Johns Hopkins Nursing Evidence-Based Practice Thrid Edition: Model and Guidelines

6 Evidence Appraisal: Research 127

content validity of a new instrument. Three issues are considered when experts

are evaluating content validity. An assessment of the relevance of an item is the

first issue. Experts are asked to rate an item on a 4-point scale of relevance to the

content being measured or to a target population. The second consideration is

the comprehensiveness of the items—that is, whether there are notable omissions

to measuring the construct. The last issue to consider is balance—for example, in

a multi-item scale, whether there are enough items to measure the construct to

ensure internal consistency in the subscales (Polit & Beck, 2017).

Construct Validity

A construct is a way of defining something that can’t be directly observed. Thus,

construct validity is particularly important for an abstract construct such as

motivation or intelligence. It concerns the measurement of abstract variables to

ensure the correctness of assuming that what the investigators are measuring

actually measures the construct of interest. If a researcher has taken careful steps

noted in establishing content validity of the instrument, the construct validity will

be strengthened (Polit & Beck, 2017). EBP teams often encounter discussion of

construct validity when instruments are used in research studies. This refers to

the degree to which the instrument measures the construct (or concept) under in-

vestigation. Questions a nurse may pose to get at threats to construct validity in-

clude these: “Did the researcher do a good job of defining the constructs? When

the researchers say that they are measuring what they call fatigue, is that what

they are really measuring?”

Cross-cultural validity is a type of construct validity. Cross-cultural validity is

concerned with the degree to which the items of a translated or culturally adapt-

ed tool perform adequately and equivalently to the original instrument. Transla-

tion and adaption require high levels of statistical applications (Polit & Beck,

2017).

Measures of Reliability

The concept of reliability is used to describe the consistency of a set of measure-

ments or of an instrument used to measure a construct. The question that is