test to the delayed-post test.
Students in the physical treatment had a mean normalized change of 0. Learning gains from the physical and virtual simulation activities. Students completed identical natural selection assessments before the lab, after the lab, and then 1 month after their lab. Normalized change represents the fraction of possible score improvement that the student achieved.
Centerlines show the medians; box limits indicate the 25th and 75th percentiles; whiskers extend to the minimum and maximum values. Students in both treatments showed an improvement from pre- to post-test and pre- to delayed post-test.
There were no significant differences in normalized gain scores between the treatments for either comparison. Thus we have no evidence that either activity elicited greater learning gains. Treatment, gender, or whether or not their native language was English were not significant predictors. To assess how often students use common misconceptions and key concepts when writing about natural selection in their own words, we compared responses to graded post-lab assignment consisting of short answers prompts and multiple-choice questions.
There were no differences between treatments in scores on the three multiple-choice questions mean scores: physical 2. For the short-answer questions, we compared the frequency of key concepts and misconceptions in student responses between the two treatments, calculated from the presence of the concept or misconception in their written response to short answer questions.
Other concepts, such as the random nature of mutation, the requirement that a trait be heritable for natural selection to occur, and using genetic drift to explain some changes, were less frequent but still occurred in similar frequencies in student responses in both treatments. Two misconceptions were fairly common in student answers in both treatments: mutations are an adaptive response to the environment, and traits change because of need Table 5.
In order to assess student enjoyment with the simulations, we administered a short survey after they completed their assignments and post-tests. Higher scores indicate higher levels of enjoyment. Student self-reported enjoyment of laboratory activities. Students completed a survey after completing the lab activities and post-tests.
They rated four statements about enjoying the lab on a scale of 1 completely false to 6 completely true ; their ratings across the four statements were averaged to calculate their overall enjoyment rating. Higher scores indicate higher reported levels of enjoyment. Students in the physical treatment reported significantly higher enjoyment. In each category, we were looking for evidence that students were involved in the lab, either through direct participation in the activity as evidenced by verbal and motor behavior or through active observation of the activity as evidenced by direction of gaze and affect.
For each of the four behavior categories gaze, verbal, motor, and affect , we compared the distributions of student behaviors into the relevant sub-categories between treatments using Chi squared tests Fig. Observations of student behavior while doing the physical simulation and virtual simulation lab activities. Student gaze a , verbal behavior b , motor behavior c , and affect d were recorded in mutually exclusive categories.
There are significant differences between treatments in the distributions among behaviors in each category. The distributions between behaviors are most different between the two treatments in the gaze and affect categories Fig. Notably, we saw more off-task behavior usually texting or other cellphone activity in students in the physical treatment, as can be seen in the gaze and motor behavior Fig.
Students in the virtual treatment overwhelming looked only at the computer screen, in contrast to the diversity of gaze directions in the physical treatment Fig. There are less striking differences in behavioral distributions in the verbal and motor categories. Students in the virtual treatment talked more to their peers Fig. Surprisingly, we observed more activity-related motor behavior in the virtual treatments typing on the keyboard, moving the mouse, or pointing at the screen; Fig.
This could be due to the fact that the group sizes in the labs doing the physical activity were larger, so although a few students in each group actively participated in the activities, many did not although they may have participated in discussion. The overall pattern suggests that most students were engaged by both the physical and the virtual lab activities, with a somewhat higher level of student involvement in the virtual treatment, given the higher proportion of students focused on the task and the lower proportion of off-task gaze, verbal, and motor behaviors.
In this study we assessed student learning gains, self-reported enjoyment, and observed involvement with two natural selection simulations, one physical and the other virtual, as implemented in sections of a large-enrollment introductory biology laboratory course.
The two simulations both attempt to address basic student misconceptions, and have the students act as active agents in the process of natural selection. Student performance improved in both treatments, with no difference in student learning between the two treatments.
Overall, our results suggest that there is no clear-cut advantage to either the physical or virtual simulations we tested. We also assessed student answers to short answer questions by scoring for the presence of particular concepts or misconceptions in each answer and comparing the frequency of concept and misconception use in the two treatments.
Students in the two groups showed no significant differences in presence of concepts or misconceptions in their responses to the short answer questions Tables 4 , 5. Together these results suggest that both simulations are equally effective at improving student understanding, and help students grapple with the key concepts of evolution by natural selection. Misconceptions about natural selection are persistent; we are more likely to overcome these with repeated instruction using a variety of approaches Kalinowski et al.
Our results suggest that either of these simulations can be a valuable tool for improving student understanding of natural selection in introductory courses, in combination with other types of instruction.
