Designing Graphing Activities

 Bowen, G. M., Roth, W. M., & McGinn, M. K. (1999). Interpretations of graphs by university biology students and practicing scientists: Toward a social practice view of scientific representation practices. Journal of Research in Science Teaching, 36(9), 1020-1043.  The authors define a framework based on the perspective of graphing as a social practice instead of an information processing, cognitive skill perspective.  This framework consists of 5 domain-specific areas to capture the degrees of purposeful participation graphing practices.  Ten undergraduate students in a second-year ecology class along with six experimental and theoretical scientists from the fields of ecology and physics participated in this study.  All participants were asked to interpret a graph depicting birth/death rates as a function of population size and how it will inform species conservation efforts.   Student data were gathered in a small group setting to capture the group construction of understanding during a seminar discussion followed by a small group interview to discuss an exam question which involved a similar task.  Experimental scientists interpreted and discussed the same graphs during an unstructured interview.  The scientists approached the task differently from each other, drawing on resources related to their expertise.  While the theoretical ecologists were able to draw more from ecology knowledge, they approached the task in a similar way to the theoretical physicist; they both focused on the graph-mathematical model transformations with little regard to the underlying ecology concepts.  In contrast, the two experimental ecologists instantiated the graph with specific examples to make sense of it as it relates to conservation.  By comparison the undergraduate ecology students exhibited several breakdowns in making sense of the graph (and a similar one on an exam) which was a reflection of their developing domain-specific knowledge from which to draw examples and apply relevant concepts (e.g. rate vs. density vs. population size).  This paper emphasizes the need to teach the concepts and practices of a discipline together with the representations of the domain.

Konold C, Higgins T, Russell SJ, Khalil K (2014)  Data seen through different lenses.  Educational Studies in Mathematics  88 (3), 305-325. This study describes the ways in which pre-college students view the nature of data and how they work with it.  The nature of students’ reasoning about data were derived from three data sources collected as part of qualitative studies conducted within mathematics and statistics education intervention programs for K-12 learners.  The authors identified four different perspectives that students use when interpreting data:  pointers, case values, classifiers, and aggregate.   Pointers refer to perspectives on data in which the data and the larger event from which the data originate are not differentiated.  Case values perspectives involve describing data based on attributes of a single case in the data set.  Classifier perspectives involve providing information about the frequency of certain observed case types.  Finally, the data as an aggregate perspective refers to the perceptual unit being the entire data distribution with features that are not specific to any single value in the distribution.  While there is a suggested hierarchy across data perspectives, with younger students tending to have data perspectives on the lower end of the hierarchy, this does not mean that there is a progression of preferred data perspectives to some higher level.  Depending on the purpose and context, different data perspectives are appropriate; there is no one single correct form. Different perspectives on data have implications for both graph construction and interpretation.  Data perspective has a strong influence on how a data display is selected and what comparisons are represented.  Further, interpreting a graph could be influenced by the viewer’s data perspective which could be different from the graph creator’s intent.

Åberg‐Bengtsson, L., & Ottosson, T. (2006). What lies behind graphicacy? Relating students’ results on a test of graphically represented quantitative information to formal academic achievement. Journal of Research in Science Teaching, 43(1), 43-62. This study describes the construction of a test of graph interpretation as well as the identification of factors that may be important to consider when teaching and evaluating secondary students’ graph interpretation.  The authors design a graph interpretation test using a set of 18 graphics comprised of line graphs, scatter plots, bar graphs, and cartograms.  The contexts for the data and graphics were drawn from textbooks the students used in classes and scenarios of everyday experiences that would be relatable to the students.  Question types were multiple choice interpretation or reading tasks in which students had to read off a value from a graph and write it down.  Students in the ninth grade from 5 different schools in Sweden completed the test and academic achievement data were gathered (n = 355 students). Knowledge of graphs and what they are used to depict, and the mathematics knowledge needed to extract values and understand the data displayed were important factors in students’ performance on the assessment.  However, knowledge and familiarity with the context and content of the data and the scenario from which they arose were perhaps more critical factors, even with experts.  This has implications for the teaching graphing to students and interpreting students’ competence with graph interpretation if the content and context of the data are not well-known to the students.

