Meta-Analysis Accelerator | AI Research Tools

Available conversion tools

• Adjusted LSM → Mean & SD — Convert an adjusted least squares mean (LSM), its standard error, and model degrees of freedom into a mean, standard deviation, variance, t-statistic, p-value, and confidence interval.
• Approximate AUC from Confusion Matrix — Estimate the Area Under the ROC Curve from a single operating point (TP, FN, FP, TN) using the trapezoidal approximation. This is NOT a true ROC-based AUC.
• Baseline & Change → Final/Post Mean & SD — By providing the baseline mean, baseline SD, mean change, SD change, and an optional correlation coefficient, this tool calculates the final (post-intervention) mean and standard deviation with automatic sensitivity analysis across multiple assumed correlations.
• Change in Mean & SD from Baseline — By providing the means and standard deviations (SD) before and after an event, along with the correlation coefficient, you can calculate the mean change and standard deviation (SD) change.
• Combine Mean & SD for Multiple Groups — This conversion process takes the means, standard deviations, and totals of multiple groups and combines them into a single set of values, making it easier to work with and interpret data.
• Combine Mean & SD for the Same Group (Subscales) — Combine two or more subscale scores measured on the same participants (e.g., SF-36 PCS + MCS) into a single composite TOTAL mean and SD, with automatic sensitivity analysis at r = 0.25, 0.50, 0.75 plus the user-specified correlation. For sums of subscales only — not for averaging paired body-side measurements.
• Combine Mean & SD from Multiple Anatomical Regions (Average per Participant) — Average per-participant calculator: combine correlated measurements from multiple regions or paired sides on the SAME participants (e.g., left/right limb, medial + lateral femoral condyle cartilage thickness, BMD across regions, muscle thickness at multiple sites, retinal quadrants) into a single AVERAGE per-participant mean and SD, with sensitivity analysis at r = 0.25, 0.50, 0.75 plus a user-specified r. Replaces the older 'Average per Participant, then Summarize' tool. NOT for questionnaire subscales — use 'Combine Subscales' for total scores.
• Combine Multiple Arms → Single log(Effect) & SE — Combine independent arms of one study with inverse-variance pooling and a RevMan-style forest plot.
• Confidence Interval (CI) → SD in 2 Groups — By inputting the upper and lower confidence interval (CI) limits, CI level, and the number of participants in each group, this tool calculates the corresponding standard deviation.
• Confidence Interval (CI) → SD in One Group — Calculate the standard deviation from the confidence interval limits, CI level, and sample size for a single group.
• Convert HR, RR, or OR → log(effect) & SE — Convert a ratio-based effect estimate (HR, RR, or OR) and its confidence interval into the natural logarithm and standard error, ready for generic inverse-variance meta-analysis.
• Effect Estimate (RR, OR, HR) & P-value → 95% CI — By inputting a ratio-based effect estimate (Risk Ratio, Odds Ratio, or Hazard Ratio) and its two-sided p-value, this tool reconstructs the 95% confidence interval using a log-normal approximation.
• Generic Inverse-Variance Meta-Analysis + Forest Plot — Pool HR, RR, OR, IRR or rates across studies with fixed or random effects and a RevMan-style forest plot.
• Hazard Ratio (HR) → Risk Ratio (RR) — Convert a hazard ratio to a risk ratio under four scenarios: exact under proportional hazards (baseline risk P0 known), rare outcome (RR ≈ HR), common outcome with no baseline (VanderWeele 2017), or a common-outcome interval [w, u] (VanderWeele 2020). Pick a scenario, then work in the spreadsheet.
• Interquartile Range (IQR) → Standard Deviation (SD) — Estimate the standard deviation from the interquartile range (IQR) and sample size using the Wan et al. (2014) method. Uses the formula SD = IQR / [2 × Φ⁻¹((0.75n − 0.125)/(n + 0.25))].
• LSM (Unadjusted) → Mean & SD — Convert an unadjusted least squares mean (LSM) and its model standard error into a mean, standard deviation, variance, and confidence interval. SD is reconstructed as SE × √n.
• Mean & SD for Individual Patients' Data — By inputting individual patient measurements, you can obtain valuable statistics including sample size, mean, standard deviation, variance, sum, standard error of the mean (SEM), and confidence intervals at multiple levels.
• Means & SDs → Cohen's d / SMD & Effect Size r — By inputting the means and standard deviations of both groups, you can calculate both Cohen's d or standardized mean difference (SMD) and the effect size correlation (r).
• Median & IQR → Mean & SD — By inputting the median, quartile 1 (Q1), quartile 3 (Q3), and the sample size of the group, you can obtain the mean and standard deviation (SD) using the Wan et al. method.
• Median & Range → Mean & SD — This conversion requires four inputs: the median, minimum, maximum, and the sample size of the group. The output consists of two values: the mean and the standard deviation (SD) using the Wan et al. method.
• Odds Ratio (OR) → Cohen's d — Researchers can input the odds ratio value, and the converter will provide the corresponding Cohen's d.
• Odds Ratio (OR) → Risk Ratio (RR) — Convert an odds ratio to a risk ratio under four scenarios: exact (Zhang & Yu with baseline risk P0), rare outcome (RR ≈ OR), common symmetric (√OR), or common interval (VanderWeele bounds). Pick a scenario, then work in the spreadsheet.
• P-value → SD in 2 Groups — This conversion takes into account the p-value between two groups, the number of participants in each group, and the mean difference to estimate the standard deviation (SD).
• Percentage Change ⇄ Absolute Change — Convert a reported mean percentage change from baseline into an absolute mean change and SD, or the reverse, ready for meta-analysis.
• Range → Standard Deviation (SD) — With three inputs: minimum, maximum, and sample size. This conversion provides an estimated standard deviation (SD).
• Standard Deviation (SD) → Correlation Coefficient (CC) — By providing three inputs: baseline standard deviation (SD), post-treatment standard deviation (SD), and standard deviation (SD) change. This conversion yields a single output: the correlation coefficient (CC).
• Standard Deviation (SD) → Standard Error (SE) — With just two inputs, the standard deviation (SD) and the sample size of a group. This conversion provides you with the corresponding standard error of the mean (SEM).
• Standard Error (SE) → Standard Deviation (SD) — With just two inputs, the standard error (SE) and the sample size of a group. This conversion provides you with the corresponding standard deviation (SD).
• Standard Error (SE) of AUC — Calculate the standard error of the Area Under the ROC Curve (AUC) using the Hanley & McNeil (1982) method. Requires AUC, number of diseased cases, and number of non-diseased cases.
• T-value & df → Cohen's d / SMD & Effect Size r — By inputting the t-value and degrees of freedom, you can calculate both Cohen's d and the effect size correlation (r). Supports both independent and dependent (paired) designs.
• TP, FN, TN, FP Calculator — Calculate true positives, false negatives, true negatives, and false positives from diagnostic test parameters such as sensitivity, specificity, prevalence or PPV, and total sample size.
• Unadjusted Odds Ratio (OR) Calculator — By inputting the counts of events and totals for two groups or conditions, this calculator quickly computes the odds ratio along with its confidence interval, Z-score, and p-value.
• Unadjusted Risk Difference (RD) Calculator — By inputting the counts of events and totals for each group, this calculator quickly computes the risk difference (RD) along with its confidence interval, Z-score, and p-value.
• Unadjusted Risk Ratio (RR) Calculator — By providing the counts of events and totals for two different groups or conditions, this calculator swiftly calculates the risk ratio or relative risk (RR) along with its confidence interval, Z-score, and p-value.