Skip to main contentSkip to footer content
 

FOR 694 Writing for Scientific Publication
Class Notes 2010

 

Date

Class

Assignments Due

Jan 20

Why Publish? 

Bring your questions

Jan 22

Discuss Getting Started

Getting Started Exercise

Jan 27

Choosing your Journal

Knowing your Journal Exercise

Jan 29

Figures and Tables

16 copies of Tables and Figures with Abstract (revised)

Feb 3

Figures and Tables

Feb 5

Outline and Objectives

Two copies of your Outline and Objectives

Feb 10

Results

Draft of Results (two copies)

Feb 12

Materials and Methods

Draft of Methods (two copies)

Feb 17

Writing exercise, in class

Bring examples of difficult passages

Feb 19

Preparation for peer review

Editing and proofreading

Submit Results and Materials and Methods Sections with Abstract (for First Peer Review)

Feb 24

Introductions

Bring examples from your field, n copies, number the paragraphs, staple

Feb 26

Statistical Considerations

Bring your questions

Mar 3

Advice, Responses to Reviews

First Peer Review Due

First Reviews.docx 
ecosystem-error-Response.pdf

Mar 5

Discussions

Bring examples from your field, include Conclusions and Summary if any

Mar 10

Progress on Introductions (shared in class)

2 copies double-spaced

Mar 12

Progress on Discussions

2 copies double-spaced

 

Mar 15-19 (spring break)

 

Mar 24

Mid-semester feedback

Anonymous feedback

Mar 26

Ethics

 

Mar 31

Readings on Peer Review

Submit  Rough Draft and Response to First Reviews

Apr 2

(no class)

 

Apr 7

Authorship and Writing

Authorship Exercise

Apr 9

Readings on Publication Productivity

Second Peer Review Due

Apr 14

Reviewer feedback and help session

Your review, your paper, or a section that needs help

Apr 16

Proposals

Bring examples of RFPs, proposals

Apr 21

Abstracts

Bring 7 copies of an example from your field

Apr 23

Work session

Sections for review

Apr 28

Abstract review

Title, by-line, key words

Abstracts due

Apr 30

Galley proofs, Steps to Publication

Final Draft of Manuscript and Response to Second Reviews

Wed, January 20:  Why Publish?

Introductions

Suzanne Conrad, FNRM, social science: Adirondack landowners and invasive plants and insects

Artem Treyger, FNRM, old-field succession and climate change

Peter Rockermann, EFB, riparian vegetation in NYS and invasive species, effects of ash removal by emerald ash borer

Carrie Rose Levine, FNRM, undergraduate thesis research, deer herbivory effects on seed bank development.

Anna Harrison, FNRM, landscape factors affecting beaver occupancy in the Adirondacks

Angela Sirois, EFB, effects of habitat alternations on bog turtle populations

Nick Pitel, FNRM, sugar maple decline following forest tent caterpillar defoliation, effect of site conditions

Laura Schifman, FNRM, water use in four varieties of willows using C isotopes.

Megan Skrip, EFB, habitat and ruffed grouse survival

Rakesh Yasarla, PBE, surface characterization of sugar maple extracts for bioalcohol production

Bali Quintero, FNRM, forest regeneration on abandoned farmlands in western Puerto Rico

Jim Arrigoni, EFB, biomonitoring for stream water quality in Hong Kong

Dan Gurdak, EFB, butterfly communities in agroforestry landscape in India

Kikang Bae, FNRM, fine root and microbial activity affect soil respiration

Bhavin Bhayani: PBE,  membrane separation of lignin from other elements of wood extracts, xylose and acetic acid.

Why Publish?

Suzanne is excited about the work and wants to share it.

Artem: it's important for your CV to get your next job

Carrie Rose: it contributes to knowledge and scientific progress.

Anna: publishing from your masters thesis shows that you have experience and some success with the scientific process.

Angela: my work will improve management practices on the ground

Nick:  My result can be used by other people and compared to other results.

Laura:  My work has management applications, and we want to affect future practices. Save the world.

Kikang:  I get ideas from other papers, and I want to do the same for others.  Writing papers will be part of my career.

Megan:  The process of peer review will improve the work.

Rakesh:  It will contribute to the completion of my thesis.  It will advance the field.

Bali:  My work could be the missing link to someone else's work.

Jim:  To fulfill obligations or meet expectations.

Dan: Publications are the currency of the academic profession.

Jim:  To achieve immortality.

Peter:  Provide baseline data against future change.

What's hard about it?  What do we need to do in this class to help you?

Dan: Everything.  Getting it all together in timely fashion.  Why does it take so long?

Jim:  It's psychologically daunting because of the immortality issue.

Bali:  Choosing what to focus on.  The Discussion is the hardest part.

Rakesh:  Avoid repeating mistakes in interpretation of previous work.

Megan: Writing technically requires focus.  I want to be more concise.

Kikang:  Choosing the journal.  Statistical considerations.

Laura:  Choosing the figures and tables, there can't be too many.

Nick:  What makes my work stand out in relation to previous work.

Angela:  Organizing and analyzing the data is challenging.  Communicating to people not familiar with the project.

Anna:  Justification for the project: who cares, to whom does it matter.

Carrie Rose: Defining the question answered by my data.

Peter: A paper is more detailed than a poster or oral presentation.  How to do it all concisely.

Artem: What if I get it wrong?  Data organization, analysis, and interpretation.

Suzanne: It's easy to get lost in the details.  The audience needs the big picture to get the importance.

Bhavin:  Writing for the selected journal; they have different requirements and different audiences.

Friday, Jan 22:  Getting Started Exercise

Jim:  How to be concise.

Ruth: think of an abstract.

Bali:  It got me thinking about my masters thesis, as it was a long time since I did it.  Focus on the main idea.

Kikang, Megan:  The light bulb: I need to change my question.

Laura:  It was hard to stick with one sentence.  There are a lot of aspects.

Nick:  Being concise.  How to elevate the conclusions beyond results.

Angela:  Results and conclusions.  Maybe the original question needs to change.

Anna:  My results don't exactly reflect my original questions.

Carrie Rose:  I found I had to adjust all the sections at once.

Peter:  Should I make the problem more general than emerald ash borer.

Artem:  My question isn't answered by my results. 

Suzanne:  I started too specific and worked towards getting more general.  I'm trying to put two groups into the same paper.

Rakesh:  I think I need a question about methods.

Dan:  I think I need more detail, not less.

We noted that in the future, copy provided for review should be DOUBLE-SPACED.

We decided that the revised Getting Started Exercise, which you will provide with your Tables and Figures and again with your Methods and Results, will be in the format required by your journal for the abstract.  This will still serve to orient your reviewers (classmates) to your paper, and it will help you develop your abstract.  The abstract does not need to be double-spaced, as we are not planning to give you feedback on it now.

We added a note that your outline, due on Feb 5, should include the full statement of your objectives (probably one paragraph).

Wednesday, Jan 27: Choosing your Journal

Factors to consider

Relevance:  You want the right people to read it and cite it.  You want it to be accepted.  The title is a clue.  You can read the statement in the Instructions to Authors or on the journal's web site.  Look through the papers you have been reading.  Look at the journal table of contents.  Look at where your co-authors are publishing.  Colloids and Surfaces has two journals with different content; choose the right one.

Prestige: The impact factor describes the average number of times each paper was cited.  This depends on the time period assessed and it differs by fields - there is a reading on this topic on my web site.

ISI Journal Citation Reports ranks journals.  The J of Wildlife Management was ranked 45 out of 125 zoology journals, and 75 out of 124 in ecology.  There are websites that present this information for specific fields.

Journal of Forestry has a lower impact factor than Forest Ecology and Management, but it's the most widely read forestry journal in the world.  So you might have more impact publishing there, even though the impact factor is not as high.  (Readers of the most applied journals are not writing research articles.)

Experts in the field (like your co-authors) will have opinions about the prestige of the journal.

Standards of the Journal:  How strong is your research relative to the expectations of the journal?  Some journals publish different types of papers.  You want to make sure that your paper is a good fit to the journal.  How well supported are your conclusions by the data?  How novel are the results?