Xiong, C., Van Weelden, L., & Franconeri, S. (2019). The curse of knowledge in visual data communication. IEEE transactions on Visualization and Computer Graphics, 26(10), 3051-3062. This study examines a well-documented psychological phenomenon in language and decision making called the ‘curse of knowledge’ within the context of noting data trends in a graph. Undergraduate student participants were told one of several back-stories related to a study whose data were displayed in a graph.  Graph data were closely aligned and salient to information given in one of the backstories.  In one part of the study, participants who were given the backstory were asked to rank salient features and also predict what the other participants would find most important in the graph.  Participants were instructed to ignore the information they knew, reflecting on what they felt would be obvious to any viewer of the graph. The informed participants predicted that others lacking the contextual information would note the same trends they did, which proved to be false. The assumption that general patterns would be innately observed in the data underscores the need to provide well-designed graphs with contextual features (e.g., graph title/caption, variable labels, color cues, etc.) which align the reader and graph creator to appreciate the important features of the graph. This study adds to the growing knowledge around the impact of viewer’s prior familiarity with the content on their ability to generate inferences from data in graphical formats resulting from top-down processes in graph comprehension. Even seemingly simple graphs can lead to divergent readings and interpretations by views depending on their familiarity with the system under study.

Schultheis, E. H., & Kjelvik, M. K. (2015). Data nuggets: bringing real data into the classroom to unearth students’ quantitative & inquiry skills. The American Biology Teacher, 77(1), 19-29. In support of instruction connecting science and mathematics for developing students’ quantitative and science process skills, the authors outline the intent, design, and use of Data Nuggets in the classroom. Data Nuggets, a freely available online resource for K-16 educators, engages students in exploring a real-world problem using authentic scientific data generated by science practitioners. Approximately one hundred Data Nuggets (at the time of summary writing) of varying reading and graph proficiency levels are posted to the author described online collection (http://datanuggets.org/search-current-data-nuggets/), with topics spanning a range of biological problems (e.g., gene expression, climate change, evolution). Utilizing a case-based design, each Data Nugget models an approach for the explicit instruction of graphing that guides students through a problem by providing (a) relevant contextual information (i.e. background knowledge on the topic and the process of the researcher who developed the experiment), (b) the scientific question and testable hypotheses, (c) data to address the scientific question and to generate a graph as a form of evidential support, (d) practice interpreting graph and tabular data, and (e) the opportunity to draw conclusions and propose future steps to extend the research. The article concludes with a discussion as to how the learning intervention attends to recommendations made in national reports (e.g. Vision and Change) and can be implemented by instructors.

Shah, P., & Hoeffner, J. (2002). Review of graph comprehension research: Implications for instruction. Educational psychology review, 14(1), 47-69.of-Class Graphing Activities Increases Student Engagement and Learning Outcomes. Journal of microbiology & biology education, 18(3). This review article summarized the cognitive science research literature up to the time of its publication, ending with an evidence-based list of implications for instruction for improving student competence with graph comprehension.  Three areas were examined in particular:  (1) visual features of the graph, (2) viewers’ knowledge of and about graphs, and (3) viewers’ knowledge about and expectations of the data depicted in graphs.  The authors make nine recommendations for graph design as well as four additional recommendations for graph instruction. The first is to teach graph reading in the context of the relevant discipline rather than in an abstract context, which allows students to apply their disciplinary and graphing knowledge as well as learn how graphs can be used to critically evaluate data. The second is to have students examine the same data in different representations.  This can help students realize how graph design can influence one’s interpretation and assist with forming connections between the presented data and effective visual display features.  The third is to explicitly focus on the links between visual features in the graph and the meaning that they have to the actual phenomena from which they came. This activity promotes better sense-making by minimizing the abstract nature that the data may take on in the context of the graph, which does not always directly translate to the actual variables in reality.  Finally, the authors recommend that instructors make graph reading metacognitive as well as an act of retrieving facts or data points from a representation. Students should be encouraged to approach reading graphs as an interpretation and evaluation task as well as to reflect on their own knowledge and expectations and how this affects their ability to read a graph.

