Abstract
Learning to write a scientific manuscript is one of the most important and rewarding scientific training experiences, yet most young scientists only embark on this experience relatively late in graduate school, after gathering sufficient data in the lab. Yet, familiarity with the process of writing a scientific manuscript and receiving peer reviews, often leads to a more focused and driven experimental approach. To jump-start this training, we developed a protocol for teaching manuscript writing and reviewing in the classroom, appropriate for new graduate or upper-level undergraduate students of developmental biology. First, students are provided one of four cartoon data sets, which are focused on genetic models of animal development. Students are instructed to use their creativity to convert evidence into argument, and then to integrate their interpretations into a manuscript, including an illustrated, mechanistic model figure. After student manuscripts are submitted, manuscripts are redacted and distributed to classmates for peer review. Here, we present our cartoon datasets, homework instructions, and grading rubrics as a new resource for the scientific community. We also describe methods for developing new datasets so that instructors can adapt this activity to other disciplines. Our data-driven manuscript writing exercise, as well as the formative and summative assessments resulting from the peer review, enables students to learn fundamental concepts in developmental genetics. In addition, students practice essential skills of scientific communication, including arguing from evidence, developing and testing models, the unique conventions of scientific writing, and the joys of scientific story telling.
Introduction
Manuscripts are one of the main products of academic scientific research. Accordingly, learning how to effectively package observations and interpretations into a compelling, logical, and mechanistic story is a major goal of graduate training. However, for most graduate students, lessons in manuscript writing become available relatively late in their training – after students have generated sufficient data. Arguably, an intimate knowledge of the manuscript-writing process should precede and drive the experimental approach. We propose that modeling the processes of writing and reviewing a manuscript in the classroom can accelerate the development of skills that are essential for independent laboratory study (Table 1).
To address the need for early exposure to manuscript writing, we developed a two-part homework assignment that enables beginning graduate or upper-level undergraduate students to experience the process of writing and submitting a manuscript (Fig. 1). This assignment, which we call Manuscript 101, was tested and improved, after written student assessment, over the course of four offerings of a first year graduate level course in developmental genetics (n = 43 students). For Homework 1, each student is provided with the manuscript-writing instructions (Fig. S1) as well as one of four sets of cartoon figures (Fig. 2 and Fig. S2-5). The figures provided are cartoon data, and are provided in unsequenced order. Students are instructed to organize their figures in an order that tells the best story, to interpret the data, to draw a model figure illustrating their proposed molecular mechanism, and to write a manuscript using all of the figures. For the manuscript, students write Title, Abstract, Introduction, Results, and Discussion sections, according to the Author Guidelines provided in the Instructions for Homework 1 (Fig. S1). In the Discussion section, students are encouraged to provide alternative interpretations of their data, and to propose experiments that would discriminate among the possible interpretations. Students are also encouraged to state the predictions of their model, and to propose future experiments that would test these predictions. Thus, Manuscript 101 exposes students to the structure and organization of a typical scientific manuscript, but also provides students with a first-hand opportunity to experience the creative aspects of scientific story-telling and exploration.
The manuscripts are submitted to the instructor, who acts as journal editor in Homework 2, the mock manuscript review. During Homework 2, each student is asked to provide an anonymous review of a manuscript generated by a peer in Homework 1, according to the Reviewer Guidelines provided in the Instructions for Homework 2 (Fig. S6). The instructor may also choose to provide a review of each submitted manuscript.
Therefore, Homework 2 is both an assessment of the manuscript writing assignment, as well as an experiential opportunity in critical evaluation of scientific manuscripts. At the end of Homework 2, each student will receive the review(s) of his or her manuscript, as well as a formal assessment of both homework assignments from the instructor. Below, we provide the teaching tools (Table 2), and framework for implementing this practical exercise.
Results and Discussion
1. Homework How-To
Instructions for Homework 1 and 2 (Fig. S1 and S6) are provided to each student, along with a cartoon data set (Fig. S2-S5). The goal of Homework 1 is to produce a manuscript, which will be peer-reviewed as Homework 2. It is important that students receive instructions for Homework 2 at the same time that they receive the instructions for Homework 1 because this will enable students to understand the assessment criteria for Homework 1. Homework 1 instructions include an overview of the writing exercise, as well as section-by-section manuscript guidelines to clarify content, format, and word limits. Depending on the class size, multiple students may receive the same data set. In this case, students are instructed that they are permitted to work together to interpret their data, as long as intellectual contributions are acknowledged by coauthorship. However, the instructions also specify that each student is to submit an original, independently written manuscript.
