513-697: Guidelines for Preparing Report
                   
                 See below for details specific to course 678
                              
Adapted from rules for Third Annual Undergraduate Data Analysis Contest;
() items in parentheses do not apply to 513-697. 

The basic idea (of the contest) is to involve (undergraduate) students in the
statistical analysis of a significant, real world data set under conditions
that simulate expectations placed on a consulting statistician.

Rules for Report:

  1. In the real world, almost all applied statistical analyses have three
     distinct stages:
        o The analyst must translate a question stated in ordinary language
          into precise quantitative terms;
        o S/he must perform the technical analysis;
        o S/he must explain his/her findings in ordinary, jargon free
          language to people who do not understand statistics very well.

     To simulate this, we are requiring that the report you submit to be
     judged consist of three parts:
     
        o An explanation as to how you translated each question into
          quantitative terms, with an argument to show that your translation
          is a reasonable one;
        o A description of your technical analysis in such detail that the
          judges can assess your purely technical ability;
        o A summary report that clearly and completely answers the original
          questions in non-technical language any interested citizen could
          understand, and which points out any unusual features of the data
          that would be of interest to the public.
          
  2. Mainly for monetary reasons, (social) scientists usually cannot collect
     new data every time they come up with a new question to answer; they
     must recycle existing datasets. And that means that, typically, they do
     not have information on all of the variables that would be relevant to
     their analyses. (For example, the datasets for Contest Question 1 lack
     at least one important variable). Be sure to mention in your
     nontechnical summary what variables you would have liked to have
     besides the ones that were given to you, and, if necessary, state
     clearly how lack of these variables might limit the force or scope of
     your conclusions. Also describe in your nontechnical summary how any of
     the variables you were given might have been recorded differently so as
     to make them more informative or helpful.
     
  3. In the real world, time is money, especially your boss's time: any
     analyst would have to give a compact, meaty presentation in relatively
     few words and almost certainly one with nice graphics. Accordingly, we
     are requiring that the three parts of your report must together
     comprise not more than (10 typed) pages of text, with (10 pages of
     pictures or charts, etc. That is, you may submit (no more than a total
     of 20 pages as your entry). You should especially aim to make the
     nontechnical summary as brief as possible without sacrificing clarity
     or correctness.
     
  4. Your report must be submitted on 8.5 by 11 inch paper with nothing less
     than a 10 cpi (12 point) font.


Additions/modifications by jh FOR  697 COURSE fall '98
======================================================

* an upper limit of ______ words for the text (excluding tables/figures)

* an upper limit of ___ pages (also 8.5 by 11) for tables/figures/output 

* use only the relevant EXCERPTS from computer output. They can be
  "cut and pasted" the old-fashioned way with scissors and glue or tape;
  you can put several on a page if you wish, but label these "exhibits" 
  clearly. Only items that the writer wants the reader to see should be 
  included. These items must be ANNOTATED; using legible handwriting to do 
  so is quite acceptable; highlighting is ok too, IF is is annotated or 
  EXPLICITLY CITED in the text. Remember that even though the instructor 
  may know what some items are, he will also read the report through the 
  eyes of a non-statistician.
  

Additions/modifications by jh FOR THIS SUMMER '99 COURSE 678
==============================================================
 
If report is presented orally...

  max   20 minutes presentation per project
  up to 10 minutes of questions/discussion 
  
  plus... an upper limit of 10 printed pages (also 8.5 by 11)
  for additional tables/figures/output 
 
  Use only the relevant EXCERPTS from computer output. They can be
  "cut and pasted" the old-fashioned way with scissors and glue or tape;
  you can put several on a page if you wish, but label these "exhibits" 
  clearly. Only items that the writer wants the reader to see should be 
  included. These items must be ANNOTATED; using legible handwriting to do 
  so is quite acceptable; highlighting is ok too, IF is is annotated or 
  EXPLICITLY CITED in the text. Remember that even though the instructor 
  may know what some items are, he will also read the report through the 
  eyes of a non-statistician.
    
  If you do not have the facilities or budget, please arrange with LeeAnn
  (room 36, tel 398 6260, email LeeAnnF@Epid.Lan.McGill.CA ) to have 
  multiple hard copies of your presentation material and annex made before
  the class on the 25th.
  
  I will also need a copy of your work, so that I can give you written
  comments (I hope you will also allow me to share my reactions with
  the rest of the class)
  
  Also, we can have a laptop available if you want to do a presentation
  from the screen rather than from acetates (i.e. transparencies)