Despite the similarity in outcomes, students differed in their reported enjoyment of the activities. To assess student enjoyment, we asked students to answer several questions using a Likert scale. Students in the physical treatment reported enjoying the activity significantly more than students in the virtual treatment Fig. The enjoyment rating of students in the virtual treatment suggested that they also tended to enjoy that activity, but did not rate it as highly. Of course, self-reported enjoyment gives only part of the picture.
Whether students were engaged by the activity and what kinds of interactions occurred during the simulation also affects the success of an activity. We observed 10 lab sections 5 for each treatment to assess student behavior—specifically gaze, affect, motor and verbal interactions.
Broadly, observations of both treatments showed most students were engaged in the simulation. However, students in the physical activity were involved in more diverse activities during the observations. During the physical simulation students moved about the room, and also had free time to socialize and engage in other off-task behaviors. These opportunities for participating in more varied activities, moving around more, and socializing may contribute to the higher enjoyment ratings of the physical simulations Fig.
Our quantification of student gaze dramatically demonstrates the difference in how students participated in the two lab activities. On the other hand, the variety of gaze direction in the physical activity could have other benefits.
Chien et al. They found that students participating in the virtual simulation activity showed longer fixation duration, which they suggested implies deeper cognitive processing. They also suggested that students in their virtual treatment could concentrate more on the relevant aspects of the task.
Our observations of verbal and motor behavior show that most students in both treatments were not interacting verbally and not actively manipulating components of the activity during any given observation Fig. Different patterns were seen in the less common verbal and motor behaviors. Students in the virtual activity were more likely to be speaking with their classmates. Not surprisingly, students in the virtual activity were more likely to be engaged in motor behavior relating to the simulation manipulating the computer , whereas those in the physical activity were more likely to be writing Fig.
Interestingly, in their study, Chien et al. In our observations, the affect shown by a majority of students in both simulations was one of focused concentration Fig. While the proportion of students showing focus was quite a bit higher in the virtual simulation, students in the physical activity were more likely to show a positive affect, consistent with the higher enjoyment ratings that the students gave to the physical activity Fig.
This may be a consequence of the fact that a greater proportion of the students in the physical simulation were not directly involved in the activity at any given time.
It is interesting that despite the greater level of off-task and negative affect, the enjoyment rating of the Physical activity was significantly higher. This points to a limitation in the self-reported enjoyment rating—we cannot distinguish between reported enjoyment being due to the activity itself or to the social setting of the activity, which allowed for more off-task behavior.
Another limitation of both self-reported enjoyment and observations of student involvement is that they are likely to be highly context-dependent and therefore not generalizable to other classroom situations. However, they can provide insight into some best practices for implementing these activities. Some small changes in the implementation of both simulations could increase the proportion of students actively involved in either movement or discussion.
Instructors in the physical activity could pose questions for groups to discuss after each round of the simulation, and both activities could have benefited from more whole-class discussion. The large group sizes in the physical sections limited involvement, so where feasible, smaller groups would allow more students to actively participate at any given time; alternatively, instructors could ask students to switch roles so that more students have the opportunity to act as predator or recorder.
Groups for the virtual activity could also be smaller, improving student motor involvement, but even in groups of 3—4, instructors can have students rotate control of the keyboard and mouse throughout the class period.
Student learning gains may be so similar between treatments because, although the simulations differ in instructional mode, the activities share similar learning objectives. They both actively illustrated that trait variation can lead to differential reproductive success, and that this can lead to change in phenotype frequency over time in a population. The student assumed the role of predator in both activities, although the participatory aspect of the simulation is much more limited in the virtual simulation.
Both simulations also allowed students to work in groups, although group size differed substantially between the two activities. These broad similarities may account for the lack of difference observed in learning outcomes. Despite these similarities, the activities do differ in many details, and given these differences, we expected we might see some differences in outcome in specific areas.
One difference was that the virtual simulation includes many imbedded questions, with automatic feedback targeting misconceptions expressed in incorrect responses.
Some questions involved visualization of data, whereas others asked students to reflect on what results meant.
In class observations, we saw that these questions prompted student discussion within their groups, and these conversations were more common than in the physical treatment Fig. Because we used two simulations that are already in wide use, we did not carefully match the physical and virtual activities in this study. Thus, there may be different inherent advantages to each activity that nonetheless led to roughly equivalent learning through different routes.
For example, benefits of the immediate feedback to questions in the virtual simulation may have been comparable to benefits from the movement-based and game-like participatory nature of the physical simulation. It is also possible that there may have been differences in learning that were not detectable with the assessments we used.
Some important differences in the implementation of the two natural selection simulations in our study could have led to different student experiences, which could potentially impact both enjoyment and learning gains. The virtual activity included carefully worded step-by-step instructions, and feedback that encouraged students to repeat questions or activities until they have completed them correctly.