Acceptance Rate:  Some journals report this in the Instructions to Authors.  Kikang found a library that provided acceptance rates for all the journals, but she didn't have access to it.  You can google the journal name and acceptance rate. You can email the editor.  The relationship between acceptance rate and impact factor is not as strong as you might think-the best papers don't get submitted to weak journals, and weak papers don't get submitted to the best journals.

Audience:  For example, some journals are read by managers, others are read by researchers. 

Time to publication: Some journals schedule certain topics for certain issues (Biomass and Bioenergy, Journal of Forestry).  This could result in a delay of up to a year, depending on when your paper gets through the gate.  Some have different timelines for different types of articles.  You can collect information by looking at the dates published in articles from that journal.  If there is a long delay to publication, find out whether your paper will be available electronically before it's in print.

Coauthor preference:  Your co-authors are probably in a good position to judge the factors above.  If you disagree, you should have a conversation. 

Cost to the author: Many journals have page charges (journals that are published by professional societies defray the costs of subscription to their members by charging authors). For example, the Journal of Wildlife Management charges $150/page for non-members, $90/page for members for the first 8 pages.  About 4 pages double-spaced makes 1 journal page.  Journal of Forest Research: about $300 per pages over 8 pages.  Oecologia charges 100 euros for pages over 10.  Who pays for this?  Your research grant.  At ESF, the Research Office can help with page charges if your grant can't cover it.

Color plates: $650 J. Wildlife Management.  I think CJFR was $1000.  J. Forestry is free!  If the electronic version will have color for free, maybe you don't care about the print version.  I don't pay for it.

Ease of access:  Articles published in open-access journals are cited more frequently than those that cost money to get.  Among subscription journals, some are more widely held than others.  Some journals allow the author to pay to make the article free for anyone to get.

Language:  Bali wants his research to be accessible to Spanish-speaking people.

The journals you chose

Bhavin:  Journal of Membrane Science:  Advisor advice, and this is the journal he uses the most.  Backup plan:  Separation and Purification Technology.

Suzanne:  Conservation Biology: Advisor agreed.  They have a new social science component.  Backup: J of Forestry.

Artem:  Ecological Applications (reviewed by the same editors as Ecology and Ecological Monographs).  It's a good journal, advisor advice.  Backup plan:  Environmental Management.

Carrie Rose:  J of Wildlife Management, coauthor advice, but the acceptance rate is 30%.  Check on acceptance rates of the other journals.  It wasn't originally intended for publication, it was an undergraduate thesis.

Anna:  Advisor prefers Oecologia, an ecology journal, but Anna is thinking of the J. of Wildlife Management. 

Nick:  J. of Forestry, because previous work was published here.  Second choice: Forest Ecology and Management.

Angela:  J. of Herpetology, advisor agrees.  Backup: Chelonian Conservation.

Megan:  J. of Wildlife Management, advisor agrees.

Laura:  Biomass and Bioenergy.  It has a good turnaround time, it fits well, and coauthors agree.  Bioresource Technology is a new journal that might also fit.

Kikang: Ecosystems publishes articles like hers. 

Rakesh:  Separation and Purification Technology.

Bali:  J. Tropical Ecology.  The Caribbean J of Science might have a better audience.

Jim:  River Research and Applications, appropriate to topic area and quality of the work. 

Dan:  A paper rejected by Global Change Biology could be revised and resubmitted to J. Geophysical Research.  The butterfly paper could go to J. Applied Ecology, or J. of Insect Conservation. 

Friday, Jan 29:  Figures and Tables

Bring 10 copies of your figures and tables, along with a single-spaced version of your revised abstract.  For at least one figure or table, provide an alternate format for us to consider.  Print them in the size you expect the journal to print them.  Provide the captions under the figures, as for a journal article.  This is not how you will submit them to the journal.

Lessons learned:

Don't ask you readers to learn new vocabulary (e.g. acronyms)

Scrutinize defaults from Excel, e.g. gridlines, box around the figure, title at the top, max and min axis scale, tick intervals, etc.

Alphabetical is not a meaningful organization of information.  Organize your table and figure entries according to some variable, the most important one.  For species, you can use a taxonomic ordering.

How do you decide how many digits to report?  You can look at the measurement error.  You can look at the uncertainty around the mean.

For P values, give to 0.01 or 1 digit if smaller.

When is a table better than a graph? Tables are better if your readers need the exact numbers.  Figures are better for seeing patterns or trends over time. 

It's hard to show sample sizes on a graph.  They could be in some other table (site description).  You could put them in the caption. Put a space between numbers and units.

Choose meaningful variation in symbols.  If you have two dimensions in your design, consider using fill pattern to describe one and shape for the other.

If you use open figures, you can see overlapping observations better.

Use consistent symbols and fill patterns across graphs.

Connect the dots rather than using curve-fitting (unless the point is to describe your observations with the curve, as in a regression analysis).  If it's just showing related observations in a time series, straight lines are best because they don't suggest that you know more than you do about what's going on between observations.

Capitalization

When is it important to show the scale relative to zero?

Choices for error bars:  s.e., s.d., 95%.  S.E. is most common, maybe because they are the smallest.  95% CI is probably more useful.

Excel is not designed for presentation graphics.  Suggestions:  SigmaPlot (not free), R is free.  MatLab is better than Excel.

Transformations:  Show the original units for the data, readers will understand them better.  You can make the scale linear or transformed, your preference.

We like legends better than text in the figure caption.  Note: Use a text box if you want to label your legend.

It's easier to compare numbers in columns than rows. 

When grouping bars, put the ones you want to be compared together.

Your legend should contain enough information to orient readers who aren't looking at your text.

Friday, Feb 5: Outline and Objectives

Kikang:  Doesn't have all her results, this will change.

Dan:  How much to put in an outline?

Jim:  I tend to try to write the final draft first.  It's hard to break that habit.

Peter:  Level of detail?

Artem:  Doesn't usually use an outline.  Like Jim, he wants to write sentences.

Bali:  He has a very general outline because it references his thesis.

Megan:  She made a paragraph outline based on her thesis.  Her journal has specific guidelines and so she used that in making her outline format.

Nick:  His journal uses section numbering.  He always writes from an outline, adding citations where he should reference them.

Angela:  This seemed to help get things organized, even though she's not used to using an outline.

Carrie Rose:  I already have sections written, I started slotting them into the template for the paper. 

Anna:  Doesn't usually use an outline, tried to describe each paragraph.  She had a new insight about how to frame her paper, especially the implications.

Suzanne:  A mix of putting in sections already written and indicating topics where they aren't written.

Rakesh:  The Introduction is the hardest.  How to decide the order of presentation of the results. 

Laura:  Outlines are something I use only when I'm struggling. 

Bhavin:  He did a google search and looked at different outlines.  He liked the one at the physics department at SU.  Needs to better organize his results.

These topics came up for discussion.

Methods: How much detail?  Three rules: An independent scientist should be able to replicate your result.  Don't include methods not relevant to your results.  Provide enough information for your readers to understand your results.  The level of detail you give will depend on your audience.

Results:  Some degree of interpretation can belong in your Results, some goes in your Discussion.  Tell us what your results mean, don't make us wait for the Discussion.  Rather than writing great descriptions of your Results in your Discussion (many of you will do this), try to write them now (for Wednesday).

Objectives:  You can add hypotheses (predictions, expectations).  Alternate hypotheses are very engaging. 

Objective 3:  To assess how is fine for a proposal but not for a paper.

Use sentences.  Use multiple short sentences rather than long sentences with numbered subsentences.

Intro vs. Discussion: some types of material (reference to other studies, introducing concepts) could go in either the Intro or Disc.  We'll be studying those sections later in the semester.

Weds Feb 10: Results

Kikang:  How to divide discussion and results

Dan:  This could be a never-ending process.  How do we know when to stop.

Ruth:  Good question.  Sometimes you just set a deadline and let the reviewers tell you whether what you did was enough.

Jim:  This was a 100-page report; what belongs in the paper?  Working with a co-author would help.

Ruth:  So those of you lucky enough to have co-authors in Syracuse should take advantage of them!

Laura:  I keep thinking of more things to do.  It can take time to even see if they are good ideas or not.

Bali:  I'm doubting my statistical analyses.  Maybe I should do another one.  This one was recommended by my advisor.

Ruth:  If you're not sure you're comfortable with it, you could get another opinion.