Dennen, V. P. (2004). Cognitive apprenticeship in educational practice: Research on scaffolding, modeling, mentoring, and coaching as instructional strategies. Handbook of research on educational communications and technology, 2(2004), 813-828. One of the ways in which students can move from novice- toward expert-like competencies is through interactions with and/or guidance from others (e.g., instructors, peers, etc.) with more advanced experience. This approach builds on the social constructivist tradition that learning is a socially negotiated process and can formally take place in apprenticeship-like models. In the context of learning concepts and the development of intellectual skills, this is termed a cognitive apprenticeship as novice learners complete authentic tasks situated in the field of practice under the guidance of a more experienced mentor. Of particular importance to cognitive apprenticeships is that the ultimate goal is for a learner to progress to a practitioner of the application of the knowledge and skills under study, not merely a knowledgeable observer. As such, there is a progression of expertise development through the use of several methods – or framework – aimed at supporting and guiding the learner, making tacit knowledge and practice explicit and visible. These methods include: (1) modeling thinking processes, (2) explaining the rationale behind activities, (3) coaching to monitor and support students when needed, (4) scaffolding to support students heavily at first and fading as they become competent, (5) having students metacognitively reflect on their own knowledge and performance, (6) having student articulate into words the conclusions of their reflections, and (7) letting the students explore new ideas related to the knowledge being learned.  The article elaborates further on each of these methods, providing links to learning theory and evidence from practice.  Given the complexity and nuances in creating, reading, and interpreting graphs, the cognitive apprenticeship model provides a useful teaching model for supporting students as they become part of the community of competent practitioners and educated citizens.

Patterson, T. F., & Leonard, J. G. (2005). Turning spreadsheets into graphs: An information technology lesson in whole brain thinking. Journal of Computing in Higher Education, 17(1), 95-115. Graph creation requires the graph maker to engage in analytical thinking to accurately analyze a dataset, select appropriate data to transform, provide correct labels , and draw logical conclusions. Additionally, the graph creator must also interject artistic, holistic, and creative thinking to create a compelling picture of data in the reader’s mind. To assist students in learning how to navigate the multiple cognitive processes needed to translate raw data into effective displays, the authors suggest a three-step graphing process. First, students need to select the data of interest, understand relevant statistics, develop a message to communicate, and physically sketch a graph on paper. Second, students should be shown examples of appropriate and inappropriately constructed graphs generated by visualization software (e.g, Microsoft Excel) with explicit discussion around graph type selection, aesthetic criteria, and how to use the software to generate graphs. Third, for visualization purposes, students then use the software for graph creation. An example intervention of these steps is discussed in which students explored the data prior to considering how to communicate it. In the year prior to the “whole brain” data transformation activity, students made many types of common analytical (e.g., selecting the appropriate graph type) and aesthetic (i.e. “look and feel”) errors. Following the teaching intervention, in a later semester with a separate student population, the authors noted a decrease in prevalence of the previously identified errors. The authors further report that students generally demonstrated a deeper understanding of data transformation in Excel and showed a greater appreciation of the emphasis on the visual message in comparison to their earlier counterparts that did not participate in the intervention.

