The roster
Analysis Studio is a statistical analysis application that runs entirely in your browser. Your data file never leaves your device: you load it, the calculations happen on your machine, and the results appear on screen with the checks, the effect sizes, and the intervals that belong with them. The Studio holds twenty-one procedures on a twenty-module roster. Module 11 carries two procedures, survey weighting and raking plus propensity adjustment for nonprobability samples, because the second builds directly on the first.
Every module follows the same pattern. You pick your variables from your own data, choose options from grouped controls, read what you will get before you run, and then run. Each result set carries assumption checks, effect sizes with intervals where they are defined, and a methods record. Files the Studio reads include CSV, TSV, SPSS, Stata, SAS, and Excel formats, with variable labels and value labels preserved.
What free means here
Analysis Studio is free to use. Nine procedures are free in full: Modules 1 through 7, Module 11, and Module 12. A subscription, at $299 per year, covers downloads and Copy Results on every module, and covers running the remaining procedures.
Promotion, stated with its date so it cannot be misread later: through March 31, 2027, all twenty-one procedures run free on screen. Downloads and Copy Results stay with the subscription throughout the promotion.
Describe
- Module 1, Descriptive Statistics and Frequencies. Means, medians, quartiles, and standard deviations for continuous variables, counts and percents for categorical ones, overall or by group. Under a survey design, estimates use the design, with intervals that respect it. Free. Open in Analysis Studio
- Module 2, Crosstabs and Tests of Association. Contingency tables with Pearson chi-square, Fisher exact tests where cell counts call for them, Cramér V, odds ratios, and relative risks. Under a survey design the module uses Rao-Scott corrected tests rather than a Pearson test on weighted counts. Free. See the Chi-Square and Crosstab Calculator page. Open in Analysis Studio
- Module 4, Correlation. Pearson, Spearman, and Kendall coefficients with Fisher z intervals, partial correlations, and a matrix view with a heatmap. Free. Open in Analysis Studio
Compare and model
- Module 3, Comparing Means. One-sample, two-sample, and paired t tests, with Welch as the default for two groups, plus one-way ANOVA with Tukey-Kramer or Games-Howell follow-ups. Effect sizes come with intervals. Free. See the T-Test and ANOVA Calculator page. Open in Analysis Studio
- Module 5, Linear Regression. Multiple regression with categorical predictors, interactions, hierarchical blocks, robust and cluster-robust standard errors, and a diagnostics panel. Free. See the Linear Regression Calculator page. Open in Analysis Studio
- Module 6, Logistic Regression. Binary outcomes with odds ratios, average marginal effects, profile-likelihood intervals on request, separation detection with a Firth option, and calibration checks. Free. See the Logistic Regression Calculator page. Open in Analysis Studio
- Module 8, Rank-Based Tests. Mann-Whitney, Wilcoxon signed-rank, Kruskal-Wallis, Friedman, and their follow-ups, with effect sizes and exact p values where the conditions allow. Subscription set. See the Rank-Based Tests Calculator page. Open in Analysis Studio
- Module 14, Count, Ordinal, and Multinomial Regression. Poisson and negative binomial models for counts, proportional-odds models for ordered categories, and multinomial logistic models for unordered ones. Subscription set. See the Poisson Regression Calculator page and the Ordinal Logistic Regression Calculator page. Open in Analysis Studio
- Module 15, Mixed-Effects Models. Linear and generalized mixed models for clustered and repeated data, with random intercepts and slopes, intraclass correlations, and Kenward-Roger or Satterthwaite degrees of freedom. Subscription set. See the Mixed-Effects Models Calculator page. Open in Analysis Studio
Plan and test
- Module 7, Sample Size and Power Planner. Required sample size, achievable power, and minimum detectable effect for t tests, proportions, ANOVA, correlation, regression, chi-square tests, logistic regression, and survey margins of error, with design effects, finite population corrections, and response-rate adjustments. No data file required. Free. See the Sample Size and Power Calculator page. Open in Analysis Studio
- Module 10, A/B Test Analyzer. Randomized experiments with two or more variants, one primary metric, a sample ratio mismatch check that stops the analysis until it is acknowledged, CUPED variance reduction, and a Bayesian summary for binary metrics. Subscription set. See the A/B Test Calculator page. Open in Analysis Studio
- Module 9, Key Driver Analysis. Predictors ranked by Shapley-based importance shares of explained variance, with bootstrap intervals, an importance table, and an importance-by-performance quadrant chart. Subscription set. See the Key Driver Analysis page. Open in Analysis Studio
Survey methods
