What a Data File Looks Like

From Market Research
Jump to: navigation, search

There are many ways to lay out data in an Excel-style file, but some are more useful than others. You typically want to have an Excel file which contains a single worksheet, as if you introduce an additional worksheet there is no way for software to work out what the relationship between the worksheets is, or which worksheet contains the data you are trying to analyze. Further differences between good and bad data layout are illustrated below.

Useless: Typical Tables of Survey Results

The term 'data file' is a bit misleading, as files are only useful if they contain data in the appropriate format. For example, the data shown below is largely useless. Yes, it shows some results from the survey. However, it is impossible to use this data to compute many useful things. For example, this data cannot be used to work out if males were more likely to be satisfied than females.


Useful: Tables Containing One Row Per Respondent

By contrast, the data below is appropriate for data analysis. This table contains:

  • A column for each of the different bits of data collected in the survey (e.g., age is in one column, education is in another).
  • A single row at the top containing the headings that describe each column.
  • A separate row for each respondent (i.e., all the data for the first respondent is shown the row under the headings, the row beneath this contains the data for the second respondent, and so on).


See also

Getting a Data File for more information about the best way to structure data for analysis.