 |
|
|
 The
TRENDS demographic database has several options for presenting and downloading
data via the web. The terms "flat" and "cross-tabulated" refer to the arrangement or presentation of columns and rows in the output file. However, the presentation format has data storage and use implications. Flat file formats are useful for statistical applications and storage in databases; cross-tabulated formats facilitate thematic mapping and change-over-time graphs. A major difference between these two file formats is that a flat file stores multi-attribute data for the same place using multiple columns and multiple rows – one column for each attribute and one row for each time period. A cross-tabulated file only stores one attribute per row for each place, but can include multiple years with each observation assigned to a separate column. In a flat file the first several columns locate the data in space (fips code, lter site, state, and county) and time (year). The next column or columns contain the attribute observations for that place and time such as number of commercial establishments [COMEST], total acres of farmland [FTOT], average wages of service employees [SWAGE], etc. Multiple observations for each place for multiple years will appear on separate rows. In a cross-tabulated file the first three columns locate the data in space (recorded by fips code, lter site, state, and county). The next column or columns contain the attribute observations for each time period. Observations for multiple years for the same place will appear on the same row at the intersection with the column for the appropriate year. |