- Module 11, Survey Weighting and Raking. Base weights, nonresponse adjustment, calibration by post-stratification, raking, or GREG, and trimming with re-raking. Diagnostics report the Kish design effect, the effective sample size, and a margin check against the targets. Free. The same module carries propensity adjustment for nonprobability samples, which builds pseudo-weights against a reference probability sample and then calibrates them; that procedure sits in the subscription set. See the Survey Weighting and Raking Calculator page. Open in Analysis Studio Open Propensity Adjustment in Analysis Studio
- Module 12, Trend Analysis. Change across repeated survey waves, tested with design-based estimates, polynomial trend contrasts on the actual time values, and two-wave change tests. Free. Open in Analysis Studio
Advanced methods
- Module 13, Scale Reliability and Factor Analysis. Cronbach alpha with a confidence interval, McDonald omega, item statistics, and exploratory factor analysis with parallel analysis and oblique or orthogonal rotation. Subscription set. See the Cronbach Alpha and Scale Reliability Calculator page. Open in Analysis Studio
- Module 16, Survival Analysis. Kaplan-Meier curves with number-at-risk tables, log-rank comparisons, and Cox proportional hazards models with assumption checks. Subscription set. See the Kaplan-Meier Survival Curve Calculator page. Open in Analysis Studio
- Module 17, Time Series and Forecasting. Decomposition, trend tests, and automatic ETS and ARIMA forecasting with holdout scoring against a seasonal naive benchmark. Subscription set. See the Time Series and Forecasting Calculator page. Open in Analysis Studio
- Module 18, Propensity Score Analysis. Inverse probability weighting and matching for a binary exposure, with balance reported as standardized differences and a Love plot before results, doubly robust estimation on request, and E-value sensitivity analysis. Subscription set. See the Propensity Score Analysis page. Open in Analysis Studio
Quality and improvement
- Module 19, Statistical Quality Control and Improvement. Shewhart control charts in Phase I and Phase II with run rules and an ARL disclosure, gauge R&R studies, and process capability with normal and non-normal routes. Design of experiments, CUSUM and EWMA charts, multivariate monitoring, acceptance sampling, and Six Sigma defect and yield metrics are coming later. Free to use on screen through March 31, 2027 under the current promotion. Open in Analysis Studio
Epidemiology
- Module 20, Epidemiologic Analysis. Measures of occurrence, direct and indirect standardization, measures of association and impact with Mantel-Haenszel pooling and additive interaction, matched designs, diagnostic accuracy with ROC analysis, agreement and reliability, outbreak investigation with an epidemic curve and attack-rate table, and actuarial life-table survival. Free to use on screen through March 31, 2027 under the current promotion. Open in Analysis Studio
One module, run once, so you can see the shape of a result
Load the built-in sample data and Module 1 summarizes its value column, twelve cases, in one pass:
- n: 12
- Mean: 34.58
- Standard deviation: 17.10
- Standard error of the mean: 4.94
- 95 percent interval for the mean: 23.72 to 45.45
- Median: 35, with quartiles at 21.5 and 48.5
- Range: 10 to 60
Every module reports at that level of completeness: the estimate, its precision, and the information needed to judge it.
Open the Studio
Analysis StudioOpen Analysis Studio →The complete how-to guide, with a chapter per module and the same worked examples carried through the screens, is linked from the product page
Questions
Does Analysis Studio upload my data? No. The application runs in your browser and your data file stays on your device. The only server contact is license verification, which sends the key and nothing about your data, your variables, or your results.
Which file formats does it read? CSV, TSV, and TXT text files, SPSS files (.sav, .zsav, .por), Stata files (.dta), SAS files (.sas7bdat, .xpt), and Excel workbooks (.xlsx, .xls). Variable labels, value labels, and declared missing-value codes are preserved on import.
What does the subscription add? Two things. Downloads and Copy Results on every module, and running the ten subscription-set procedures: rank-based tests, key driver analysis, the A/B test analyzer, propensity adjustment, scale reliability and factor analysis, count and ordinal regression, mixed-effects models, survival analysis, time series and forecasting, and propensity score analysis. The price is $299 per year.
Can I try the subscription procedures before paying? Yes, during the promotion. Through March 31, 2027, every procedure runs free on screen. Downloads and Copy Results are the parts that stay with the subscription during the promotion.
Does it handle survey data properly? Survey work is a first-class part of the Studio. You declare the design once, weights with their type confirmed, strata, clusters, and replicate weights included, and every module either uses design-based methods or declines to run, with the reason stated. Weighting, raking, trend analysis, and propensity adjustment are built for survey data specifically.
How were the results checked? Every module was validated against an independent statistical environment before release, with fixed tolerance standards per method type, and end-to-end tests compare the numbers on screen, the numbers in exports, and the numbers in generated scripts against the reference.