STATISTICAL METHOD
Diagnostic Accuracy Analysis
Statistical Consulting for Test Accuracy Studies
We apply diagnostic accuracy analysis within your study design, building the 2x2 table against the reference standard, checking verification and spectrum bias and reporting sensitivity, specificity and predictive values with 95% confidence intervals.
No Commitment Required
Strictly Confidential
In-House PhD-Level Statisticians
Fixed Project Fee
Same-Day Response
METHOD OVERVIEW
What Are Sensitivity and Specificity?
Sensitivity and specificity measure how well a diagnostic test classifies people against a reference standard. Sensitivity is the proportion of people with the condition whom the test identifies, and specificity is the proportion of people without it whom the test correctly clears. Both come from a 2×2 table at a fixed threshold and do not depend on prevalence, whereas positive and negative predictive values do. They are reported with confidence intervals, and their validity rests on the reference standard and on how participants were sampled and verified. OnlineSPSS, a specialist statistical consultancy for academic, clinical and professional research, provides diagnostic accuracy analysis within your study design, with the relevant checks, effect estimates, uncertainty measures, interpretation and publication-ready reporting.
WHEN TO ENGAGE
When Are Accuracy Measures Appropriate?
Accuracy measures fit a test with a defined positive result judged against a reference standard, and these are the research situations in which researchers most often ask us to build or confirm the analysis.
Validating a New Diagnostic Test A rapid assay, imaging rule or screening questionnaire is compared with an established reference standard, and its accuracy at a prespecified threshold must be reported.
Predictive Values for a New Setting Clinicians need positive and negative predictive values where prevalence differs from the study sample, which means recalculating them from sensitivity and specificity.
Comparing Two Tests Head to Head Two tests were applied to the same participants, and differences in sensitivity and specificity need paired methods, not two separate sets of intervals.
Reviewing an Accuracy Analysis Only percentages were reported, or a reviewer asked about verification bias or STARD. Independent Statistical Review rechecks the 2x2 table and the bias handling.
ANALYTICAL APPROACH
How We Analyze Test Accuracy
Within the family of Clinical, Survival & Epidemiological, an accuracy study is judged from its reference standard and sampling first, and from its 2x2 table second.
Reference Standard and Sampling We examine how the reference standard was applied and who was verified. Verification bias, spectrum bias and an imperfect reference standard distort sensitivity and specificity.
Fixed Threshold, Interval Method The threshold is set before testing. Sensitivity and specificity get Wilson or Clopper-Pearson intervals, which hold up better than Wald intervals near 0 or 100 percent.
Prevalence and Predictive Values Predictive values are recalculated for the target setting's prevalence, with likelihood ratios alongside, because the study sample rarely matches where the test will be used.
Indeterminate Results and STARD We decide in advance how indeterminate and missing results are counted, then report participant flow and the 2x2 counts in a form aligned with the STARD checklist.
DELIVERABLES
What You Receive
As part of our Quantitative Data Analysis service, your diagnostic accuracy study analysis is delivered with the estimates, checks and files needed to report and defend it.
Data & Variable Review Test results, the threshold, reference standard outcomes and verification status are reviewed, with indeterminate and missing results counted before the 2x2 table is built.
Method-Specific Checks Verification pattern, participant spectrum and reference standard quality are assessed, with sensitivity analyses where verification was partial or the reference imperfect.
Statistical Results The 2x2 counts, sensitivity, specificity, predictive values and likelihood ratios with 95% confidence intervals, and paired comparisons where two tests share participants.
Publication-Ready Tables & Figures A participant flow diagram, the 2x2 table and an accuracy table, prepared in line with STARD reporting or the journal's requirements, ready to insert.
Interpretation Accuracy is read as a property of the test in a defined population and reference standard, and predictive values are tied to the prevalence they assume.
Reproducible Analysis Files R scripts, Stata do-files or SPSS syntax reproduce every table, interval and sensitivity analysis, so the results can be rerun if the data change.
ALTERNATIVE METHODS
When Another Method May Be More Appropriate
Fixed-threshold accuracy fits one test result at a time. These methods fit better when the marker is continuous, studies are pooled, tests are paired or accuracy varies by patient group.
01
ROC Curve Analysis
When the test is continuous and no threshold is settled, the ROC curve shows accuracy across all cutoffs and its AUC summarizes how well the test discriminates.
02
Diagnostic Test Accuracy Meta-Analysis
Combining several accuracy studies calls for a bivariate model that pools sensitivity and specificity jointly, not an average of reported percentages.
03
McNemar's Test
When two tests are applied to the same participants, McNemar's test compares their paired results, separately among those with and without the condition.
04
Logistic Regression
To ask whether sensitivity or specificity varies by patient group, logistic regression models the probability of a positive result separately by disease status.
CASE EXAMPLE
How We Work in Practice
An illustration of how an accuracy study moves from a prespecified threshold to 2x2 estimates that can be reported and defended.
ILLUSTRATIVE EXAMPLE · DIAGNOSTIC LABORATORY VALIDATION · UNITED STATES
A diagnostic laboratory validated a rapid point-of-care antigen assay against a laboratory reference method and needed accuracy estimates for its verification file.
