WRITING A PAPER OR THESIS

For help in writing see web site

Publications

  • Are important for your education
  • Benefit the research group as a whole
  • Build your resume

Publications have an order of value that is summarized in Table 1. When given a choice, always try for a refereed journal article in a Science Citation Index journal. The Science Citation Index is in the Science and Technology Library.

Table 1. Types of Publications and Priority Order of Importance.

Priority / Publication Type / Comments
1 / Books / Not practical for graduate students
2 / Refereed Journal Publications listed in Science Citation Index / Best
3 / Refereed Journal Publication not listed in Science Citation Index
4 / Proceedings or Preprints from presentations at Professional Society Meetings / AFS, AICHE; these are usually easy to prepare
4 / Poster papers at Professional Society Meetings / AFS, AICHE
4 / Non-refereed Journal Publications and Pedagogical publications
5 / Invited presentations
6 / Non-published presentations at Professional Societies
6 / Government Reports on funded research projects.
1Writing a paper

You should write in a style that makes it easy for the reader to understand. Keep your writing short and concise. Use examples of other papers to show you how.

Organize your thoughts in a logical format. For most papers you can organize your paper in the format of:

  • Abstract
  • Introduction (including objective of the work)
  • Literature Review
  • Theory
  • Experimental
  • Results and Discussion
  • Conclusions
  • Acknowledgement
  • Notation
  • Reference

In the abstract you should answer 5 questions. Each question can be answered in one sentence or one paragraph. Usually abstracts are not more than one page. The questions are:

  • What is the purpose or hypothesis of your work? (What are you doing?)
  • Why did you do this work? (What is the economic or technological significance?)
  • How did you do this work? (What theoretical or experimental approach did you use?)
  • What were your results?
  • What is unique about this work?

You should have a particular message in mind that you want to tell your reader. Build your paper around this message.

A engineering/scientific work needs a clear objective. The objective is usually stated at the beginning of the paper, sometimes in the introduction. The end of the paper should state whether the objective of the work was achieved.

Always acknowledge the organizations that fund the research.

Select an appropriate journal and obtain the journal’s guidelines to authors. See the separate guide to writing references.

2Writing a Thesis or Dissertation

The graduate school guide for writing a thesis or dissertation is at You must follow the instructions in that guide.

A typical thesis is organized as:

Abstract

Acknowledgements

  1. Introduction
  2. Literature Review
  3. Theoretical Analysis
  4. Experimental Setup and Procedures
  5. Experimental Results
  6. Discussion and Analysis of the Results
  7. Conclusions and Recommendations

Nomenclature

Bibliography

Appendices

The literature review summarizes the relevant literature on your topic including the importance of the work and gaps in the literature in need of research.

The objectives of your work should be specific (derive a correlation, measure the effect of…, optimize the design…). You objective of the work is NOT to do a “study” or some other vague activity.

Put material in the Appendices that may distract from the flow of your thesis. In particular, put raw experimental data in tables or figures in the appendices.

Some comments on thesis writing from faculty committee perspective:

  • In the abstract and introduction you should address the 5 questions:
  • What is the purpose or hypothesis of your work? (What are you trying to achieve?)
  • Why did you do this work? (What is the economic or technological significance?)
  • How did you do this work? (What theoretical or experimental approach did you use?)
  • What were your results?
  • What is unique about this work?
  • In the introduction chapter include your objectives (be specific) and a brief description of the organization of your thesis. Write your objectives to fit what you actually accomplished; sometimes this is different from what you started out to do.
  • After you have written your draft, go back and re-read your abstract, introduction, and conclusions. Are they consistent? Did you achieve your objectives? Why or why not?
  • The introduction should address the big picture. Why would someone read your thesis? If your work is aimed at a particular industry, why should people in that industry read your thesis? What did you do that will make a difference to their work/industry?
  • Your conclusions should restate each objective and state whether you accomplished the objective or if you discovered that an objective may not be achievable. It may be that you could only partially achieve an objective.
  • The conclusions chapter must also address future work.
  • In computer modeling you must (1) make sure numerical methods have converged (either by refinement of the mesh size to show no further improvements in calculations, or by comparison to a special case analytical solution), (2) check the sensitivity of your calculated parameters (this can be done as a parametric study, hold all parameters but one constant; vary the one parameter and determine how it must change to cause a significant change in the calculated results.) Parameters that are very sensitive may require additional attention to make sure of their accuracy.
  • In fitting models to experimental data you must show the goodness of the fit (see separate file on Bootstrap and Chi-square methods).
  • When running experiments, part of the experimental analysis is error analysis. Each measurement has an associated accuracy.
  • The error of an individual data point of some value F, where F=abc, is estimated by dF/F = da/a + db/b +dc/c (this is obtained by differentiating F). The errors of a, b, and c themselves may be estimated from functions or from the accuracy of the calibration of your measuring instruments; or by the scale accuracy of the instrument; whichever is greater. Sometimes this is simply engineering judgement.
  • The error of a functional fit can be displayed analogous to the 95% confidence interval provided in statistics.
3.Literature reviews
  • Use SCIFINDER scholar software
  • Seek assistance by the librarians