Renninger, K. A., & Hidi, S. E. (2021). Interest development, self-related information processing, and practice. Theory Into Practice, 1-12. Drawing on findings from research literature in psychology, education, and neuroscience, the article explores the centrality of interest in learning, and how educators can support students’ interest development. This focus stems from the position that interest, one’s willingness to engage with content, is both a cognitive and affective motivational variable that benefits attention, goal setting, sustained engagement, and performance. Specifically, the authors argue to the instructional value of helping learners identify self-related connections to the content (or self-related information processing) in benefit to their interest development by (a) triggering interest, which encourages learners to seek out further information and persevere in understanding material, or (b) deepen the interest through sustained engagement for those with existing interests. In explanation to the importance of supporting interest development, the authors summarize previous studies from various content areas and learning settings regarding the benefits of interest on learning for learners of all backgrounds. The Four-Phase Model is shared for distinguishing learners’ developing interest from (1) triggered situational interest to (2) maintained situational interest to (3) emerging individual interest to (4) well-developed individual interest, with evidence-based insight as to how instructors can help cultivate student interest by phase. The authors additionally provide background on self-related information processing, highlighting the benefits when learners work with personally relevant content to increase engagement and activation of the reward circuitry in the brain. Allowing students to generate meaningful personal connections with the content promotes interest as well as sustains opportunities to advance conceptual understanding. For example, student use of data of interest to them could promote stronger interest in learning graphing and related concepts.  The article concludes with a discussion of general instructional implications, which can be readily considered through the lens of promoting student motivation in developing data representation skills.

DeBoy, C. A. (2017). Student Use of Self-Data for Out-of-Class Graphing Activities Increases Student Engagement and Learning Outcomes. Journal of Microbiology & Biology Education, 18(3). In this article, the author investigated how differences in data acquisition (i.e. self-generated versus instructor provided) affected the learning and engagement of students in a 200 level “biology of women” capstone course for biology majors and non-majors at a historic women’s college. During a unit on hormonal regulation, students were separated into two treatment groups differing in data acquisition on basal body temperature, in which one group was provided data by the instructor and a second collected self-generated data based on their own reproductive cycle (30 continuous days of data collection). All students were trained to collect pulse and stress-level perception data. To assess if the mode of data affected learning, students were assessed three times over the semester: after a lecture on hormones but before a graphing activity, after a graphing activity, and on the final exam. Comparisons of direct (quizzes) and indirect measures (surveys) for students using self-generated versus provided data suggest that while both activities increase learning outcomes, use of self-data compared with provided data has a greater impact on increasing learning outcomes and enhancing confidence in graphing skills and graphing efficacy.

Hug, B., & McNeill, K. L. (2008). Use of First‐hand and Second‐hand Data in Science: Does data type influence classroom conversations?. International Journal of Science Education, 30(13), 1725-1751. Increased availability of public datasets provide students with opportunities to increase their understanding of diverse phenomena by exploring data that they may be unable to gather in standard K-16 classrooms.  However, it is unclear how working with these second-hand data are similar or different to working with self-generated data (first-hand data) as it relates to student learning.  This study aimed to understand how the use of first- or second-hand data affected students’ interaction with data within small group and whole-class discussions.  The context of this study were middle school chemistry and biology classrooms in which each class completed units working both with first- and second-hand data.  Analysis of video and audiotaped class sessions focused on themes relating to: the nature of discussion around measurement, data source, data manipulation, limitations of the data, patterns/inferences, conclusions, use of content knowledge.  With the exception of discussions around measurement, which only occurred when students worked with first-hand data, all other themes were present in classroom discussions in varying frequencies regardless of whether the data were first or second-hand.  When students engaged with first-hand data they more often engaged in discussions regarding the collection, source, and limitations of the data.  In contrast, when students engaged with second-hand data, they often manipulated the data, identified new data patterns, drew conclusions from the data, and considered content knowledge.   These results suggest that students meaningfully engage in many science practices around data, but may focus on different features when working with their own or other’s data, an important consideration for instructors when planning learning activities.

