1/48
Comprehensive vocabulary flashcards covering measurement validity, research validity, subject/investigator/logistics threats, reliability coefficients, diagnostic accuracy metrics, and likelihood ratios.
Validity
The degree to which a concept, conclusion, or measurement is well-founded and accurately reflects the real world.
Measurement Validity
The extent to which a test and/or instrument measures what it claims to measure (e.g., a goniometer measuring range of motion).
Research Validity
The extent to which a quantitative study of any type produces truthful, believable, and generalizable results.
Diagnostic Validity
The extent to which a diagnostic tool accurately identifies or predicts a specific health condition (e.g., EMG and nerve conduction diagnosing muscle function and nerve damage).
Face Validity
The extent to which an instrument or test appears to measure what it should on the surface; it is an intuitive and quick yes/no judgment that remains subjective and prone to bias.
Content Validity
The degree to which an assessment tool covers all relevant aspects and the entire domain of the construct being measured, typically judged by a panel of experts.
Construct Validity
The degree to which a test measures the theoretical construct it is intended to assess, consisting of convergent validity and discriminant validity.
Convergent Validity
A subtype of construct validity demonstrated when a test correlates highly with other instruments that measure the same or similar constructs.
Discriminant Validity
A subtype of construct validity demonstrated when a test does not correlate strongly with measures of unrelated constructs.
Criterion Validity
The degree to which the scores of an instrument are related to scores obtained with a current reference standard (gold standard).
Concurrent Validity
A subtype of criterion validity that reflects how well a test correlates with a current reference standard measured at the same time or within a reasonable time interval.
Predictive Validity
A subtype of criterion validity reflecting how well an assessment or test score predicts future performance, outcomes, or events (e.g., predicting falls over the next 6 months).
Internal Validity
The believability of a study, representing the extent to which observed results reflect the truth in the studied population without being caused by methodological errors or bias.
External Validity
The extent to which the findings and conclusions of a study can be generalized to other populations, settings, and conditions.
Selection Threat
A threat related to subjects where a nonrepresentative sample is chosen relative to the population of interest, producing inaccurate study results.
Assignment Threat
A threat related to subjects where the allocation process yields unequal baseline characteristics between study groups that could influence outcomes.
Attrition Threat
A threat related to subjects caused by the loss of participants over time (due to death, drop-out, or relocation), leading to reduced sample size or group baseline inequality.
Maturation Threat
A threat related to subjects involving natural internal biological or functional changes over time (e.g., spontaneous recovery or healing) that influence the outcome independently of intervention.
Compensatory Rivalry
A threat related to subjects where participants who discover they are assigned to a control group alter their behavior and work harder to prove they can improve anyway.
Resentful Demoralization
A threat related to subjects where participants who find out they are in a control group become resentful or discouraged and decrease effort or engagement.
Diffusion or Imitation of Treatments
A threat related to subjects occurring when control group participants communicate with experimental participants and learn or adopt parts of the intervention.
Statistical Regression to the Mean
The mathematical tendency for participants who score at the extreme high or low ends on baseline outcome measures to score closer to the average upon repeat testing.
Compensatory Equalization of Treatments
An investigator-related threat occurring when clinicians or researchers inadvertently or purposefully provide extra attention or alternative treatments to the control group.
History Threat
A study logistics threat involving external concurrent events occurring outside the experiment that are beyond the researcher's control and affect the outcome.
Instrumentation Threat
A study logistics threat arising from the inappropriate selection, malfunction, limitation, lack of calibration, or inaccurate application of measurement devices or techniques.
Testing Threat
A study logistics threat in which outcome scores change due to a participant's increased familiarity, practice with the procedure, or cues given by the tester.
Reliability
The ability of a measurement, test, or instrument to produce consistent and repeatable results when administered under identical conditions.
Intra-rater Reliability
The consistency of repeated measures taken by the very same evaluator on the same individuals under identical conditions.
Inter-rater Reliability
The consistency or agreement among measurements performed by different evaluators assessing the same individuals under identical conditions.
Test-Retest Reliability
The consistency of scores obtained when the exact same test is administered to the same individuals under identical conditions across two separate occasions.
Percent Agreement
A categorical reliability metric indicating the percentage of matching ratings; limited because it cannot distinguish chance agreement from true systematic agreement.
Kappa Coefficient (κ)
A statistical measure of inter-rater agreement for categorical data that accounts for and removes agreement expected by chance alone: κ=1−chance agreementobserved agreement−chance agreement.
Kappa Interpretation Thresholds
Standard guidelines for Cohen's kappa where values >0.75 indicate excellent/perfect agreement, values between 0.40–0.75 indicate fair to good agreement, and values <0.40 indicate poor agreement.
Index Test
The new or experimental diagnostic test undergoing appraisal and comparison against an established gold standard.
Reference Standard (Gold Standard)
The most accurate, established, and reliable diagnostic method used to establish the true presence or absence of a specific pathology or condition.
Spectrum Bias
A diagnostic appraisal pitfall where the index test is studied in a nonrepresentative spectrum of subjects, typically overrepresenting extreme severe cases and healthy controls.
Verification Bias (Work-Up Bias)
A diagnostic appraisal pitfall occurring when the decision to perform the reference standard depends on the index test outcome or prior clinical suspicion rather than being applied to all subjects.
Incorporation Bias
A diagnostic appraisal pitfall occurring when the index test being evaluated is already included as an active part or component of the reference standard.
Observer Bias
A bias that arises when assessors evaluating the reference standard or index test are not masked/blinded to previous results or patient status, influencing subjective interpretation.
Sensitivity
The ability of a diagnostic test to correctly identify individuals who truly have the condition, calculated as A+CA (True Positives divided by all individuals with the disease).
Specificity
The ability of a diagnostic test to correctly identify individuals who do not have the condition, calculated as B+DD (True Negatives divided by all individuals without the disease).
SnNOUT
A clinical rule stating that with high Sensitivity, a Negative test result effectively rules OUT the diagnosis because false negative rates are very low.
SpPIN
A clinical rule stating that with high Specificity, a Positive test result effectively rules IN the diagnosis because false positive rates are very low.
Positive Likelihood Ratio (LR+)
A diagnostic ratio assessing the odds of a positive test in diseased vs non-diseased individuals, calculated as 1−SpecificitySensitivity.
Negative Likelihood Ratio (LR−)
A diagnostic ratio assessing the odds of a negative test in diseased vs non-diseased individuals, calculated as Specificity1−Sensitivity.
Large and Conclusive Likelihood Ratio
A diagnostic likelihood ratio of LR+>10 or LR−<0.10, representing a large and decisive shift from pre-test to post-test probability.
Moderate Likelihood Ratio
A diagnostic likelihood ratio of LR+=5–10 or LR−=0.10–0.20, generating a moderate shift from pre-test to post-test probability.
Small Likelihood Ratio
A diagnostic likelihood ratio of LR+=2–5 or LR−=0.20–0.50, indicating a small but sometimes clinically important change in disease probability.
Negligible Likelihood Ratio
A diagnostic likelihood ratio of LR+=1–2 or LR−=0.50–1.0, indicating a negligible or clinically insignificant change in pre-test probability.