In short, what you do with your data during the data collection until up to the end of the project. Why? How? Disciplined use of spreadsheets for data capture The datadef worksheet contains the variable names and the variable definitions which briefly acquaint us about the background of each data field.
variable names: factors being observed used also as column headers
variable definitions: short descriptions for each factor observed datadef sheet Disciplined use of spreadsheets for data capture The data worksheet on the other hand contains the research data. It includes the column headers and the data values.
column headers: names of the parameters in the research study, which correspond to the variable names.
data values: values that have been observed or measured for each of the parameters data sheet Disciplined use of spreadsheets for data capture 1. Column header alignment: center format: all characters in lower case, bold, no space (but can use underscore '_') 2. Data values Date oalignment: align right oformat: dd-mmm-yyyy (e.g. 01-Jan-2009) Numerical data oalignment: align right oformat: specified number of decimal places Text data oalignment: align left Same length character data oalignment: center General format oColumns: same column spacing for all worksheets, no grid lines oFonts: same style, same size, same color Data validation and checking Data validation ensuring that data inserted into an application satisfies defined formats and other input criteria discovers inconsistencies in the data set
Quality checking ensuring that the data is fit for analysis inconsistencies are either confirmed as valid or discarded (not included) Data validation and data quality check are prerequisites before doing an actual analysis.
'Garbage in - garbage out' Hands-on Documentation & Metadata Documentation Show how data or data sets were created Datas structure and content Any modifications, etc Designed to be human-readable Documentation & Metadata Metadata Subset of core data documentation Designed to be read by a computer (may or may not be human-readable) Set of structured information: Purpose Origin Creator/Author Access