Kjelvik, M. K., & Schultheis, E. H. (2019). Getting Messy with Authentic Data: Exploring the Potential of Using Data from Scientific Research to Support Student Data Literacy. CBE—Life Sciences Education, 18(2), es2.77(1), 19-29.related information processing, and practice. Theory Into Practice, 1-12. Authentic data from scientific research can be used in the classroom to engage students and develop their cognitive skills in problem-solving and quantitative reasoning (e.g., interpretation, argumentation). In this article, the authors outline five main features of authentic data to characterize the complexity of data-centric learning activities: scope, selection, curation, size, and messiness.  Scope of data is determined by the number of variables and information contained within those variables. Data sets narrow in scope can be used to identify specific relationships between variables, whereas, data sets broad in scope may challenge students to select appropriate variables that address the question at hand. Selection of data can range from the instructor providing the exact variables of interest to students independently defining the dataset. Here, opportunities for autonomy in data selection benefits student understanding and ownership in the learning activity. Curation of data or data handling, consists of organizing and preparing the data for visualization, such as transforming raw data (i.e. sums, means, percentages) or merging multiple datasets into one dataset (i.e. datasets with asynchronous collection time frames, missing data points, mismatched scales). The total number of data points that exist in a dataset are defined as the Size and can determine if students can use paper and pen or data software to visualize data in graphs or use data as evidence to support a claim. The last feature of authentic data sets is the Messiness of data (i.e. outliers, missing data points, unexpected trends, variability). Although messy datasets can be frustrating for students, they create learning opportunities to strengthen critical thinking skills and highlight the normalcy of data variability. The authors suggest the five features of authentic data are correlated to some degree and should be instructionally scaffolded over time, starting students with small datasets limited in scope and with clearly defined variables to progressively larger, less-defined, and “messy” data sets. Deliberate practice with authentic data will give students the opportunity to understand the value of data and improve their critical thinking as they collect, analyze, and interpret data.

Schultheis, E. H., & Kjelvik, M. K. (2020). Using Messy, Authentic Data to Promote Data Literacy & Reveal the Nature of Science. The American Biology Teacher, 82(7), 439-446. The authors in this article offer advice to K-16 instructors on engaging students with “messy” data (i.e. real-world data that contains variability).  This variability arises from natural variation or from the data collection methods itself (e.g., missing values, outliers, unexpected trends). Messy data is a valuable teaching tool in promoting data literacy for future science practitioners as well as the general public because it evokes critical thinking, increases an understanding of the nature of science, and inspires additional research questions. Further, experience with messy data may help students overcome common misconceptions about the quality of data that demonstrates variability, which may result in-part from “smooth” trend lines often depicted in textbooks and other public data displays. Data types can fall into two broad categories: first-hand (collected directly by the students) and second-hand data (from outside sources). Although students may be limited to the types of phenomena and data they can collect in the classroom setting, first-hand data are important for engaging students and helping them build an understanding of the natural world. Second-hand data may allow students the opportunities to study data collected over long periods of time (e.g., variations in climate), but the lack of data familiarity may lead to the lack of understanding of the variables or the datasets themselves. Therefore, teaching with both first and second-hand data with appropriate practice and scaffolding is vital for providing productive learning experiences. The authors provide a sample sequence of implementation of authentic messy data in the classroom and suggestions on scaffolding data.

