Course Preparation

In order to get the maximum value out of the multilevel modelling course, it is recommended that students do a bit of preparation in advance.

Title:   Multilevel Analysis - An introduction to basic and advanced multilevel modelling
Authors:  T. Snijders and R. Boskers
Date:  1999
Publisher:  Sage
Cost:  $55 (incl. all taxes) (a reduced price since we are submitting a bulk order)

If you want a copy of this book, contact James Valcour (jvalcour@upei.ca) BEFORE April 1st. You can pay for the book at the time of course registration.

1. Title
   
2. Short introduction - background to the study
3. Description of your data:
   1. specify the levels of organization (e.g. cow, herd, county) and  number of units at each level
   2. identify the key dependent variable(s) and the level at which they exist
   3. identify the key independent variables and the level at which they exist
4. Purpose:
   1. what is the most important hypothesis you want to study (this should fit with the dependent variable identified above)
   2. what are your expectations (based on literature or previous work with the data

These project descriptions will be photocopied and distributed to the class.
        Note: If you are not bringing any of your own data to the course, that is OK. You can either team up with a student who has data or work on some sample datasets that we will provide
   
    If you are willing to volunteer to be one of the students who will present their problem to the rest of the class, prepare a single acetate (for overhead projector) with the key elements of the description of your project. Presentations will take place on the first evening and will be limited to 10 - 15 minutes.

1. One record per observation at the lowest level of the hierarchy (e.g. if the dataset contained data from lactations, within cows within herds, the dataset should have 1 record per cow)
2. Make sure that each observation is uniquely identified (e.g. herd id, cow id and lactation number)
3. Identify the key variables of interest and create a dataset with just those variables in it (rather than brining the whole dataset if it is very large)
4. If there are a lot of missing values, you might want to prepare a dataset that consist of those observations for which complete data are available
5. Prepare a description of the data for your own use (list of variables and their definitions)
6. You can bring the data in any computer format you like, but we would suggest that some form of spreadsheet (Excel, QuattroPro or Lotus) would be the easiest to work with
1. If you bring the data in a spreadsheet, have the variable names in row 1 and the data immediately below (starting in row 2) - do not include anything else in the spreadsheet