Nick:  Results vs. Discussion.  Should my statistics go in a table or in the text? 

Answers:  How many are there?  Some journals publish more ANOVA tables than others.

Angela:  Summarizing data for presentation can be different than the statistical analyses.

Megan:  How to rise above the minutia.

Anna:  Order of presentation? 

Answers:  Start with the most general?  This could be different than starting with the most important.  Some have chronological structure.

Carrie Rose:  Writing captions for the figures is the hardest. 

Answers:  Maybe redundancy is good here.  Nick:  Start with the most description possible, edit after. 

Carrie Rose:  Describing statistical results is hard to organize.  (Are the tables enough?)

Artem:  How much detail to include?  It's deadly to read too much minutia.

Peter:  The Results have to be consistent with what belongs in the Discussion, which we don't have yet.  The outline helps.

Rakesh:  Results and discussion: he asked his co-author, who suggested more experiments!

Bhavin:  Same thing he asked for help and got suggestions for more experiments.

Remember:

You can have a different judgment from your advisor.

You will write other papers after this paper.

Advice on writing results

Orient the reader at the start of this section.  Carrie Rose, Megan, Bhavin

Some orientation to methods can be useful: Angela, Dan

Use past tense for describing results:  Bhavin, Anna

Direction of effects is more important than significance.

When should you spell out numbers?  There are many rules:  If less than 10.  If it starts a sentence.  If it's followed by a unit, use a numeral.

Table 1 contains is usually a waste of space.  Describe what you want us to see and cite the table parenthetically.

Avoid the solidus (/) except to indicate division.

Avoid introducing unnecessary acronyms.

Avoid respectively  Nick

Friday, Nov 12: Methods

Rakesh needs to say almost the same thing in more than one place. 

It's not a bad thing to use the same sentence structure to mean the same thing.

Artem, Suzanne, Anna:  Is there enough information or is this too long?

Carrie Rose spent a long time on phrasing, combining sentences rather than having too many short, choppy ones.

Anna:  What to include, especially for the site description.

Angela:  How to keep it short.  I'm using previous data that weren't published, so I have to describe them all.

Megan:  Justification of the approach.

Justification for your approach might belong in your Introduction, if it's important.

Nick:  How to refer to the earlier study.

Laura:  Site description, and length.

Jim: What about the rationale for decisions that were made along the way.  We used an uncommon statistical procedure in an uncommon statistical package. 

Bali:  The methods section was well done in my thesis in the first place, because we worked it through before I started the work.  Mainly I need to eliminate detail.

Kikang:  How much detail? 

You are an important member of your reading audience; you may be the person who most refers to this document in the future.  Imagine yourself reading this paper after you've forgotten the details, if it helps you decide what to include.

Data not shown

Use paragraph breaks to keep one topic per paragraph.

was performed, is a waste compared to active verbs like, weighed, measured, analyzed

We'll talk about the active voice, with examples, next time.

Feb 17 Writing Exercise

3 voted for Brown:  Didn't start sentences with - But or However or use the first person.  Suzanne liked the sentence structure.

12 voted for Smith: Shorter sentences are easier to read.  They both started sentences with However!  Paragraphs breaks are good.  Parenthetical material must not be important.  Smith used fewer words. 

We identified subject and predicate in Smith and Brown (Smith's were better, more to the point) and in our own writing.  We looked at adjectival nouns.  We looked at some writing samples and worked on improving them.

Feb 19 Peer Review

Review Criteria

Is the title appropriate for the article?  --Later

Are the keywords, abstract, and summary informative?  --Later

Is this work novel?  --Introduction and Discussion

Is it relevant to a broad audience?  --Introduction and Discussion

Are the methods described clearly enough so that the work could be repeated?

Is the paper accurate in terms of methods, procedure, replications, and statistics?  Are significance statements justified?

Are the conclusions supported by the data?

Are there other issues of scientific quality?

Is a tightly reasoned argument evident throughout the paper?  Or does it wander?

Do the paragraphs flow smoothly, is the manuscript readable?

Please point out ambiguities but do not substitute your style for the author's.

Length relative to content: include only what's relevant and necessary.

Is there redundancy across figures, tables, and text?  Are all the figures and tables necessary?  Are they readily interpretable?  Are the captions concise and do they describe the figures?

Are the references up to date and adequate?  Can you suggest additional references?  Are all the references cited listed in the reference list and vice versa?

Does the paper conform to journal requirements for units and other formatting issues, including Latin names, soil names, chemical names, etc.?

Reviewer Assignments

Anna and Megan review each other's papers.

Review the paper by the person following your name:

Nick Rakesh Bali Bhavin Jim Angela Artem Carrie Rose Dan Kikang Laura Peter Nick

Advice on Writing Reviews

The review will be read by the author as well as by the Associate Editor (in this case, me).  You need not write additional comments for the Editor (journal reviewers may do this).  We will not be recommending acceptance or rejection!

Criticize the science, not the scientist.

Begin by describing the paper.

What did you like about the manuscript?

Start with the general comments, big-picture issues, end with specific comments.

Make your suggestions specific, give examples.  Line references can be helpful.

Acknowledge that you may not be the intended audience for the paper.  Share your reactions.  Point out areas where you may not be an expert reviewer.

Tone is important. Use positive constructions.  Don't say, this is badly written, say the writing could be improved.

Humor is good.

Indicate your interest in the topic. 

Indicate strengths as well as weaknesses. 

Your review should be clear and concise. 

You can write on the manuscript and return it to the author, but these comments will not be seen by the Editor.  If a comment is important to the Editor's evaluation of the paper, it should be included in your written review.

If you include comments in your written review that are not important to the Editor's evaluation of the paper, you note them as such.  As you will soon learn, the authors will be asked to respond to the reviews, and we don't want to have to respond to every suggestion about comma placement.

Feb 24 Introductions

Select a paper that has an introduction of the type you plan to write (similar field, organization, maybe from your target journal).  Number the paragraphs so we can quickly review them together.  See if you can identify the purpose of each paragraph.  Bring 8 copies, and we'll share.

What's in an Introduction?

Problem: big picture

Broader context

Describe object of study

Background (historical)

Comparison to other studies

Techniques, especially if unusual

Problem to be solved in your study

Objectives

Bali:

  1. Broader context
  2. Problem statement
  3. Objective

Nick:

  1. Background, problem
  2. Object of study, sugar maple
  3. Background, defoliation stress
  4. Background, decline
  5. Background, damage
  6. Background, climate factors
  7. Background, stand dynamics
  8. More stand dynamics
  9. Background, soil chemistry
  10.  Objectives

This paper was written for practitioners, not scientific peers (J of Forestry).  So the long background might be very suitable for this audience.

Anna:

  1. Background, problem
  2. Techniques used in the past
  3. Objectives: demonstrate a better technique
  4. Acknowledgments

The Acknowledgments used to be in the introduction of papers in the J Wildl Manage., they don't do this any more.  (Thanks, Megan)

Megan:

  1. Problem, limitations of previous studies
  2. This study, need for specific design
  3. Background, continuing objectives
  4. Predictions, importance.

She would reorganize this introduction, the paragraphs are too long and the topics are mixed together.

Carrie Rose:

  1. Problem statement, deer
  2. Additional problem, plants
  3. Lack of information (problem to be solved in this paper)
  4. Background on the specific problem.  Objectives.

It's probably not ideal to mix references to previous work in the same paragraph as your objectives.

Artem:

  1. Broader context, successional change, seed rain
  2. Background: environmental conditions
  3. Background: conditions over time
  4. Background: theory involving above factors
  5. Objectives.  Includes references to previous work in introducing hypotheses.

The introduction is long, it was published in Ecological Monographs.  There are references in with the objectives, but we thought it was okay.

Kikang:

  1. Background, history, definitions
  2. Problem with methods
  3. Advantage of this method.  Objectives.

There is no introduction to ponderosa pine, nor to stand age, which are the contexts for the study.  They wrote other papers focusing on these aspects.  The title suggests that these topics are important and should have been introduced.

Laura:

  1. Background: water availability
  2. Background: wet and dry years
  3. Background: species.  Gap in knowledge.
  4. Background: water use physiology
  5. Background: variation in individuals.  Gap in knowledge.
  6. Background: isotope techniques
  7. Background: isotopes in tree rings
  8. Objectives

Looks like all the right topics, we might recommend a different sequence (3 after 4).