When all students have submitted Homework 1, each manuscript is assigned an anonymous identifier that does not include student names. Each manuscript is then redacted to protect student identity, and then each manuscript is assigned to another student as Homework 2. Ideally, the data sets used by author and peer reviewer should differ. To help keep track of student names, their data sets, and anonymous identifiers, we provide an organizational spreadsheet (Table S1).
The goal of Homework 2 is to provide students with a realistic peer-review experience, from the perspective of the author and the reviewer. Homework 2 instructions include clear criteria for evaluating a manuscript. Students should also understand that Homework 2 cannot be assigned until each student has submitted Homework 1. Prior to each homework assignment, it is beneficial to review homework instructions (Fig. S1 and S6) during lecture, and to assess student familiarity of the structure and style of scientific writing (Table 3). Although most students will have received training in the reading and critical evaluation of scientific manuscripts, most have not yet engaged in the writing of manuscripts or manuscript reviews.
2. Homework Assessment
Tools for assessing student performance, as well as homework efficacy, are listed in Table 3. Logistically, it is helpful for the professor to create a spreadsheet to keep track of manuscript identifiers, reviewer assignments, and final grades. A sample organizational spreadsheet is provided (Table S1). For Homework 1, students receive two independent reviews of their manuscript – one written by an anonymous student peer, and one written by the instructor. If time permits, the instructor can also provide a more detailed assessment of Homework 1 by annotating each student’s manuscript.
In addition, students receive a grade for each of the two homework assignments. To facilitate grading, we have provided two scoring matrices (Tables S2 and S3), which closely track the expectations stated in the homework instructions. Results for the scoring matrices may be returned to the students, or summarized as comments. Typical assessment commentary includes a paragraph highlighting the major points that drove the grades for each assignment.
3. Designing New Data Sets
We have included sample data sets that were designed for a graduate level course in vertebrate developmental genetics. Manuscript 101 can be adapted to include additional model organisms and biological processes by devising new cartoon data sets. Here, we describe our approach to creating the cartoon data sets in order to provide a path for creating new data sets.
To devise our sample data sets, we used four criteria, which are generalizable to multiple biological disciplines. Our first criterion was that the cartoon data set should focus on processes and methods that have been described in lectures. In this way, the manuscript writing and reviewing homework can synergize with the didactic lesson. The second criterion was that each data set should highlight techniques for creating gain and loss of function (in genes or tissues) in order to test specific hypotheses. For example, our data sets focused on various model organism-specific approaches for altering gene expression. In this way, the homework teaches experimental approaches as well as concepts. The third criterion was that each data set should highlight standard approaches for evaluating experimental outcomes. For example, the endpoints of altered gene expression levels in our model data sets were embryo morphology and gene expression. Accordingly, our data sets included embryo morphology and gene expression data that would have been generated using microscopy, quantitative PCR, RNA-sequencing, in situ hybridization, immunofluorescence, flow cytometry, and western blotting approaches. Thus, the manuscript writing and reviewing homework is an opportunity to expose students to multiple experimental approaches for evaluating experimental endpoints. The final criterion used in creating our data sets was that several of our cartoon results could be interpreted in multiple ways. For example, a band shift on a gel could be due to alternative mRNA splicing or to protein cleavage. In this way, students are challenged to consider multiple alternative interpretations of an experimental test and to propose how they could discriminate between the possibilities in future experiments. New figures can be hand-drawn, or prepared using any computer illustration software.
Conclusions
Manuscript 101 is designed to model some of the most important scholarly activities of working scientists, so that students can experience these activities early in their training. An additional benefit of this activity is that it exposes students to the creativity that is intrinsic to experimental science, as well as the elegant structure imposed by the scientific method and manuscript conventions. From the instructor’s perspective, it is fascinating to observe diverse models that arise from a singular set of figures, and rewarding to coach beginning scientists through some of their first forays in science communication.
Supporting Figures
Figure S1. Instructions for Homework 1
Figure S2. Dataset A
Figure S3. Dataset B
Figure S4. Dataset C
Figure S5. Dataset D
Figure S6. Instructions for Homework 2
Table S1. Instructor’s organizational spreadsheet
Table S2. Homework 1 Scoring Matrix
Table S3. Homework 2 Scoring Matrix
Homework 1 Instructions – Write your own manuscript!
