Digital Systems RESEARCH PRESENTATION Digital Systems Research Presentation

Digital Systems RESEARCH PRESENTATION Digital Systems Research Presentation Mohammad Sharifkhani © 1

Lecture 3: Writing an academic manuscript Digital Systems Research Presentation Mohammad Sharifkhani © 2

Outline � General Considerations �Title, author list, etc. � Abstract � Introduction � Body and Result Presentation � Conclusion and Discussion � Response to reviewers Digital Systems Research Presentation Mohammad Sharifkhani © 3

HOW TO WRITE A GOOD PAPER � Full Unequivocal Data Set � Focus �A good paper can not be written 3 months � Careful Planning � Organization � Attention to Detail � Communication with Colleagues

HOW TO WRITE A GOOD PAPER FALLACY: It’s only the number of papers that count FACT: One good detailed paper in a prestigious journal is worth far more than smaller component papers in lesser journals!

HOW TO WRITE A GOOD PAPER A) B) C) A) B) AUTHORSHIP Before Starting to Write, Define the Following: Who will be coauthors Who only will be acknowledged Order of authors POTENTIAL PROBLEMS Senior Authorship/Independence Co-contributing 1 st and 2 nd Authors

HOW TO WRITE A GOOD PAPER � Choose the journal before you start writing � Carefully review several recent issues � Study the “Instructions to Authors” � Identify publication costs

HOW TO WRITE A GOOD PAPER � � � TITLE PAGE Choice of Title Short Description of major conclusion Running Title Not longer than SPECS Include words from title No unusual abbreviations Authors’ Names Check instructions Footnotes/Symbols Author for correspondence Key words

HOW TO WRITE A GOOD PAPER ABBREVIATIONS � Check instructions � Standard abbreviations � Frequent cumbersome phrases/words � Abbreviation list vs in text � Generally don’t begin a sentence with an abbreviation

HOW TO WRITE A GOOD PAPER ABSTRACT/SUMMARY � “Is the entire paper worth reading? ” � A brief account of paper’s relevant points � Length and format specified for each journal (i. e. continuous vs. specific sections) � It must stand alone without reference to text body

How to write a good paper � You may include the following in the abstract: �Features �Verification methods �Facts (numbers, improvement factors, etc. ) Digital Systems Research Presentation Mohammad Sharifkhani © 11

HOW TO WRITE A GOOD PAPER � Outline Key Aspects of Paper! � Introduction � Method � Results � Conclusion � � � Review with Senior Author and Key Collaborators Write First Draft Review with Senior Author/Key Authors Review Subsequent Drafts Penultimate Draft Sent to all Co-Authors

HOW TO WRITE A GOOD PAPER INTRODUCTION � Long enough to place paper in context � Reasons for doing study � Methods used � Indication of conclusions � DON’T MAKE INTRO TOO LONG!! � Don’t provide a complete history of field � Basically: What you did, why you did it, what you found!

How to write a good paper � Intro (example) �Preamble (problem statement general) �Method 1 (brief, problems, solutions) �Alt. Method �Most recent methods �Brief presentations, etc. �Paper construction Digital Systems Research Presentation Mohammad Sharifkhani © 14

How to write a good paper � � MATERIALS AND METHODS General/conceptual explanation of the art Step by step, hierarchy, Definitions (variables, acrons…) Every paragraph needs to have a message! �Introductory sentence �Short sentences �Conclusive sentences Strong sentences, companied by numbers etc. � Reference � Digital Systems Research Presentation Mohammad Sharifkhani © 15

HOW TO WRITE A GOOD PAPER MATERIALS AND METHODS � Provide sufficient detail for others to repeat and/or understand differences from past/future studies � Use subtitles � Patient populations: Tables, � Inclusion/exclusion criteria � Use generic vs. trade names � If unusual complexities, consider appendix � Make sure statistics are appropriate

How to write a good paper � Everything leads to something (meaningful) � Simulation � Analysis �… �Consistent and clear notations �Graph: X/Y unit, Z, condition, �Clearness. EPS waveforms, etc… ○ Always use Vector based graphics (Visio, Meta file, etc. ) ○ Never use paint! Digital Systems Research Presentation Mohammad Sharifkhani © 17