Signal words:  Something may explain,  Something is poorly understood.  Something could arise from something else.  None of these studies examined.

Use these types of statement only if you're going to deliver to fill the need!

Jim:

  1. Problem: water quality, Thailand, monitoring
  2. Method of biotic indices (this is well accepted and could have been omitted), problem of local calibration
  3. This study tests a new method, locally calibrated.  (Objectives)
  4. Introduces BMWP (unnecessary detail)

5-12.  Previous attempts to modify for local applications

  1. His previous work.

This introduction is too long and not focused on the current study.

Remember, only your committee (who reads your thesis) cares whether you know everything that has gone before.  Your journal audience doesn't want to read a comprehensive lit review.

Hydrobiologia publishes weak papers and Jim intends to do better.

Angela:

  1. Problem, threatened species, road mortality and crab traps
  2. Background, site-specific, road mortality
  3. Objectives
  4. Never mind the crab traps
  5. Secondary objectives

Maybe the dismissal of the crab traps could have gone in the Discussion.

Do we like the primary and secondary objectives?  It seems like it's asking us to learn a classification that might not be important.

Bhavin:

  1. Background, problem statement (gap in knowledge)
  2. Objectives, what is covered and not covered.

There are no references in the Introduction!  There is no global context, why is this important. 

Oh, this is a review paper. 

Peter:

  1. Background, species
  2. Background on study, site, problem
  3. Other studies, more background
  4. Objective

Is there a best number of paragraphs?  It might not have to be more than 3 or 4.  Some of the papers that had only 3 or 4 paragraphs were lacking in paragraph breaks!  There is not a single right way to write your Introduction, but there are a lot of possible solutions depending on your specific needs. 

Feb 26 Statistical Considerations

Laura:  ANOVA, MANOVA or ANCOVA

Bhavin: Trend lines (polynomial fit)

Kiakang: ANOVA

Jim: regression, PCA

Peter: ANOVA

Artem: MANOVA, PCA, NMS?

Bali:  Mann-Whitney, NMS ordination

Nick: Regression, linear, maybe logistic, model selection AIC

Angela: regression, GLM and GAM

Megan:  logistic regression, model selection AIC

Carrie Rose:  ANOVA, paired t-test

Anna:  regression, AIC

Rakesh: graphs

Dan:  2-stage ANOVA, other

Mar 3 Reviews, Responses to Reviews

Kikang used line numbers to reference her comments.  Looking at someone else's paper can help me improve my paper. 

Dan:  It was interesting being on the other side of the process, and learning how to do it.  Nobody trains us to do this.

Jim:  The sheet of editorial notations was too constrained.  Maybe we don't need such a specialized system.  Some of the symbols are generally known.  Stacking them up in the margin seems hard.

Bhavin:  It was an opportunity to get an update or introduction to a field I was interested in.  And it gives me better ideas of how to organize my own paper.

Artem:  It was hard to decide how much to put in the specific comments area.  Some are clearly not important, but where to draw the line?

Megan:  My comments not requiring a response was the longest section.

Ruth:  You'll have a better feeling about this after you've written a response to reviews.  What's worthy of a response?

Nick:  It was hard to understand the paper without the Introduction.  I did some background reading.   (Wikipedia!)

Angela:  Not being familiar with the subject matter, the Introduction would have helped. 

Ruth:  We had only the Abstracts.  We should encourage people to include Intros and Discussions, if they have them, in the first round, even though we're not ready to review them. 

Megan:  It was helpful to be on the other side of the review process.

Carrie Rose:  I couldn't tell whether my suggestions were personal preference.  I read the paper several times and changed my mind about things. 

Ruth: The author doesn't have to take your advice.  You may have had a problem for a reason that the author will understand differently from you.

Carrie Rose:  I was glad I felt competent in the subject matter. 

Now when I read published papers, I see them more critically.

Anna:  Reviewing a similar study gave me ideas for things I want to do in my own paper, since this is a new field for me.

Pete:  It's easier to make specific suggestions than the big-picture suggestions.

Bali:  I didn't know much about the topic, but there are similarities in structure and opportunities to learn about writing.  I needed more information to understand the methods.

Laura:  First I made specific comments, it took rereading to get to the more general observations. 

Rakesh:  I wasn't familiar with the subject matter, so I wasn't sure whether my comments were appropriate.

Ruth:  In this class, we aren't always confident whether we represent the intended audience for these papers.

Response to Review

Bali:  The reviewers are the jurors, the Editor is the judge, and the author is the defendant.

Your goal is for the Editor to accept your paper quickly, without another round of review.  If it goes out to review, you will have a delay of several more months and another round of comments to address.  Make the editor's job easy.   Write your review so it can be evaluated without reference to your paper (either the old or the new).

Instructions:

This document is addressed to the Editor. 

Date it (actually, with electronic submission, they know the date).

Give the title of the paper and the number assigned by the journal (not relevant in this class).

Introductory remarks.  You might need to explain things that don't fit into the format of responding to reviewer comments.

Thank the editor for handling the paper.

The editor may have made requests to you, in addition to the reviews from the reviewers (not in this class).

Include the reviewer comments verbatim.  (Please email your reviews to your author.)  You could distinguish them with italics.  Color is good.  I'll send you an example and put it on my web site.  This one had black for the reviewers, blue for my narrative response, black for quoting my text, and green to show changes in my text. You are not required to use color.  The requirement is that the reviewer's comments are clear and your responses are clear.

You may need to clarify the reviewer's comment for the editor's benefit.  If the reviewer says,
51-52  This sounds important but I don't understand it.

You will need to include the original sentence so the editor doesn't have to go find it.  The sentence in question read or The sentence was and provide it.

Then say, The revised sentence is and provide it.

In some cases, the comment is no longer relevant because the material has been removed.  In this case, you probably don't need to copy the material to the Editor.

Tone:

Try to avoid saying, The reviewer is stupid.  There is probably a reason that the reviewer got it wrong, and there is usually something you can improve so that another reader doesn't have the same problem.

Don't sound defensive.  The editor doesn't want to hear why you did something the way you did, just fix it.

Writing rules

I usually pull - anonymous - examples from your papers to illustrate common problems with writing or presentation.  This time, I said I'd set up a document for you to provide your own examples.  Here are some of the categories.

Methods:  Too much?  Too little?

Use past tense for Results.

Use active verbs.  Use the subject and verb of your sentences to convey the main meaning.

Throw-away phrases or sentences: The results of this analysis are shown in Table 2.

Describe your results, citing statistics (don't focus on statistics)

Don't make the reader learn new acronyms.

Grammar and punctuation rules.  Have you heard of these?

Don't split infinitives

Don't end a sentence with a preposition

Dangling modifiers

Don't use a comma in a list of two items.

"which" introduces a non-restrictive clause; "that" is restrictive.

Use words for their expected grammatical roles.  Avoid adjectival nouns. 

Hyphenate compound adjectives (or avoid using them):

You can't start a sentence with a number or an abbreviation.

Don't use a solidus (/) except to indicate division.

Avoid the possessive apostrophe, which is rarely seen in technical writing.

Don't use pretentious words where a common word has the same meaning.  To me, these sound affected:

"comprise" for "compose"

"impact" for "effect"

Phrases to omit (you can find long lists of unnecessary phrases in books on writing)

"in order"

"Respectively" rarely improves a sentence.

How many digits?

  1. precision of the measurement
  2. what do you need to compare it to?
  3. uncertainty in a mean is indicated by the standard error
  4. converting units with a constant doesn't add precision
  5. You might want to report insignificant digits to make your calculations consistent.

Suggestions for improvement, using anonymous examples.

Mar 5 Discussions

What's in a Discussion?

Interpretation of Results, reasons for what you saw, alternative interpretations

Relate to theoretical framework, relate to your objectives,

Compare your results to those of other studies

Speculation

Limitations of your approaches

Implications for management or application.  What is the population of inference?

Suggestions for future work

Conclusions

Summary

Make sure you bring things back to the bigger picture.  Importance of your findings.

