Statistics 770

How to Write a Seminar Report

Updated 2009-09-15

---

Some General Comments

• The report should be no more than 4 or 5 pages in length, double-spaced.
• The report must be prepared in LaTex. A good visual presentation is important.
• The emphasis of your report should be on the ideas presented in the seminar, not on the formulas and proofs. Give formulas only as necessary to illustrate specific points. If you want to describe a derivation or a proof, you should outline the approach used without repeating all of the details.
• Organize the report into headings and, where appropriate, sub-headings. Write concisely.
• If you cite any papers (for example, papers from the reading list), include a list of references at the end of your report.

Suggested Organization of the Report

INTRODUCTION

Say something about the speaker. Where is the speaker from? What is the speaker's background? Why is the speaker interested in this topic?

Introduce the topic in non-technical words. Explain the context in which the work was done. Get your reader interested in the topic.

What were the speaker's objectives for this seminar?

CONTENT

Give an overview of the seminar. Illustrate this with specific details you feel are particularly interesting or important. Quote formulas or details of derivations or proofs if you think they are of interest in their own right, or if they are essential to an understanding of the material.

DISCUSSION

Is the work original, novel, or elegant? Is it useful? Does it have applications in other fields of research? Could it be useful to you? How does it relate to what you already knew? Does it improve on earlier solutions to the same problem? Can you see applications in addition to ones the speaker described? Did the speaker suggest future directions? Do you have any other ideas for future work on this topic?

CONCLUSIONS

Summarize, very briefly, in plain language, what you learned from the seminar. How important do you think the topic is? Does this work make a significant contribution?

• We should not have to find your errors of grammar, spelling and formatting for you; your first submission must be perfectly clean. Remember things like:
  • verbs must agree with singular or plural nouns;
  • capitalize only proper names and book titles;
  • put just one space after any punctuation;
  • in formulas, variables must be in italics, function names must not be in italics, vectors and matrices must be in bold;
  • insert punctuation after formulas as though they were phrases or clauses in a sentence;
  • indent at the start of paragraphs but not where a paragraph continues after a displayed formula;
  • put one space before, not after, an opening parenthsis ( and one space after, not before, a closing parenthesis ) if it is not followed by punctuation.
• Use past tense for what was said or done, present for what is true, present progressive for what is happening now, future for what will happen.
• Use no article if you are referring to a general class of things, "the" if it is an identified member of the class, and "a" if it is any member of the class. For example: "In a study of [no article here] tests of hypothesis, the t-test is an important test."
• The first time you mention a person by name, give their first and last name with no title. On subsequent mentions, give only their last name.
• All works cited must be in the references, all works in the references must be cited in the text. References must follow Biometrics style exactly. Get a recent copy of Biometrics and note the use of periods, parentheses, and italics. Note that no abbreviations are used for names of books and journals. Note the different formats for citing a paper in a journal, a complete book, a contributed chapter in an edited book, an unpublished work, a web site, etc. If you want to include background reading in the references, you must cite those works in the Introduction.
• Think of writing a report as though you are writing a story. You don't just list everything that happened, you have to present it in such a way that one idea flows into the next. The Introduction has to set the scene and provide the background, context and motivation. In a story, the Introduction will introduce the main characters and tell you enough about who they are and what has happened to them already so that you will be eager to know what happens to them next and how they will interact with each other. In a technical report, the Introduction has to introduce the ideas in much this way. Your Introduction has to be sufficiently non-technical that the intended reader (in your case, a trained statistician) can understand it without reading the rest of the report. It has to be interesting enough that the intended reader will want to read the report.
• The Discussion will relate the ideas to each other and to whatever else we might know about them. The Conclusions will summarize what has been learned, looking back to the Introduction to recall what the speaker was trying to achieve.
• You can't introduce new material in the Discussion or Conclusions unless you are certain that the intended reader will know it already. All the ideas of the seminar must be outlined or foreshadowed in the Introduction, presented in the Content and discussed (not introduced) in the Discussion and Conclusions.
• Get someone else to read your report and look for typos, missing words, "empty" words, confusing explanations, etc.

---

Statistics 770