ollecting valid data is an art unto itself. Ensuring that your results are valid and unbiased is a skill beyond the scope of this module. However, even quality data can be corrupted by human error and inattention to detail.

Some of the most common data errors include:

• Data that are not read properly
• Dates are interpreted improperly
• Hidden characters that result in improper fields (such as a tab character in a field that looks blank but causes SPSS to classify the field as a "string")
• SPSS auto-assigning incorrect levels of measurement
• Data that are out of range
• Impossible categories
• Calculating a nonsensical mean (if 1 = blonde, 2 = brunette, 3 = red hair, 4 = black hair, etc., then a mean hair color of 3.1 provides no useful information)
• Lower- and upper-case letters used interchangeably (such as "f" and "F" for sex)
• Inconsistency of measurement and coding

Be sure to review the demonstration video in the Learning Resources, which explains how to identify and fix common data errors. It also discusses specific errors commonly introduced when Excel data are imported into SPSS.