Kastens, K., Krumhansl, R., & Baker, I. (2015). THINKING BIG: Transitioning your students from working with small, student-collected data sets toward “big data”. The Science Teacher, 82(5), 25-31. Engaging students in inquiry and gathering data to answer questions of interest to them has many advantages, including a greater sense of engagement and ownership as well as a deeper understanding of the data and  study system.  The increase in publicly available data for use by people outside the community of experts who collected them offers an opportunity for educators to engage students with large, diverse sets of information not easily gathered in a classroom context for studying complex problems. However, engaging students with these datasets has its challenges including their size, larger number of different variables present, and details available in the metadata.  Therefore, the approaches students use for data analysis may be insufficient when working with ‘big data’.  In this pedagogical essay, the authors provide four evidence-based classroom activities to scaffold learning as they transition from working with simple, small student-collected datasets to complex large datasets collected by others:  1) Data puzzles is an activity in which students are given graphs of large datasets to interpret while paying attention to how the data shown relate to biological phenomena.  The nested dataset approach allows students to connect with the system under study by (1) collecting their own small dataset, (2) then working with datasets collected by other student groups, and (3) eventually expanding out to examining professionally collected data on the same topic.  The predict, observe, explain approach helps to connect students to the data that they will explore (e.g. in a database) by creating a motivating factor to get them to explore more deeply as they evaluate their prediction.  Finally, hypothesis arrays can be used when students might not know a lot about the system under study by giving them choices to explore and evaluate using information drawn from the dataset they are examining.  These strategies were developed for high school teachers, but are relevant in undergraduate classrooms to provide initial scaffolding before students embark on open inquiry or course-based research experiences over longer periods of engagement and higher student autonomy, for example.

Rosenberg, J., Edwards, A., & Chen, B. (2020). Getting Messy with Data. The Science Teacher, 87(5), 30-34. In this short informative article, the authors provide a summary of free data tools (i.e. Desmos, Google Sheets, JASP, R, and CODAP) and provide implementation strategies on data collection, analysis, and statistical modeling.

Science Education Resource Center (SERC) at Carleton College. The SERC website provides a range of teaching activities intended to strengthen K-16 students’ higher-order quantitative reasoning skills. The site currently hosts approximately 400 graphing activities for the lecture and laboratory college classroom across disciplines, including biology (n=~40), environmental science (n=~130), and health sciences (n=5). The posted teaching activities provide lesson outlines (e.g., learning goals, assessment practices) and resources, data sets to engage students with real-world “messy” data  (e.g., climate change and plant distribution, water quality), provide math and statistics modules, and explain the use of graphing tools (e.g., how to use Excel to store and present data).

Roth, W. M., & McGinn, M. K. (1997). Graphing: Cognitive ability or practice?. Science Education, 81(1), 91-106. In this review article, the authors advocate for viewing the creation and interpretation of graphs as a competence developed through regular collaborative activities in the field rather than as an innate cognitive ability. Framed by practices experts (e.g., scientists) engage in, the authors describe the use of graphs in three ways:  as semiotic objects (collections of signs and symbols representing different aspects of measured variables), rhetorical devices (conveying information to communicate and convince viewers), and conscription devices (a way to invite and engage others into discourse).  In all of these roles, creating and making meaning from graphs is contextualized within the community of others who understand the discipline and system under study.  Viewing competence with graphing through this lens has classroom implications for the teaching and assessment of students’ graphing competence.  The authors advocate that students learn the mechanics of graphing (creating the semiotic object), the reason for graphing (rhetorical use), and collective sensemaking with graphing (conscription use) within authentic contexts and the community of practice.  Here, students should be encouraged to use graphing not merely as an end point to an activity, but as part of a cycle of collective meaning making with their instructor and peers.  From the perspective of assessment, the authors argue that graphing competence is not measured as some universal set of skills, contending that it is  less about creating the graph artifact itself or reading off the symbols and trends in a graph, but rather about using graphs to understand the systems under study and in a way that is grounded in the discipline and communities of others familiar with that discipline.  Teaching and assessing graphing in this way can not only lead to increased competence with graphing, but can facilitate a broader learning of science.