HOW TO WRITE A GOOD PAPER FIGURES/LEGENDS � Focus on key points � Don’t clutter Illustrations � Define symbols/abbreviations in legends � Label axes � Use same units as in text � Photo authenticity (in view of recent scandal) � Obtain permission for use of previously published material

How to write a good paper � Comparison (Apple to apple comparison) � Recent works � Conv…. � Comparative condition � Power, speed, area, reliability (offset, etc. ) � Use graphs � Effect of scaling � � Layout, gate/trans count, Measurement … state all conditions � Test setup � Verify the consistency � If not, try to explain Digital Systems Research Presentation Mohammad Sharifkhani © 19

How to write a good paper RESULTS � Logical sequence of presentation � Minimize “unpublished results” � Use subheadings � Use appropriate tables/illustrations � Don’t repeat data in text and tables/ills � Number of significant places after decimal � Patent issues

HOW TO WRITE A GOOD PAPER � � � � DISCUSSION “Reviewers rarely complain a discussion is too brief” Provide summary of results and how results led to conclusion Place results in context of current knowledge Impact of data (present and future) Unanswered questions Limitations of study Use subheadings Don’t overemphasize your own work

HOW TO WRITE A GOOD PAPER ACKNOWLEDGEMENTS ● Those not listed as authors who helped with study -Intellectual input -Encouragement -Special measurements/reagents -Techs in lab -Secretarial assistance ● Research support ● CONFLICTS OF INTEREST

HOW TO WRITE A GOOD PAPER REFERENCES � CHECK WITH “INSTRUCTIONS TO AUTHORS” � Numerous variations � IEEE/ACM Standard is well known Full names vs. initials Number of names before et al Page inclusitivity Punctuation Order in text vs. alphabetical No unpublished data or personal communications in this section � Limit “historical” references � Keep number of references as short as feasible � � �

HOW TO WRITE A GOOD PAPER ● ● ● SUBMISSION Address letter to Editor Describe paper in 3 -4 sentences Why it is appropriate to this journal If allowable, suggest reviewers Define any competitors who you feel would give unfair review Signatures of all coauthors on statement of authenticity/responsibility

HOW TO WRITE A GOOD PAPER ● ● ● ● RESPONSE LETTER WITH REVISION Segregate responses to Editor and each reviewer Repeat responses to each reviewer even if already responded to another Respond to each comment (consider numbering) Start with polite intro to each reviewer: “We are grateful to reviewer A for useful comments. . ” Consider placing Reviewer’s comment in letter immediately above your response Make sure pagination of initial and revision are synchronous and differences stated Don’t try to “muscle” the Reviewer! See an example by Nitin Mohan

HOW TO WRITE A GOOD PAPER � � � � COMMON MISTAKES Failure to follow instructions Too many fonts on the title page Inconsistent formatting in the body of the manuscript Errors in the punctuation of references in the List of References Failure to indicate the variability and/or reproducibility of results Results that are given to a degree of accuracy that far exceeds the accuracy of measurements Graphs and histograms without indications of standard deviations From Korner, AM. Guide to Publishing a Scientific Paper

HOW TO WRITE A GOOD PAPER COMMON GRAMMAR/STYLE ISSUES � Spelling: British vs American � Spell-check: false security � Active vs passive voice � Multiple adjectives � Incorrect use of “due to” � Ambiguous use of “this” � Hyphenation � Abbreviations � Run-on, sentences, use of semicolons � ALWAYS save your results in the original simulation tool platform/figures/ etc. while you prepare the paper. You need that for revisions and/or modification of the paper. � Punctuation IS important. � Use spaces properly: space after comma, fullstop, etc. � A paragraph never starts with Also, However, But, etc.

Technical Documentation Ø Ø Ø Organization (Abstract, Introduction, Background and History, Chapters, Conclusions, Reference List, Attachments) Logical Ordering, Exactness, Clear Problem Definition and Its Solution Challenges, Trends Format (Consistency, Hierarchy, …) Quote All References In Place Include Necessary References (Downloads in CD, Copies, Call No, …)

Avoid Plagiarism � Copying someone else’s work/words is absolutely unacceptable. � Make sure you maintain your scientific integrity in all of your submissions � Your track record and credibility is more important than your paper count! Digital Systems Research Presentation Mohammad Sharifkhani © 29

Digital Systems Research Presentation Thank you! Digital Systems Research Presentation Mohammad Sharifkhani © 30