The physical simulation included directions at the beginning, but students may have been able to alter portions of the procedure, unless a TA redirected them. Based on our classroom observations, it appeared that the role of the teaching assistant tended to be quite different in the two activities. In the virtual activity, after helping the students access the module, most TAs then only answered occasional questions.
In the physical activity, the TA was more actively involved in all stages of the activity. In addition, some of the TAs had taught the same lab course in previous years, and therefore had experience with the physical activity, whereas none had taught using the virtual simulation before.
The TAs may have had different levels of enthusiasm for their activity; both the prior experience and the enthusiasm of the TA would likely influence the student experience of the activity. In our observations, the TAs for the physical activity were more likely to engage the whole section in general discussion; TAs for the virtual simulation may have benefited from more direction from course instructors on how to incorporate more whole-class discussion in the activity.
While we did survey students about their enjoyment of the two activities, we did not ask them to self-assess their own learning gains or their opinions about the value of the activity. Had we done so, we may have uncovered other student opinions about the lab in addition to simply enjoyment Pyatt and Sims ; Wiesner and Lan In a study comparing physical engineering lab activities to remote and simulated labs both conducted by students outside of the lab classroom , Corter et al.
Anecdotally, we heard some students and TAs express that they felt the physical activity was too juvenile and not appropriate for a college lab; however, many students clearly enjoyed themselves, and opinions about the value of physical manipulation and the game-like competition of the physical simulation are likely to vary substantially among students.
Our results align with other recent studies finding no difference in learning gains between physical and virtual activities Chen et al. Thus future research should start to focus more on other aspects of student experience so that we can learn more about the contexts in which one or the other type of activity might be preferable.
Given the equivalent learning gains we found in the physical and virtual natural selection simulations we studied, decisions about which activity to use can focus on what is most useful and feasible in a specific classroom context. Each of the simulations we studied has benefits and limitations. Students reported enjoying the physical activity more, but they were also more likely to engage in off-task behaviors, most likely due to the necessarily larger group size—this in itself might make the activity worth avoiding, depending on the specific student population and classroom environment.
This finding points to what may be a shortcoming in participatory simulations in general—benefits may be more likely to accrue to the participants than to the bystanders, so careful design and implementation of activities can maximize the proportion of students who participate.
On the other hand, the physical activity may be helpful to some students who will enjoy or learn better from the social and tactile interactions in such a task Corter et al. We feel the activity could be improved by more frequent breaks for reflection, answering questions, and discussion as a class. The virtual activity has the distinct advantage of not requiring a large group, space, or materials, although it clearly requires access to computers and is susceptible to technology failures.
It could be performed as an out-of-class activity, although this could then reduce the known benefits of group discussion Linton et al. Virtual simulations in general can be more flexible and allow for students to view events from multiple perspectives National Research Council ; Rutten et al.
Some previous authors have pointed to simulations being less costly than some comparable physical lab activities that require expensive equipment, supplies, or animals Dewhurst et al. Virtual simulations in some cases can be a more efficient use of class time Gibbons et al.
The implementation of the virtual activity in a lab setting could be made more effective with more support and direction for TAs about how they might more actively guide the lab and interact with the students, such as suggesting points in the simulation where they might pause for a class-wide discussion, and providing discussion questions and ideas for whole-class data collection and comparison of results.
These changes would likely increase the satisfaction of both students and TAs with the activity. When choosing which active learning strategy to use for natural selection, instructors must necessarily base part of their decision making on available resources and financial considerations; but of course, the relative effectiveness of a given activity for achieving the desired learning objectives should be a key consideration Freeman et al.
In our study, we assessed two popular simulations of natural selection and found them both to be effective, and saw no differences in learning gains between the simulations.
Both of these activities are suitable choices for improving student understanding of natural selection. Depending on the particular circumstances and learning objectives, one or the other might prove more effective. The results from our student survey and classroom observations may help inform considerations about how the simulation might work in a specific context. Although in our study the two simulations we compared show similar learning gains, it is unlikely that all activities or strategies for teaching natural selection will prove equally helpful Nehm Furthermore, as argued by Freeman et al.
Addressing undergraduate student misconceptions about natural selection with an interactive simulated laboratory. Evol: Educ Outreach. Google Scholar. Perspective: teaching evolution in higher education. Article PubMed Google Scholar. Altmann J. Observational study of animal behavior: sampling methods. American Association for the Advancement of Science.
AAAS science assessment website. Washington: American Association for the Advancement of Science; Development and evaluation of the conceptual inventory of natural selection. J Res Sci Teach. Article Google Scholar. Student conceptions of natural selection and its role in evolution.
Using Avida-ED for teaching and learning about evolution in undergraduate introductory biology courses.
0コメント