Measurement Scales (6 of 6)

previous


In this example, a statistical test could validly rule out chance as an explanation of the finding that 10-year olds lost more points from distraction than did 5-year olds. However, the statistical test will not reveal whether a greater drop necessarily means 10-year olds are more distractible. So, the conclusion that distraction costs 10-year olds more points than it costs 5-year olds is valid. The interpretation depends on measurement issues.

In summary, statistical analyses provide conclusions about the numbers entered into them. Relating these conclusions to the substantive research issues depends on the measurement operations.


previous