++
++
In Chapter 3, you learned that outcomes cannot be simply focused on the disease or physical condition of an individual person. Rather, outcomes exist along a disability continuum that ranges from the pathophysiology of a condition to societal limitations put on the individual person by the disease or physical condition. Within this context, you also learned that different types of outcomes exist for assessing different points along this continuum. These included outcomes that are focused on the disease, as well as outcomes that are meaningful to the clinician and to the patient or client. However, for any of these outcomes to be useful, they must do more than simply make sense at face value. Rather, they must have certain fundamental properties that address their quality as outcomes. These properties tell us that the outcomes are consistent, quantify what they are purported to measure, and behave in meaningful ways as the status of the patient or client changes.
++
In this chapter, you will learn about the properties needed for an outcome to be meaningful. You will start by learning how outcomes of individual clients are compiled into summary indices that quantify the data's central tendency (e.g., the mean). You will also learn how the standard deviation is used to quantify the variability of the data. The four fundamental categories that differentiate all data are described. For an outcome to be meaningful, it must be reliable (consistent). You will learn about the different types of reliability and how they apply to your given question. An outcome must also be valid by measuring what it claims to measure. Like reliability, validity comes in a variety of forms, and knowing the differences is important in selecting the best outcomes. Finally, for outcomes to be meaningful in clinical cases, other properties such as important differences, responsiveness, and interpretability are also required. You will be introduced to these different concepts and how the properties are determined.
++
Learning Outcomes
After reading this chapter, you should be able to answer these questions:
When is it best to use the mean, median, or mode?
What is the difference between a standard deviation and a standard error?
How is a confidence interval used?
What is the difference between standard error of the measure, minimal detectable change, minimal clinical important difference and global rating of change?
How are reliability and validity related?
What is the difference between interpretability and responsiveness?
++
Key Terms
Accuracy
Bias
Ceiling effect
Concurrent validity
Confidence interval (CI)
Construct validity
Convergent validity
Criterion validity
Divergent validity
Face validity
Floor effect
Internal consistency
Interpretability
Intersession reliability
Interval data
Intrasession reliability
Mean
Measurement error
Median
Minimal detectable change
Minimal important difference
Minimal clinical important difference (MCID)
Mode
Nominal data
Ordinal data
Precision
Predictive ...