Some papers combine Results and Discussion

Examples you brought

Jim:

  1. It starts with excuses and limitations.  Best not to start (or end) with this topic.
  2. Relation to other work (but leave out the details that belong in the Results section)
  3. Relation to another study, too much detail.
  4. Justification for the approach.  Might be worth a sentence.

5, 6.  Effort required in this study and for future applications.

  1. Declare success.

What to do, what not to do.

  1. Recommendation for future research, maybe not the best sentence to end with.

Rakesh:

The Results and Discussion is all one paragraph!

It could have been two paragraphs, one about filtration and one about coagulation.

There is no comparison to other results, the specific objectives are novel.

Some of the differences in the style of publications across fields is cultural; disciplines have evolved in different ways and use journal articles differently.

Nick:

Subheadings are good!

  1. Comparison to other studies (soil and foliage)
  2. Comparison to other studies (foliar and soil depth)
  3. Soil properties: interpretation, suggestions for future work.
  4. Stand condition: interpretation
  5. Geology: interpretation, caution
  6. Caution explained
  7. Comparison to another study, disagreement, explained by scale.

8, 9, 10: Three more topics under one heading.

8 is about the topic, mostly using other studies.

  1. Relates back to objectives.
  2. Back out to the big picture.

These two paragraphs probably should have had a change in section headings.

Carrie Rose:

  1. Summarize results, relevance to bigger questions
  2. Relate to theoretical framework
  3. Narrows down to the current topic (good justification paragraph for an Introduction)
  4. Comparison to other (possibly irrelevant) studies
  5. Importance of the result
  6. Implications for management
  7. Speculation on implications
  8. Conclusion and future implications

Summary section has its own heading.

Anna:

  1. Interpretation of results
  2. Justification of study design (could go in Introduction)
  3. Justification of site selection (could go in Introduction)
  4. Limitations of this approach (sampling error)
  5. Background (could go in Introduction)
  6. Other similar studies
  7. Need for future work

Heading for Management Implications

  1. GIS
  2. Wetland inventories
  3. Requirements
  4. Limitations

Angela:

  1. Novelty, comparison to other studies
  2. Surprising result, interpretation
  3. Continues interpretation
  4. Comparison to other studies
  5. Recommendation for future research
  6. Back to the big picture, implications for management

Bali:

Three main headings, which had confusing language

  1. Results
  2. Comparison, interpretation
  3. Evaluation of hypothesis, why not supported
  4. Justifies importance of the topic

Topic heading narrows.

  1. 5, 6  Interpretation of results
  2. 7, 8  Third topic heading, address hypotheses.
  1. Conclusions, implications for management.

Megan:

  1. Review objectives and predictions, were they correct and why.  (Redundant with the Introduction.)
  2. Most interesting finding
  3. Continues with interpretation
  4. Switches topics, relates this aspect to other studies
  5. Complications, other considerations
  6. Limitations of this approach  (improve is a positive word)
  7. Value of this study
  8. Management Implications

Laura:

  1. Summary of results, interpretation
  2. Results, comparison to other studies, interpretations
  3. Results, comparison to other studies

4-8:  More detail, relates to other studies

9 explains a specific result, which is a new finding

10 interpretation and comparison to other studies

11 speculation as to cause

New topic heading

12 relates to other studies

13 declare success

14 last paragraph, some of everything, focus on most important finding.

Refer to your figures and tables in your Discussion if it will help orient your readers.

Dan:

5.1    Comparison to other studies, which differed in scale (6 paragraphs)

5.2    Sampling limitations (5 paragraphs)

5.3    More specific topics (11 paragraphs)

5.4    Implications and conclusions (3 paragraphs)

Artem:

  1. Explanation of results
  2. Comparison to other studies
  3. More explanation, with reference to theory
  4. Variation in results, comparison to other studies
  5. Soil moisture regimes, interpretation, reference to other studies
  6. Size distribution, ditto
  7. Stem ages, ditto.  The last sentence is the conclusion.

Kikang:

Summary of results, divided by old and young forests. 

Start with fine roots, because it was more important.  Results could have been reorganized to match the order desired in the Discussion.  Parallel is good.

Limitations and future research are at the end.

Conclusions section.

March 10 Introductions

Bhavin:  Does it need more literature review?  How does it fit with the results?

Dan:  I've gotten it down to 5 paragraphs, it was too long.  Does it still make sense?

Laura:  I found that I had repetition across my sentences.

Ruth: You can make a topic outline of your paragraphs.  Bali says 5 is good.

Nick:  I'm trying not to get into too much depth.  I'll go into more depth in the Discussion.

Jim:  I need help with an abrupt transition.

Artem:  Same problem with a transition between paragraphs.  I have a secondary objective without a hypothesis.

Laura:  How to sequence the topics.  Now I'm thinking about the funnel, it's harder to organize.

Anna:  How much background information is needed?  Is this too much?

Megan:  The funnel:  we want to convince the reader that this study is the best way to get at the big question.

Angela:  I left out the specifics; my hypotheses are broad.  I thought it would take too long to go into detail.

Carrie Rose:  I'm having trouble with sequencing, too.  I have the site justification at the end but it seems like it should go earlier.  The outline could help.

Peter:  How to decide how much detail is useful.  Invasive species in general vs. emerald ash borer.

Bali:  I have the broad context but I have trouble communicating my problem statement.

Rakesh:  Our papers have short introductions, mine is 5 paragraphs.

Carrie Rose:  What about style? 

March 12 Discussions

March 24 Mid-Semester Feedback

On the Wednesday when we get back from break, we will discuss how the course has gone so far and how to make the most of our remaining time.

If you write your answers to the following questions and print them out, then we can pass them around in class and your feedback will be read anonymously by another student.  (If you don't mention your paper topic, this can be really anonymous.)

  1.  Below is a list of the topics and activities in the course so far.  Please indicate which you found especially useful, which least useful.  Give specific suggestions for improvement if you have any.

Why Publish?  :  Reassuring to hear that others are struggling with similar things.  3 votes  Suggestion:  Why not to publish.  (There is such a paper!  I'll find it and share it with you.) This helped me focus more and narrow down my objectives.  Could combine with tips for choosing a journal.

Getting Started:  8 votes.  Most useful part of the course, gives shape to the paper.  We should revisit this exercise later in the class. 

Choosing your Journal:  6 votes  Thinking about who the audience is and the correct format saves revising.

Figures and Tables:  8 votes

We should have a class on how to use SigmaPlot and other graphical software.  Datagraph?  R?  Excel.  Powerpoint, photoshop, Image-J, Adobe illustrator.

Could we project one copy instead of printing so many copies?

It's nice to have the whole class focus on each set of figures and tables.  Useful, style and appropriateness.

Outline and Objectives:  7 votes.  Least useful 1.  It helped me focus my objectives.  Clarify the assignment.  Could this be combined with another exercise?  Give examples of outlines.  Explain the value to people who have text already.  There could be two different exercises for people who haven't started writing and those who have.  Objectives doesn't have to be in the same exercise.

Results (yours)  8 votes.  I got good feedback.  It's hard to give feedback on unfamiliar topics, the getting started exercise would help. 

Materials and Methods (yours)  10 votes. 

Writing exercise, in class  7 useful.  2 not.  Made us more conscious of word placement.  How to write more concisely and accurately.  Easy to get stuck on one sentence and lose track of time.  Or didn't get the point of the exercise. 

Preparation for peer review:  6 useful.  Examples are valuable, we wouldn't see these until after we publish. Learn how the process works.  No one example has all the information.  (There are examples on my web site.  There oughta be a web site to find the instructions to reviewers for the journal you're targeting.)

Introductions (examples):  5 useful. 2 not.  Bad examples are particularly instructive.  It takes a long time to go through all the examples; we had to rush.  Could be done in small groups.  But seeing a lot of examples is valuable.  You see that the good ones share characteristics; there are many ways to be bad.  If they were emailed in class, they could be projected instead of printing so many copies.

Statistical Considerations:  3 useful, 2 not.  It could have been useful, we need a better way to do it.  Maybe in groups of similar statistical approaches.  Most of us are listening to things that aren't useful most of the time.

With a smaller class, it could be more personal. 

Most people are done with their statistics at this point; should we have it earlier?

Students could provide more specific questions in advance of the class.

Maybe we could bring examples that are relevant to our type of analyses.  The types of analysis would define the small groups.