Samples came from symptomatic patients at several sites, and many negative rapid results were never sent for the reference test, which would have inflated sensitivity if ignored. Each patient contributed one sample. The plan, written before the results were unblinded, fixed the manufacturer’s threshold, counted indeterminate results as test failures in a worst-case analysis, and specified Wilson intervals for the verified samples and a correction for partial verification with bootstrap intervals, assuming verification depended only on the rapid result. We reported both sets of estimates, and predictive values at two stated prevalences, not one.
Outcome: The laboratory received the 2x2 table, sensitivity and specificity with Wilson 95% confidence intervals, verification-corrected estimates with bootstrap intervals, predictive values at both prevalences and the scripts.
This example does not describe a real client or project. We never share details of actual client work, in line with the confidentiality we promise every researcher. This illustrative example is included to show the type, scope, and standard of work involved in this kind of analysis.
SOFTWARE
Statistical Software for Accuracy Studies
Diagnostic accuracy measures and their intervals are run in the statistical software researchers already use, with syntax or scripts provided where applicable.
RESEARCH DISCIPLINES
Test Evaluation Across Research Fields
Diagnostic accuracy analysis is used across research fields, each with its own tests, reference standards and reporting conventions.
A Defined Process for Your Analysis
We review your dataset, research question and design before any work begins, so the scope, deliverables, timeline and fixed project fee are clear from the outset.
01
Share Your Research Materials
Send your dataset, research question, study design and any existing analysis plan or output. We will let you know if anything else is needed to assess the project.
02
Receive Your Assessment
Your statistically trained Research Coordinator reviews the project and confirms the required analysis, deliverables, timeline and fixed project price before work begins.
03
Analysis & Delivery
Once you confirm the scope, an in-house statistician completes the agreed analysis and delivers the outputs, interpretation and reporting files.
RESEARCHER VOICE
What Researchers Say
Feedback from a laboratory medicine researcher who needed estimates and confidence intervals calculated and reported correctly.
“Our first pass at Sensitivity & Specificity (Diagnostic Accuracy) produced output, but I was not comfortable signing it off because of how effect sizes and confidence intervals should be calculated and reported. The analysis was rebuilt carefully and the interpretation was much clearer. The final tables included the effect estimates and 95% confidence intervals we needed for reporting.”
RESEARCH RESPONSIBILITY
Statistical Support Without Taking Over Your Research
We provide the statistical work for your accuracy study, and the study, its meaning and its authorship stay entirely with you.
Agreed Statistical Support We provide the agreed methodological, analytical and statistical reporting support for your diagnostic accuracy study, within the scope confirmed before work begins.
No Intellectual Contribution OnlineSPSS claims no intellectual contribution to your research. Our work is statistical support, and authorship remains yours.
Your Research Decisions The choice of reference standard, the clinical use of the test and your conclusions remain with you, and our interpretation stays within what the estimates and design support.
Confidential Handling Research data and project details are handled confidentially and are not published. An NDA is available where required, and can be arranged before sensitive materials are shared.
FREQUENTLY ASKED QUESTIONS
Questions About Test Accuracy Analysis
Short answers to what researchers ask before reporting test accuracy, from predictive values and bias to confidentiality and the assessment process.
Are predictive values a fixed property of a diagnostic test?
No, predictive values depend on prevalence, so the same test has a higher positive predictive value in a high-prevalence clinic than in a screening population. Sensitivity and specificity do not depend on prevalence, although they can change with the spectrum of patients. We recalculate predictive values for the target setting and report the 2×2 counts behind them.
Which confidence interval should I use for sensitivity?
Sensitivity and specificity are proportions, and the simple Wald interval performs poorly with small samples or values near 0 or 100 percent. The Wilson score interval or the exact Clopper-Pearson interval is preferable, and we state which one is used. Intervals should accompany the 2×2 counts, since percentages alone hide how few diseased participants there may be.
What is verification bias and does it affect my results?
Verification bias arises when only some participants receive the reference standard, often those with a positive test, which inflates sensitivity and lowers specificity. It matters whenever verification depended on the index test result. We examine the verification pattern and, when verification depended only on observed results, report corrected estimates beside the uncorrected ones.
Can you review an analysis that has already been completed?
Yes, we can review a completed accuracy analysis before submission or a response to reviewers. We check the reference standard, the threshold and when it was set, the handling of indeterminate results, the interval method, the predictive values against prevalence and the reporting against STARD, and confirm what stands or needs to change.
How do you protect my dataset and research confidentiality?
Your test results, reference standard data and any study protocol or analysis plan are handled confidentially, and project information is not published. An NDA is available where required and can be arranged before sensitive materials are shared. Read more about confidentiality and data security.
What happens when I request an assessment, and how long does it take?
You share your research question, design and relevant materials, and a statistically trained Research Coordinator reviews the request and responds the same day. The analysis, deliverables, timeline and one fixed project fee are then confirmed before any work begins, so the timeline follows from the agreed scope.
Still have a question? Find more answers in our full FAQ.
NEXT STEP
Need Support With a
Test Accuracy Study?
Share your research question, reference standard and data. We assess how your accuracy study should be analyzed and confirm the approach before any work begins.
No Commitment Required
Strictly Confidential
Fixed Project Fee
In-House PhD-Level Statisticians
Same-Day Response
Serving Researchers Since 2014