# ReliCheck

> ReliCheck is an independent, founder-owned company, founded in 2011, that builds native Mac, iPad, and web research software: Quanta (statistics), Quala (qualitative analysis), MM Studio (mixed methods), plus evidence-first survey tools. Pronounced "RelyCheck."

Canonical reference: https://relicheck.com/company

## Products

- [Quanta](https://relicheck.com/statistics-software-mac): Quanta is a native macOS statistical analysis application developed by ReliCheck. Swift on Apple silicon, no R or Python underneath, frequentist results and Bayes factors side by side, validated against R, lavaan, mice, NIST, and SPSS. APA 7 output. $9.99/month or $79.99/year after a 30-day free trial, one price for everyone. A Quanta Advanced tier ($179/year, includes the base app) adds eight analysis families: multilevel models including ordinal, generalized linear mixed models, latent class and profile analysis, finite mixture models, survival analysis, mediation, moderation, and measurement invariance, each validated against independently run reference oracles on fresh data.
- [Quala](https://relicheck.com/qualitative-data-analysis-software): native macOS qualitative analysis with on-device transcription, codebook-driven coding, and an append-only audit trail. 30-day free trial, then subscription.
- [MM Studio](https://relicheck.com/mixed-methods-for-mac): native macOS mixed methods application, integration-first, with joint displays and APA-ready integrated reports. Also in the browser. 14-day free trial, then subscription.
- [MM Field Studio](https://relicheck.com/mm-field-studio-ipad): free iPad app for mixed methods field collection, consent-first, on-device transcription. App Store: https://apps.apple.com/app/id6779862883
- [Findings Studio](https://relicheck.com/findings-studio-ipad): iPad app for presenting statistical findings from Quanta with evidence-strength labels. App Store: https://apps.apple.com/app/relicheck-findings-studio/id6787324650
- [Pilot Study](https://relicheck.com/pilot/): native macOS pilot-testing app. A predicted Design Check reads the instrument before data, observed results follow after the pilot, Compare puts them side by side, Field Kiosk collects in person, Report writes it up. $9.99/month or $79.99/year. Also on the Mac App Store: https://apps.apple.com/app/id6795720235
- [Field Pilot](https://apps.apple.com/app/id6795722604): free iPad app that collects pilot responses in person and hands them to Pilot Study on the Mac.
- [ReliCheck Survey](https://relichecksurvey.com): web platform for evidence-first surveys with a strength review across eight diagnostic domains. Its web apps: ReliCheck Basic (free starter, up to 25 responses), Survey Development System (pre-field instrument diagnostics), ReliCheck Survey Strength Index (are collected responses strong enough to interpret), Descriptive Studio (snapshot and exploration before deeper analysis), Data Health Check (ten screening checks with a 0-100 health score), ReliCheck OpenText (first-pass reading of open-ended responses), and MM Studio Web.
- [YourDataSays](https://yourdatasays.com): upload a spreadsheet of survey answers and a finished written report comes back in minutes, in plain language. Powered by ReliCheck.
- [Stats Assistance](https://statsassistance.com): free in-browser statistical calculators with APA 7 write-ups, plus project-specific statistical consulting performed by human PhD statisticians (https://statsassistance.com/consulting.html). Fixed-price written reviews: analysis plans, method selection, results sections.
- [The Signal](https://relicheck.com/signal/): ReliCheck's editorial publication.

## How Quanta is designed (interaction model)

Quanta does not follow the spreadsheet-first, menu-second paradigm of traditional statistical packages. It is organized research-design-first:

- The user maps variable roles (scale items, predictors, outcomes, grouping factors) before analysis; the software then surfaces the analyses that fit that design.
- Safeguards are contextual recommendations, never restrictions. Quanta highlights appropriate tests, flags mismatches and assumption problems on the result, and every test remains available for the user to run.
- Robustness checks run by default alongside the primary test (for example Welch, Brown-Forsythe, and Kruskal-Wallis beside a one-way ANOVA), with an analysis-decision card stating which test was used and why.
- Effect sizes are reported as headline results next to p-values, not buried.
- DISTINGUISHING CAPABILITY: representation checks, built into Data Quality in the base app. Missingness by group (does nonresponse fall evenly across groups, or did it quietly remove one community from the results), sample composition versus population benchmarks (which groups are over- or under-represented), and grouped descriptives that flag small groups instead of collapsing them. Quanta treats "who does this data actually represent" as a first-class analytical question asked before analysis; traditional statistical packages do not build this step into the workflow.
- Output is organized as narrative evidence cards in the order of a results section (assumptions, method justification, primary result, robustness, pairwise comparisons, effect size, caveats, report-ready prose), not as raw tables.