Shofner, M. A., & Marbach-Ad, G. (2017). Group Activity to Enhance Student Collaboration, Graph Interpretation, and Peer Evaluation of Ecological Concepts in a Large-Enrollment Class. Journal of Microbiology & Biology Education, 18(3), 18-3. The authors provide examples of two inquiry-based graphing activities that were implemented in a large-enrollment introductory biology course. Graphing goals for students in this course are to: interpret graphs, compose hypotheses, incorporate biological concepts into their graphs, work in groups, and engage in the peer review process. Prior to the first classroom activity focused on biogeography, students were asked to read about the biology scenario in the textbook and were given a brief introduction in class. Students were divided into small groups and given a worksheet focusing on writing hypotheses related to the biological context and interpreting a graph. To promote collaboration, students were given 15 minutes to work in their small groups with one student assigned to record responses. Afterwards, students were asked to exchange their worksheet with another group and were given 10 minutes to review and comment on another group’s answers. Students then received their own worksheets with feedback and asked to review suggestions prior to large group discussion. A similar process was taken with the second activity as students constructed a graph from a data table to explore the relationship between climatic change and population dynamics of a species. Following the first activity, students were asked to reflect on two questions regarding working in a group and the process of peer review. Most students (80%) rated their experience positively with a smaller group (15%) that reported negative feedback attributed to their preference of listening to information in lecture, working by themselves, lack of communication between their team members, and lack of confidence on the biology concepts to provide substantial feedback to others. To ease students into teamwork, the authors recommend assigning each group member a specific task (e.g. recorder, facilitator, presenter).

Group Work  Guide. This evidence-based teaching guide provides a rationale for and practical tips for implementing group work in undergraduate biology courses.

 Tanner, K., Chatman, L. S., & Allen, D. (2003). Approaches to cell biology teaching: cooperative learning in the science classroom—beyond students working in groups. CBE-Life Sciences Education, 2(1), 1-5. The authors review the features of cooperative learning and contrast it with other types of learning in science classrooms.  They present five essential elements of successful implementation of formal cooperative learning in addition to some pedagogical approaches for informal group work.

diSessa AA (2004)  Metarepresentation:  native competence and targets for instruction.  Cognition and Instruction  22:  293-331. This review article aims to describe the difference between representational competence (RC, one’s ability to produce and use a variety of standard representations) and metarepresentational competence (MRC, one’s critical, reflective and inventive strategies that can be applied to any representation).  MRC entails five key principles:  (1) inventing or designing new representations, (2) critiquing and comparing the adequacy of representations and judging their suitability for different tasks, (3) understanding the purposes of representations, generally, and in particular contexts and how they are useful, (4) explaining representations, and (5) learning new representations with ease.  The MRC perspective acknowledges that there is no singular ‘best’ representation for a given set of data, but rather that a number of existing and new representations could be appropriate for a given context and purpose, but that each has affordances and limitations that need to be explored. The author provides an overview of the vast representational competence literature and how that has shaped the knowledge base that has informed instruction. The author then describes, with examples from the literature, how the lens of MRC can provide further insight into the development of student competence with representations and an acknowledgment of intrinsic ability that students possess which is a resource for further learning.  The strict adherence to norms and a singular ‘best’ solution for a representation is challenged and the author highlights the importance of an exploration and critique of multiple representations.

 Angra, A. & Gardner S.M. (2016).Development of a Framework for Graph Choice and Construction.  Advances in Physiology Education 40: 123–128. The ability to create and make use of visual representations to solve problems is known as representational competence.  In the context of graphing in biology this requires a knowledge of different graph types, data types, and features of the biological system from which the data arose.  Metarepresentational competence (MRC) extends from this and incorporates reflection as an important component for successful construction and reasoning with graphs and other external representations. In graphing, reflection reveals students’ own awareness of their understanding of graphs and gaps in their knowledge. In this paper, the authors present two validated learning and instructional graphing tools, a Step-by-Step Guide and Guide to Data Displays that can be used to teach students to reflect on their decision-making processes. The Step-by-Step Guide is divided into three phases, with the last phase being a five-step reflection phase that prompts the student to: (1) check the alignment of their graph to the research question and hypothesis; (2) provide the advantages of the representation; (3) provide the disadvantages; (4) the take-home message; and (5) other ways to represent data. The Guide to Data Displays organizes  the common types of visual representation (graph and tabular data) used in biology as well as their respective advantages and disadvantages in a table. Provided in the supplemental materials is a blank Guide to Data Displays which can be used as a tool to help students of varying experiences reflect on graphs throughout the semester.