Responses to Reviews:  7 useful.  Now I know how to deal with a reviewer. 

Useful to read the other (Ruth's) review of the same paper I reviewed.  (Journals do this now)

Discussions (examples):  5 useful.  4 not.  Bad examples are important.  Discussions are longer and it was harder to evaluate them in the time available.  Each example could be critiqued before sharing it with the class.  For a large class, half the class could bring in Intros and half bring in Discussions.  If each person had to bring a good and bad example, we'd be evaluating them in advance.

Progress on Introductions (yours):  8 useful.  Thanks to the class for you input, it improved my work. 

It seemed like too much to do in the same week, but Intros and Discussions are interrelated. 

Progress on Discussions (yours):  9 useful.

Spreading them out would have allowed me to spend more time on the second one.

Sequence them with bringing examples, or just intersperse another session.

Mid-semester feedback for improvement:  This would go faster if the votes and comments were compiled electronically.

  1.  What future sessions would most help you prepare your manuscript for publication?  Are there other related topics you would like to address that aren't on the list?  Here are the proposed future topics:

Ethics:  2 useful,  1 not.  Maybe this should come earlier. 

Readings on Peer Review: 2 useful. 

Authorship:  3 useful.  Add a writing element.

Readings on Publication Productivity:  4 useful. 

Reviewer feedback and help session:  6 useful,

Proposals:  7 useful.

Abstracts (examples) 2 useful. 1 not.  We've all seen enough abstracts.

Work session:  1 useful.

Abstracts (yours):  5 useful.

Galley proofs, Steps to Publication: 2 useful.

Suggested topics:  Grammar and writing.

We decided to add an exercise with writing in the Authorship session (did I remember that right?)

  1. There are many ways we can work together.  Do you have comments or suggestions on the value of these alternatives?

For Figures and Tables, the whole class looked at each example.  For all the other parts, we used small groups.

Is it more useful to stick with the same papers than to switch every time? 

We want to learn to communicate clearly and effectively; we should be able to get the point or point out what's not clear if we don't.  New eyes are good.

It would be possible, in triplets, to rotate them so that there would be new and old readers each week.

Keep small groups constant for a few sessions and then change them.

The optimal answer could change over the course of the semester: seeing a wide range of issues is especially important early in the class (results and methods).

Pairs or triples, trading feedback in class

Some people said pairs were better; more people said triples were better

Writing formal reviews out of class

Informal peer assistance outside of class

Group discussion

Going around the table to equalize participation

Going around the class gets repetitive, we hear some issues repeatedly.  Sometimes unnecessary.  But good issues are raised.  And we get to know each other.

  1.  We have a very diverse group of papers, and we don't all know each other's fields.  Do you have any suggestions for improving our effectiveness in spite of this diversity?  How much diversity is desirable?

When we review papers, it's most useful to the author to have a reviewer who knows the field. 

Diversity is good and forces you to communicate clearly.

Technical comments can come from your committee (or from your outside reviewer).

Providing a background reading could help orient reviewers who don't know our field.

March 26 Ethics

Selection of Data (Megan, Laura)

Two students and a post-doc got data at a national lab; they had two points they didn't like.  Normally when this happens, we would repeat the experiment; they can't.  They could analyze the data with and without the outliers.  They don't have good evidence for the cause of the problem.  Sometimes it's more obvious that a measurement is wrong.  If you can describe the criteria you used to omit points from analysis, then you are not by-passing the peer review process.  These researchers were looking for a particular pattern.  We don't see it!

Conflicts of Interest (Bali, Nick)

A student's work is funded by a single private company.  She can't publish if her results are patentable.  Angela's data were coveted by the state, TNC, and ESF, and they had to work out an agreement.  Megan had a MOU memorandum of understanding between Ft Drum and ESF.  Links with industry can fund academics and provide graduates jobs in industry.  We acknowledge funding sources, and this can affect perceptions of the results.

We're lucky in ecology that our answers are site-dependent; someone else's study doesn't make ours unpublishable.  In conservation, we share values that override our personal gain.

Allocation of Credit (Angela, Jim)

A student's work was not acknowledged by a researcher who published results from his suggestions without acknowledging the source.  Could the faculty advisor play a role in his defense?  Can we afford to be generous?  How do we know who is trustworthy?  It's better to discuss the nature of a collaboration in advance.  If it's in email, you would have evidence of the communication.  Choose your collaborators carefully.

Authorship (Anna, Jen)

Graduate students want lead authorship on publications of their results.  Untenured professors also want lead authorship.  Increasingly, the specific contributions of authors are detailed in publication.  Increasingly, the impact of publications counts and not just the number.  I think it's more common for grad students to be listed before their major professors on their thesis publications.

Misconduct (Carrie Rose, Bhavin)

A student exaggerated the publication status of a paper in preparing a fellowship proposal.  If he had talked to anyone about it, he wouldn't be alone in his defense.

Misconduct (Rakesh, Artem)

A grad student suspects that another student, whose results she depends on, is fabricating data.  She thinks the advisor will side with the other student.  Could she approach someone on the committee with whom she has a better relationship.   It's a major accusation; should she approach the student first?

March 31  Peer Review

These people review the person after their name:

Laura - Jim - Anna - Carrie Rose - Nick - Peter - Dan - Artem - Bali - Laura

Angela and Megan review each other.

Rakesh and Bhavin review each other.

I'm happy to point out that 7 people got their first choice!  

Authors should bring to class on Wednesday two (2) copies of their papers and three (3) copies of the response to reviews.  Nobody is getting the same reviewer as before, so that simplifies the instructions.  Some reviewers may ask for electronic copy. If you want to save paper, you could send your documents electronically and ask who wants paper.  I want paper!  You can print on scrap paper, that's fine.

We had a good suggestion for educating our reviewers: authors should provide a background reading, maybe a review paper, that would help orient reviewers who are not familiar with the territory.  Next year I'll suggest this with the first round of reviews.  WIth this round, your Introductions should both lay out what a reader needs to know to understand the paper and also provide references to appropriate background reading.  But it wouldn't hurt to provide suggested readings--we'll talk about this briefly on Wednesday.

Effect on the quality of peer review of blinding reviewers and asking them to sign the reviews (Megan)

This was an experiment using an accepted paper and 8 introduced errors.  They used four groups of reviewers.  Full factorial:  Blind to the author or not, signed or unsigned, and a control group that didn't know it was an experiment.  Blinding was not completely effective (people can guess the author if they are expert in the field).  The quality of the reviews was similar, but reviewers who knew the identity of the author were more likely to reject.  Anonymous reviewers were insignificantly more likely to recommend rejection.  They conclude that changing the practices of blinding and signing would not make a big difference.

The effects of blinding on the quality of Peer Review. (Peter)

This experiment concealed author identity of 130 papers, which was 73% effective (27% guessed the authors - Megan's study was 26%!).  The editors rated the reviews as more knowledgeable when blinded.  Blinded reviewers gave lower ratings in several areas.

Does editorial peer review work? (Artem)

Review paper.  There was a study that took accepted papers and sent them out for review again with the authors concealed; most of them were rejected.  Another study looked at the changes that occurred during the review process and questioned whether the changes were improvements.  Other findings:  Only 10-15% of papers are useful.  Roughly half of all papers published are never cited!  (We hope this is getting better with electronic access.)  Reviewers most (90%) likely to produce a good review are under 40 years old, at top institutions, well known to the editor, and blinded to the author.  If none of these characteristics were present, 7% were good.  Suggestion:  Limit the number of papers reported in a CV, to focus on quality more than quantity.  (Is this getting better with impact factors and h-indices?)

The effect of blinding on the acceptance of research papers by peer review (Bali)

Reviewers blinded to the author were more likely to reject papers (this is the opposite of the paper Megan reviewed), even though their scores weren't different.

Peer review is a two-way process (Laura)

Proposal review:  There is a fear of bias in the review board, so grantees are not on the review board.  This doesn't get you the best experts in the field. 

Ruth: US agencies ask current recipients to server on review panels!  But they don't allow us to review or even observe the discussion of proposals with which we are in conflicts.  Conflicts are defined by advisor-advisee relationships, collaborations, and institutions.