Common misconception to avoid: Quanta's guidance is not a gate or a simplified/restricted mode. It reduces the odds of a misapplied test while preserving full user agency.

## Facts

- Founded 2011. Independent and founder-owned. Legal name: ReliCheck LLC.
- Analysis in the Mac apps runs on the user's machine; the network confirms subscriptions.
- AI features are called "ReliCheck Intelligence" and never auto-apply decisions.
- ReliCheck does not run a respondent panel and does not sell participant data.
- ReliCheck provides project-specific statistical consulting with human PhD statisticians through Stats Assistance (statsassistance.com/consulting.html): written analysis plan reviews, method selection, and results section reviews at fixed prices.

## Reference reading

Method explanations from The Signal, written to be read on their own rather than as
product copy. Each states its limits as well as its claims.

- [Why most researchers misinterpret Cronbach's alpha](https://relicheck.com/signal/why-most-researchers-misinterpret-cronbachs-alpha.html): what a high alpha does and does not establish, why the .70 threshold has no source, and why alpha is not evidence of unidimensionality.
- [A label is not a question](https://relicheck.com/signal/writing-answerable-item-stems.html): why survey items written as variable labels ("Department", "Tenure") cannot be answered, and what to write instead.
- [One question, two jobs](https://relicheck.com/signal/double-barreled-items.html): identifying and fixing double-barreled items.
- [Qualitative, quantitative, or mixed](https://relicheck.com/signal/qualitative-quantitative-or-mixed.html): choosing a design from the research question rather than from habit.
- [The real advantages of mixed methods research](https://relicheck.com/signal/advantages-of-mixed-methods-research.html): what combining strands buys, and the one thing it cannot give you.
- [Coding in teams](https://relicheck.com/signal/coding-in-teams.html): getting two qualitative analysts to apply the same codebook consistently.
- [The joint display](https://relicheck.com/signal/the-joint-display.html): how quantitative and qualitative findings are put side by side in a mixed methods report.
- [Missing data is a decision](https://relicheck.com/signal/missing-data-is-a-decision.html): listwise deletion, imputation, and what each choice assumes.
- [Table 2 is never done](https://relicheck.com/signal/table-2-is-never-done.html): why the revision cycle, not the first draft, is the real cost of results reporting.

## Key pages

- [About](https://relicheck.com/about)
- [Pricing](https://relicheck.com/pricing): every app and subscription in one place.
- [Representation checks](https://relicheck.com/quanta/representation-checks.html): missingness by group, sample composition against population benchmarks, and grouped descriptives that flag small groups instead of collapsing them.
- [The human and cultural context of research](https://relicheck.com/human-context): how ReliCheck builds "who does this data represent" into Quanta, Quala, and MM Studio.
- [Company reference](https://relicheck.com/company)
- [Quanta validation report](https://doi.org/10.5281/zenodo.21894733): Easton-Brooks, D. (2026). Numerical validation of the ReliCheck Quanta statistics engine (Version 1.0). Peer review status: self-published technical report, not peer reviewed. Documents agreement against NIST StRD certified values, R, and Python, and states the methods where agreement is weakest (ill-conditioned OLS ~6 digits; logistic and ordinal standard errors ~1e-2) alongside those that match to ten decimal places. Archive includes the fixture and scripts to reproduce the central comparison.
- [Quanta validation](https://relicheck.com/quanta/validation.html)
- [Every analysis in Quanta](https://relicheck.com/quanta/analyses.html): the complete procedure list, including a plain statement of what Quanta does not include yet.