McFarland, J. (2010). Teaching and assessing graphing using active learning. MathAMATYC Educator, 1(2), 32-39.  In this article, the author identifies three common challenges to graph learning in the college classroom as well as instructional approaches to overcome these potential challenges. The three challenges were:  1) generating student interest in effective graphing practices (e.g., selecting data appropriate graph types), 2)  encouraging students to critique graphs in their textbooks and the popular media, and 3) the student assumption that using graphing software will automatically produce appropriate graphs for a given data set. To overcome these challenges, the author uses the principles of active learning to share a 90-minute activity that requires biology undergraduates enrolled in a laboratory (or lecture) course to collaboratively practice cognitive and metacognitive skills to improve graphing skills. After an introduction on graphs, students are divided into small groups and asked to hand draw a graph from a provided data set, and reflect on their decision-making through a series of questions around the purpose of graphs, criteria for appropriate graph construction, and alternative graph types. After graph construction, students are engaged in a class discussion about the importance of graphs and published examples generated by other professionals and scientists. Finally, each group shares their graph, observations about the data trends, and questions evoked by the graph. To translate this activity to long-term student learning, students are required to complete two graph self-assessment questions on graph choice and quality, which they must submit along with their lab report. End of the semester feedback from students on the graphing activity and subsequent reflections reveal the impact of multiple practice and reflections rounds.

Matuk, C., Zhang, J., Uk, I., & Linn, M. C. (2019). Qualitative graphing in an authentic inquiry context: How construction and critique help middle school students to reason about cancer. Journal of Research in Science Teaching, 56(7), 905-936. This study investigated how the incorporation of a qualitative, social practice-approach to graphing in scientific inquiry impacts pre-college biology students’ understanding of both graphs and science. A qualitative approach to graphing is described by the authors to differ from a quantitative approach by focusing on the identification of general data patterns and trends rather than pointwise details (i.e. the plotting and interpretation of specific data points). In the two-part study, data were collected from middle school students (n= 147) before and after a socially-contextualized multi-week unit on cell division and cancer to assess the value of integrated qualitative graphing activities. For each study, the authors analyzed pre- and post-test assessments designed to measure learning gains on key concepts and targeted graphing skills (i.e. critique, construction) using standardized rubrics. The results demonstrate that the incorporation of qualitative graphs in instruction benefited students’ conceptual understanding as well as developed competency in using graphs as narrative tools to express this understanding. In addition, the authors reported on how the actions of critiquing and constructing qualitative graphs distinctly benefited students in unique ways that improved their ability to integrate their knowledge of science and graphs. Specifically, critiquing graphs helped students improve their scientific explanations within the unit, while constructing graphs led students to link key science ideas within both their in‐unit and post‐unit explanations. The authors conclude that the inclusion of qualitative graphs into inquiry-based activities can simultaneously strengthen students’ graphing competencies and conceptual understanding. Design considerations for critique and construction activities are discussed in the context of extant learning science and science education literature.

Tanner, K. D. (2012). Promoting student metacognition. CBE—Life Sciences Education, 11(2), 113-120. This resource is an essay that translates metacognition research into practical recommendations for instructors. The essay provides actionable advice that instructors can follow to promote metacognition in their students as well as in themselves.

Student Metacognition Guide. Graphing is a social practice of science in which the community of scientists question each other and jointly make meaning together from graphs.  This evidence-based teaching guide reviews seminal literature on the benefits of and practical tips for supporting student metacognition.  In particular, the authors provide a section on social metacognition in which students support each other’s metacognition through the sharing of their own metacognition and reasoning and questioning and supporting each other.  The role of instructors in promoting and scaffolding social metacognition are reviewed.