Nepotism and sexism in peer-review (Carrie Rose)

197?:  44% of biomedical PhDs are women, but they hold only 7% of professorial positions.  Proposals rankings of 114 proposals were compared to the credentials of the authors:  women had the same impact factors as men but were ranked lower.  Women had to be much better than men to get the same ranking.  The same thing happens with affiliation with members of the board (units of impact points).  This was in Sweden, the leading country in the world for equality of men and women.

When peer review fails: problems with a giant laser project (Nick)

Livermore National Lab proposed a project that was supposed to cost $400 million, it became $4 billion over 10 years, and the date for completion was 4 years past due.  The reviewers were not disinterested; they were affiliated with the lab or the DOE.  Congress should stop funding and link it to progress.

Gender bias in the refereeing process (Anna)

24 journals in ecology and evolution were solicited, 7 participated.  Data on 2600 manuscripts: more men are publishing than women.  Senior researchers are male-dominated more than junior researchers.  Acceptance rates of papers by women and men were similar except for one editor.  The 17 editors who didn't respond might also be biased.  Non-English-speaking authors had a higher rejection rate. 

The philosophical basis of peer review and the suppression of innovation (Angela)

Reviewers think that the aim of peer review is quality control.  But they see a broader aim (in medicine, in this case, the goals of healing people), which might not be served by conservatism in quality control.  For example, trials with Li as a mental health treatment would never pass peer review now.  Solving big problems is disadvantageous to researchers because the funding dries up!  (polio)  As a reviewer, think about the overall implications and not just the quality control.  They gave examples of important work that was difficult to publish for this reason.

NSF is concerned about this bias against innovation.  Now they are asking for transformative research.

Drawbacks of peer review

This experiment showed that peer review of papers was not reproducible.

Writing an effective manuscript review

The choice of approach is a matter of opinion.  Ask the author to explain why an approach was used, and judge whether it yielded useful information.  (In other words, don't reject it just because it wasn't the approach you would have preferred.)

Focus on the positive, don't spend all your attention as a review on the criticisms!

April 7  Authorship

Point system: 

Angela:  It came out exactly as I expected.  It was helpful to go through the exercise.  It helped me evaluate whether I should add an author.

Laura:  My third and fourth authors came out not in the order I expected, and I added weight to the third based on source of funding.

Rakesh:  It came out in the order I expected.

Carrie Rose:  There was an order that the authors expected at the time of the thesis.  Now they have contributed more, but it would be awkward to change it.  I'm going to leave it the way it was because I don't think they care.

It helped me appreciate what people contributed.

Ruth:  When I initially list the authors, I say - in some order - to suggest that the order hasn't been decided yet.  You can't always predict the importance of people's contributions.

Carrie Rose:  It's easy to be biased and give more weight to recent contributions.

Ruth:  What if a co-author puts in a lot of work that isn't reflected in the final product?

Nick:  The Hunt system was better because it has finer categories.  I'm using data collected by other people, who need to be acknowledged as authors.

Angela:  Data were collected by contractors, they don't get acknowledged.

Ruth:  Paid help is less likely to be awarded authorship.  If you ask someone to help with your paper and you can't pay them, you can offer authorship.

Bali would have like to be an author of a paper he worked on as a technician.  If he had asked earlier, it could have worked out.

Ruth:  Identifying authors early on means that you can ask them for more help.

Dan:  My advisors have a very inclusive authorship practice, I have 11 authors.  Some of them probably didn't know they were authors until they saw this draft.  I'm changing the order now (they weren't supplied in any order).  Should they really all be authors?  Many contributed data that haven't been published before, and only some of those have participated in the paper.

Jim:  The agency that funded the project want to be authors.  Jim consulted with his supervisor, who confirmed that they didn't understand the project and weren't likely to contribute.  But it's politically expedient to include them.  They will probably be the last two authors.  The agency contributed data, but they are freely available on the internet.

Ruth: Chalk it up to cultural difference.  And hope they don't interfere. 
Authorship is culturally defined.  Even the order means different things in different cultures.

Is everyone the first author of their paper?  Yes.

Is anyone a sole author?  Bali is thinking about it.

Warning:  Last year's weakest paper was written by the only sole author in the class.

Does it matter how many co-authors you have?

Jim: It matters to the second author when you add a third author, if the citation format doesn't give names of more than two authors.

In some academic cultures, the last author is the head of the lab or the PI of the project.

In some cultures, the authors are listed alphabetical after the first.

It's more common now for journals to include information about author contributions.

The journal may require that all the authors certify their support of the paper.  Some want to know that all the authors can defend the paper.

What goes in Acknowledgments?

Funding agencies

Field technicians

Landowners, Agencies, staff of the research station.

Friendly reviewers, anonymous reviewers.

Committee members

Order of the acknowledgments:  How about chronological?  Funding sources are often last.

Who is in your Acknowledgments?

Artem: The person who collected the second inventory.  He was paid for it.  He had some creative input, too.

Anna:  Some people helped a lot and others helped only a little.

Angela:  You can say and all the other people who helped in the field.

Send a copy of your paper to your field crew, or they'll never know you acknowledged them.

April 9  Publication Productivity

Publish or Perish: Some reasons for not publishing

You're not done until you publish; the thesis is not enough.  Being too busy is not an excuse.  Yes, it's difficult.

The role of the advisor should be to strike the balance of letting the student make mistakes and learn from them but not mistakes that would prevent them from having publishable results. 
What can an institution do to prevent failure of advising?

Form 3B is intended to promote communication with the committee, but it's all about classes.  There should be more review of proposals, beyond just the major professor.  The candidacy exam is an early check for PhD students.

Publication Productivity among Scientists: A Critical Review

Success in science is measured by publication.  Why do some people publish a lot and others not? 

Individual characteristics:  strong motivation, autonomy, or self-direction.

Environmental location:  it matters where you got your degree, regardless of your current work environment.

Most scientists do not alter their ideas, approaches, and commitments after graduate school.

Productivity declines with age. 

Gender, Family Characteristics, and Publication Productivity among Scientists

A survey of 1200 US scientists across a lot of fields looked at factors that might relate to publication productivity. 

For women, 52% had no children; 21% of men had no children.

Productivity of women with pre-school children was high, then it's lower until they leave home.  Women do better in the second marriage.  It doesn't matter so much for men.  The productivity of men was less sensitive to family status.  Women do better if they're married to another professional or a non-academic scientist. 

Child Care, Research Collaboration and Gender Differences in Scientific Productivity

This was a survey of professors in the 4 Norwegian research universities. 

There was no difference between men and women in co-authorship.

Publication rate of men was not affected by having children.  It was low for women with children under 6.  By the time the kids are 10, there was no difference.

People in their 40's published the most.  Women under 40 published less, but afterwards they were as productive as men.

Career patterns of women and men in the sciences

This study surveyed NRC and NSF postdoctoral scholars. 

Men published 2.8 papers/year; women 2.3.  The women's papers were more comprehensive and got more citations on average than the men's: 24.4 vs. 14.4.  Write fewer papers and make sure they're good ones!

Women had more collaboration in graduate school but less collaboration after graduation, maybe because of discrimination.

Biology is the only field in which women were equal to men.

Women academic scientists in India

Women in their 50s have a high level of research activity.  Bhavin says all the women in his family went into medicine.  They don't go into engineering.  Rakesh says that engineering is more important in the south of India. 

Married women are more successful than single women, maybe because their male colleagues are more comfortable with them.

There are not enough women scientists: they say 10-15% would be enough.

The paradox of critical mass for women in science

This study interviewed people in 30 departments in 5 disciplines in the US

Women share the views of their age group in terms of work and family.  The pioneers came in under different conditions and don't relate well to the younger women.

A strong minority of 15% could start the ball rolling.

Motherhood and Scientific Productivity

In Norway, women are more productive after their kids are about 10.  Married and divorced people are more productive than single people.  Marriage is not as good for women as for men.  Unmarried childless women are less productive than men.

Cumulative Disadvantages in the Careers of Women Scientists

Women have a harder time finding mentors and don't have as good relationships with their advisors.  They more often move to accommodate spouses than men do. 

Angela's experience involved more female than male role models.

Are things improving?

Bhavin says chemical engineering is improving faster than mechanical engineering.

Nick learned more from female professors.

