QHS Guidelines for Collecting Data via Excel Templates Page 4/262016
1.0 Introduction
This document describes how to optimally develop a data base in MS Excel. With these guidelines
• the data collection process may be streamlined,
• accuracy and consistency of the data may be increased,
• higher data quality may be achieved, and
• data analysis facilitated.
This document is broken down into three main areas; philosophy of developing a database, general layout and data
collection considerations. A Checklist is included in Appendix A. Appendix B gives a list of other documents you
may find helpful. They are available on the QHS website
http://www.mcw.edu/Quantitative-Health-
Sciences/Resources.htm.
2.0 Philosophy
2.1 Development of framework for database
When developing a database there are several general considerations:
• What is the purpose of the data collection?
o What are the aims?
• What will be the source(s) of the data?
o Data entry?
o Data uploads?
• What variables will be needed?
o How are the variables related?
Do they naturally divide into “forms”?
Are there repeated measurements?
o What will be the format and values?
• How will security be maintained?
o How and when will data be unidentified?
2.2 Data Dictionary
The data dictionary describes the variables and values that variables can take. It is best to create your data dictionary
first and then set up your template. The data dictionary can be circulated to other investigators to ensure there is
agreement on the data entered. This may save time for the person(s) abstracting data from charts or entering the data
into the spreadsheet. It may be helpful for your statistician working with your data as well, especially if the variable
names are not as meaningful to the statistician. Some examples are included below:
• For all variables “-9” indicates “missing” and “-8” means “not applicable.”
• ID: 4-digit number.
• Sex/gender: M=Male, F=Female
It’s possible to set up the data dictionary to be used for data validation as well. See section 4.3.
2.3 Maintenance of database
When maintaining a database there are several factors for consideration:
• Who is responsible for maintaining the database?
• Is there a procedure for modifications?
o Will there be one correct database?
o How will an audit trail of corrections be maintained?
o Will any changes be kept in a database with a new name?