The ideal dataset will have:
a) a large number of numerical (interval, ratio) and categorical (ordinal, nominal) fields;
b) a large number of cases (e.g., at least 250, ideally over 1,000);
c) data fields that can act as independent variables (typically demographic, descriptive, or situational variables) and dependent variables (typically indicators of impacts, outcomes, or performance);
d) a sufficient range of variables (i.e., characteristics, attitudes, behaviours) that allow you to develop a variety of hypotheses (or guesses) as to what relates to what; and
e) data relevant to a business context, either for-profit or not-for-profit (at the event, customer, organization, or industry level).
Non-business or business data that has a faith or religious element (e.g., data on faith-based organizations, religious attitudes) would also be an interesting option. Students are welcomed to discuss with the instructor any data that may not fit the above description. Non-business data are acceptable as long as the data allow for developing hypotheses and analyzing relationships on topics that are of interest to a general audience (e.g., social issues, public attitudes, economic trends).
Peter Turkstra Library, Redeemer University , 777 Garner Road East, Ancaster, ON, L9K 1J4, Canada Circulation Desk Telephone: 905.648.2139 ext. 4266, Email: library@redeemer.ca