Variables

Predictive Validity and Prognosis Studies

Variable that is presumed to predict an outcome of interest

Intervention studies

presumed cause of or influence upon a measured effect

Levels of independent variables

number of forms the independent variable takes in a study

Single-factor (one-way)

effect of one variable

Two-factor (two-way)

effect of 2 variables as well as their interaction with one another

Factorial design

number of independent variables or factors included

Dependent variable

variable hypothesized to be caused or predicted by the by the independent variable

The dependent variable is the

outcome of interest

Extraneous variables

confound the relationship between the independent and dependent variable

Extraneous variables need to be

anticipated and controlled for if possible

Discrete variables are the same as

categorical variables

Continuous variables are the same as

scale variables

Discrete variables

can assume only distinct values

discrete variables can be

dichotomous or polytomous

A likert scale is a

discrete variable

Continuous variable

theoretically can assume infinitely finer degrees of measurement depending upon the instrument utilized

Measurement

method of assigning quantitative or qualitative values to variables

Utilizing measurements requires

clearly defined rules that are consistently applied during a study

Nominal

classification without value placed on the category and no ranking or order

Nominal data can use

names or numerals

Ordinal

classification with order, but without equal intervals between levels

Ordinal data can

use numbers to label categories, but can’t do math

Interval

order and interval distance known, origin is unknown

With interval data you can do

addition and subtraction

Ratio

order, interval distance, and origin are known

With ratio data you can

add, subtract, multiply, and divide

Measures are a combination of

true value (signal) and error (noise)

Sources of error for reliability

errors made by examiners, subject variability, instrumentation flaws or failures

Categories of measurement reliability

instrument and rater

Instrument reliability

test-retest, internal consistency (constructs)

Rater reliability

Intra-rater (within), inter-tester (between or among)

Measurement validity

the degree to which a measure captures what is is intended to measure

reliability is a… but… condition for validity

Necessary, not sufficient

Types of measurement validity

face, content, construct, criterion related

Construct validity types

convergent and discriminant

Criterion related validity types

concurrent and predictive validity

Face validity

does the measurement appear, on the face of it to assess what is intended

Face validity is

all or nothing

Face validity is addressed from the standpoint of

the tester and the patient or family member

face validity is

rather subjective

Content validity

extent to which a measurement is judged to reflect the meaningful elements of a variable

content validity is judged by

content experts or people with experience with the variable

content validity is usually only pertinent to

multidimensional measurements

Examples of things that content validity would be assessed in

disability measures, functional measures, self-report tools, knowledge assessment

Construct validity

validity of abstract concepts that underlie the measure

construct validity is achieved via

operational definitions, logical arguments, theoretical arguments, and research evidence

convergent validity

comparison of scores between two similar instruments expected to produce similar results

discriminant validity

differentiation among characteristics or levels of the same characteristic

Criterion validity

extent to which one measure is systematically related to other measures or outcomes

Criterion validity requires

direct comparison of index measure with a standard (criterion) measure or with known outcome

Concurrent validity

ability of an index measure to capture an outcome similar to that of another measure

To ascertain concurrent validity you compare

the index measure to the criterion measure, that were obtained at approximately the same time

Predictive validity

the ability of an index measure to predict a future outcome

To ascertain predictive validity you compare

the index measure to the criterion measure that was obtained at a later point in time

Responsiveness to change

ability of a measure (instrument) to detect change in the phenomenon of interest

responsiveness to change fit

between instrument and operational definition of variable

Responsiveness to change depends upon

number of values on the measurement scale

Standard error of measurement (SEM)

extent of which observed scores are dispersed around the true score

Minimal Detectable Change (MDC)

tiniest amount of change needed to detect difference

If a measurement has less error and is more reliable, it will have

a lower MDC

All measurement tools have

error

Floor effect

failure of a measure to detect lower scores for patients whose status has declined

Ceiling effect

failure of a measure to detect higher scores for patients whose status has improved