Overview
Goal
How-to
Assessment
Author Guidelines
Title
Abstract
Introduction
Results
Model figure with legend
Discussion
References
Format
Manuscript-writing resources
1a. Goal: write a manuscript from cartoon data
1b. How-To
Follow Author Guidelines (Section 2 of this talk)
You write: Title, Abstract, Introduction, Results, and Discussion sections
You draw: a model figure and (write its legend)
All writing must be your own
Collaboration on interpreting figures is encouraged
Collaborators must be listed as co-authors of your paper
Use your creativity to transform evidence into argument
Persuade your audience of your important new insight
Discover a new mechanism regula0ng the biological process at hand
1c. Assessments
2a. Author Guidelines - Title
Limit: 20 words
State the main lesson from the study
2b. Author Guidelines - Abstract
Limit: 200 wordsPocket-sized synopsis of the paper
Should entice the reader
Should provide pithy summary of which new lessons are learned
Should explain the significance:
Why would I have undertaken this study?
How could I get others excited about this study?
Should explain the discovery:
What do these pieces of data reveal about the process?
What is the mechanism by which X molecule regulates Y process?
Should explain the novelty:
Why is this new?
How does this differ from other studies?
2c. Author Guidelines - Introduction
Limit: 300 wordsSets the stage for the story
Provides background needed to understand the results
Cite real papers for bonus points!
Differs from Abstract and Discussion
Abstract is the summary
Discussion is the interpretation
2d. Author Guidelines - Results
Limit: 400 words per sectionUse section headings that summarize the key finding for each figure
sections should follow the figures in sequence
Refer to each figure panel parenthetically, the first time it arises
Include in each section:
Brief rationale for each experiment (i.e., the hypothesis tested)
Brief description of the experimental approach, including controls
Your observations
Brief interpretation (i.e., do your observations support or reject the hypothesis?)
Remember that data can often be interpreted in multiple ways!
2e. Author Guidelines - Model Figure and Legend
Legend Limit: 50 wordsA graphical summary of the mechanism you describe in the Discussion
Legend should explain the model briefly and define colors/symbols used
2f. Author Guidelines - Discussion
Limit: 600 wordsIntegrate all the interpretations into a mechanistic model
The model is a new discovery about a biological process (e.g., the way that a protein regulates a developmental event)
Refer to the model figure
State the predictions of the model
Propose experiments to test the predictions of the model
Propose experiments to distinguish between alternate interpretations of results
Avoid restating observations already explained in the Results section
2g. References
No limitCite real or fictinal papers throughout the text
Use Endnote or other reference manager
Use Cell reference style or similar
2h. Author Guidelines - Format
The Text11 point Times New Roman or Arial
Double-spaced
1-inch margins on all sides
Page number, lower right hand corner
Print to PDF
Number the figures and arrange in the order described in the text
Edit legends if needed
Print to PDF
Merge Text and Figure PDFs, in that order
Submit PDF of the complete manuscript to the instructor by email.
3. Other Resources
Development Author Instructions:
http://dev.biologists.org/site/misc/submissions.xhtml
Understanding the publishing process (Elsevier):
Homework 2 Instructions – You are the reviewer!
Overview
Goal and how-to
Assessment
Reviewer Guidelines
Summary
Assessment Criteria
significancee
Observation
Interpretation
Model
Clarity
1a. Goal and How-To
1b. Assessment
2a. Reviwer Guidelines - Summary
Limit: 300 wordsDescribe the biological process explored in the manuscript
Summarize the key findings of the manuscript
Briefly describe the paper’s strengths and weaknesses, using the five assessment criteria (next slide)
2b. Reviewer Guidelines – Assessment Criteria
Address each of the following with specific examples and suggestions for fixing the problem.
significancee
Do the authors explain the significancee of the study? Is their rationale for undertaking these studies compelling?
Observation
Are the authors’ descriptions of the data accurate? Do they include a description of positive and/or negative controls?
Interpretation
Are the authors’ interpretations supported by the observations? Are there alternative interpretations of the data that are not discussed? If so, what are they?
Model
Is the authors’ molecular mechanism explained clearly in the text and in an original figure? Is the model supported by the observations? Did the authors state the predictions of the model and did they propose experiments to test the predictions of the model? Would the proposed experiments definitively test the model? Are there experiments that would be better?
Clarity
Is the manuscript structured and formatted according to the Author Guidelines? Is the manuscript easy to read and free of jargon, typos, and grammatical errors?