The post-liberal feminist perspective

The reward system reflects traits of males: Aggression, competitiveness, and dominance are rewarded.  Women are more collaborative and not as independent.  It won't change unless people are aware of these biases.

Equity and Equality in measuring faculty productivity

The system rewards characteristics typical of men.  We should be better at appreciating the differences in contributions by women.  They tend to teach more, fill more service roles, and be co-authors rather than lead authors.  They are more likely to write books.  Co-authorship should be valued more.  Equity is not equality. 

In physics, 44% of women are married to other professionals (80% for mathematicians).  Women have to work around having children.

Publication is the life blood of science.

The internet attracts a broader audience, and the public can be more involved in science.

April 16  Proposals

Abstract:  You may have a word limit or a space limit.  Read the instructions.  This document is important because more people will read it than read the rest of the proposal.

Objectives are really important, we don't have results, so this is what they're funding.  They seem more elaborate than what you see in papers.  You don't have to worry about matching your hypotheses to your results!  I like writing proposals even better than I like writing papers, because they aren't constrained by unfortunate outcomes.

It's important to show how the methods will allow you to meet the objectives.  You can show this with headings in the Methods section.  You can conclude with a section on data analysis and interpretation, explicitly linking the methods to the objectives.  It's disappointing to read a proposal where it's not clear that the proposed work is going to solve the problem.

The importance of the work may be highlighted in a section called Project Value, Justification, or Rationale and Significance.  This may be required by the proposal format.

How does a proposal end?  Products, outcomes.   Papers to be written, including lead authorship.

Timeline of tasks or a flow sheet.  Allocation of resources, including personnel.

You will rarely be held accountable for the specific products you proposed, but you need to be productive.  Anna's proposal doesn't even mention beavers, which she's working on from that funding.

Results of Prior Support:  NSF requires this.  Reviewers will comment on prior productivity.

RFPs (Request for Proposals)

These should describe priorities for funding.  How do you know where to send your proposal?  You can often find documentation of what's been funded in the past.  You can ask the program manager!

Technical requirements:  follow the rules. 

There are rarely rules for the format of in-text citation.  Make sure that your reference list is uniformly formatted.

Beware of criticizing authors you might get as reviewers.

Supplementary documents:  In addition to the project description, you may be required to submit:

CV for all the PIs.  There may be specific instructions for these.  Number of pages, font size, margins.  NSF asks for 5 papers relevant to the current proposal and 5 other. 

Conflict of Interest list:  previous collaborators, advisors, advisees.

Letters of support from people providing data, organizations providing sites, enthusiastic members of the stakeholder groups.

Documentation of cost-sharing.

Current and Pending Support: grants you are already committed to, and time committed. 

List of suggested reviewers is really important if they're having trouble finding people who aren't in conflict.

Facilities and equipment.

Ethical treatment and disposal of animal or human subjects.

Figures and tables cost space.  The page limits commonly include figures and tables.  Sometimes we squeeze them into small spaces.

Budget:  There may be a required form for the budget.  There may be a Budget Justification or Budget Narrative.  Including detailed calculations can help convince reviewers not only that you need the money but that you've thought through the details of your work. 

Your proposal will be reviewed by experts in the field.  NSF uses ad hoc reviewers as well as panelists.

How do you know about these RFPs?  Linda Galloway makes a monthly compilation of research opportunities.  Professional societies do it, too.  You can meet with her to learn how to set up automatic notification of opportunities in your area of interest, including those specific to graduate students.

April 21 Abstracts

Laura:  Three sentences for rationale.  Objectives (1 sentence).  Methods (2).  Results, Discussion, Results, Discussion.  You might want to mix things this way.

Dan:  1 sentence rationale.  1 for objectives.  1 for methods (pretty general, we learn the details in the Results).  4 for results.  2 interpretation.  1 conclusion relating back to the problem statement.

Yes, your most important results need to be in the Abstract, or people won't read the paper who ought to be interested in it.

Bhavin:  3 sentences for rationale.  Then the abstract describes the objectives of the paper, outlining the topics to be covered.  It does not describe the results.  It looks like a review paper, but it describes the results of this author's work.

Jim:  This abstract is not written for a general audience.  Water quality and invertebrates are not mentioned in the abstract.  There is no problem statement or background.  Like Bhavin's example, the abstract tells us what the paper will provide, rather than abstracting the contents of the paper.  It's not an experimental study.

Nick:  1 sentence for rationale.  1 for objectives.  1 for methods.  Many sentences for results.  No conclusion.  There is a keyword for acid deposition but nothing in the abstract. 

Angela:  No introduction or problem statement.  1 sentence for methods and objectives.  5 sentences for results.  4 sentences at the end about implications for management.  Some of this material could be moved to the front to make a problem statement.  We discussed the final sentence, which is inflammatory, but makes its point.

Anna:  First sentence is objectives.  1 on methods.  4 on results.  There is a paragraph break, followed by 2 sentences.  The last sentence is definitely a conclusion.  A good introduction might be to explain that colonization is hard to study but they had a situation with zero beaver following a fire. 

This is the first abstract with a paragraph break.  Check whether your journal allows it. 

Carrie Rose:  It starts with methods (2 sentences).  Results.  2 sentences of discussion.  The last sentence is a prediction but it's restricted to their study site.  We would like to see something relevant to a broader audience.  They need a problem statement:

White-tailed deer are at  record population densities.  There is a potentially interesting question about non-native species. 

Rakesh: Two paragraphs.  The first is background and objectives.  The second paragraph has methods, results, and conclusions.  Rakesh says more commonly, in a two-paragraph abstract, the second paragraph starts with the objectives. 

The conclusion seems weak: economics must be considered.  There's probably a way to state it that sounds more promising.

April 23 Last Chance

Laura:  Introduction: Why I chose my study sites.

Dan: figures and tables:  Some of the data in tables might be more effective in figures.

Bhavin:  Methods section is a mess.  Please review for consistency.

Jim:  There is going to be new material in the Discussion (not ready for today). 

Some results are needed for later methods.  Instead of IMRD, he may need I MR MR MR or even MRD MRD, and on to the next step.

Artem:  The laundry list results section:  Make sure the reader gets the point and isn't lost in detail.

Nick: Discussion: Cutting unnecessary stuff. 

Angela:  How to describe all the past studies, which are different and complicated.  She tried to put it in a table.  It's not published anywhere else.  Are the tables enough?  Does she need a narrative with it?

Megan:  She has a new section of her Discussion, regarding a surprising result.

Anna:  The Discussion should be rearranged.  Should the big picture come first or last?

Carrie Rose:  The graphs were in Excel and she wants to learn SigmaPlot.  The plots are only 10m on a side; is it misleading to report the results in hectares?  Species counts should be reported as counts.  Stem density should be per hectare.

Bali brought his abstract.

Rakesh:  Graphs could be better formatted. 

Bhavin has so many symbols you can't read them.  How about a line?

April 28 Abstracts and Titles

Rakesh: Flocculation dynamics of hot-water extracts of sugar maple wood

Pete:  Abundance of ash species in riparian forests in New York: implications for invasion by emerald ash borer

Bali: Divergent patterns of forest recovery on abandoned farmlands of the Anasco River watershed of western Puerto Rico - succession

Anna:  Landscape factors affecting long-term site occupancy by beaver in the Adirondacks

Megan:  Broad-scale factors affecting ruffed grouse population decline in New York State.

Carrie Rose:  Impacts of white-tailed deer herbivory on forest regeneration patterns in an exurban landscape

Angela:  Long-term assessment of bog turtle populations in Massachusetts: investigating effects of habitat changes

Nick:  Soil chemistry predisposes sugar maple to decline following defoliation [by forest tent caterpillar]  this one is hard

Laura:  Water stress and biomass production in short-rotation shrub willows in three contrasting sites (location?)

Effect of site conditions and water status on biomass production by different varieties

Artem:  Changes in tree seedling composition within power-line corridors are consistent with climate change in New York State

Bhavin:  Ultrafiltration of wood extracts: a study of the nature of clogging in membranes with different molecular weight cutoffs.

clogging during Ultrafiltration of wood extracts using membranes with different molecular weight cutoffs. (do we care about the feedstock)

Dan:  Woody debris dynamics along an elevational gradient of mature tropical forest in the Peruvian Andes  (respiration?)