Chapter 10: Concepts of Measurement Validity

Introduction

—with K. Douglas Gross

Clinical decision-making is dependent on the accuracy and appropriate application of measurements. Whether for diagnosis, prognosis, or treatment purposes, without meaningful assessment of a patient’s condition, we have no basis for choosing the right interventions or making reasoned judgments. Validity relates to the confidence we have that our measurement tools are giving us accurate information about a relevant construct so that we can apply results in a meaningful way.

The purpose of this chapter is to describe the types of evidence needed to support validity and the application of this evidence to the interpretation of clinical measurements. Statistical procedures related to validity will be identified, with full coverage in Chapter 32.

Defining Validity

Validity concerns the meaning or interpretation that we give to a measurement.¹ A longstanding definition characterizes validity as the extent to which a test measures what it is intended to measure. Therefore, a test for depression should be able to identify someone who is depressed and not merely tired. Similarly, an assessment of learning disabilities should be able to identify children with difficulties processing information rather than difficulties with physical performance.

Validity also addresses the interpretation and application of measured values.² Beyond the tool itself, validity reflects the meaning ascribed to a score, recognizing the relevance of measurements to clinical decisions, as well as to the person being assessed.^3,4

We draw inferences from measurements that go beyond just the obtained scores. For example, in clinical applications, when we measure variables such as muscle strength or blood pressure, it is not the muscle grade or blood pressure value that is of interest for its own sake, but what those values mean in terms of the integrity of the patient’s physical and physiological health.

Validity addresses three types of questions:

• Is a test capable of discriminating among individuals with and without certain traits, diagnoses, or conditions?

• Can the test evaluate the magnitude or quality of a variable or the degree of change from one time to another?

• Can we make useful and accurate predictions about a patient’s future status based on the outcome of a test?

We use those values to infer something about the cause of symptoms or presence of disease, to set goals, and to determine which interventions are appropriate. They may indicate how much assistance will be required and if there has been improvement following treatment. Such scores may also have policy implications, influencing the continuation of therapy, discharge disposition, or reimbursement for services. Measurements, therefore, have consequences for actions and decisions.⁵ A valid measure is one that offers sound footing to support the inferences and decisions that are made.

Validity and Reliability

A ruler ...