Validity and Reliability A Prof Dr Maha ALNuaimi
Validity and Reliability A. Prof. Dr. Maha AL-Nuaimi
Medical or epidemiological study, major consideration is to obtain: Valid measurement Reliable measurement on the exposure factors and outcomes of interest in the study population “WITHOUT BIAS and ERRORS” or to minimize them to the least as possible
To achieve a high standard quality study: Ensure right answers to study questions Good the study design Valid and reliable the measurements. Control for any possible bias Good cooperation between * research group and * study population
How good is the instrument or tool? • instrument • tool measurement – without bias or error – minimize bias • true value truth • measurement – valid/accurate – precise/reliable
PRECISION DEFINITION : A precise measurement in one that has nearly the same value each time it is measured SYNONYM • reliability • repeatability • reproducibility • consistency • agreement
Precision is usually affected by random errors, there are 2 main sources of errors: Subjective: - subject variability due to intrinsic biological variability ex. : fluctuation in mood, B. Pr. , heart rate… Objective : 1 - Observer: hand-eye coordination 2 - Instruments …. Standardization … why? ?
The more precise a measurement the greater the statistical power. Reliable (qualitative) …. Precise (measured & dependable) Assessing precision: usually by S. D. of repeated measurements, also the Co. V. or correlation coefficient. Test – retest. . Kappa test ….
Strategies For enhancing precision: Through decreasing random errors, and this is done by: 1 - Standardizing the measurement method 2 - Training & certifying the observer 3 - Refining the instruments 4 - Automating the instruments 5 - Repetition: repeat the measurement & use the mean
ACCURACY DEFINITION: The degree to which the results of a measurement correspond to the true state or truth SYNONYM: • validity • conformity
IMPORTANT POINTS • It is attributed to: – Methodological aspect of study design or analysis – Selection of subject – Quality of information obtained – Confounding – Misclassification
IMPORTANT POINTS • accuracy is a function of “ SYSTEMATIC RROR” • VIP influence on the internal and external validity of the study • the greater the systematic error , the less accurate the variable
Accuracy is affected by systematic errors (bias) of two types: 1 - Subjective: Co-operation, re-call, illiteracy… 2 - Objective : -observer: conscious or un-conscious -Instruments: which are not calibrated recently. .
If the test is reliable but not accurate. . Is it dependable?
Strategies for enhancing accuracy. 1 Standardizing measurement methods. 2 Training and certifying the observers. 3 Refining the instruments. 4 Automating the instruments. 5 Making informal measures that the observer is unaware of. 6 Blinding. 7 Calibrating the instrument
. . . good precision poor precision good precision poor accuracy good accuracy . . . poor precisio poor accuracy Illustration of the difference between Precision and Accuracy
Types of validity • Construct validity: the degree to which the measurement agrees to measure what it claims to measure (content validity) • Internal validity: the tool measured what intended to measure precisely and accurately. . So reach to real association • External validity … generalizability. .
- Slides: 16