Back to Home

One Engine.
Total Privacy.
No Hallucinations.

How it Works

Unlike Generative AI (LLMs) which "guess" the next word, APA Gen uses a Deterministic Parsing Engine. This means we wrote specific logic for every supported test. If the engine sees a T-Test output, it knows exactly where to look for the t-value, the p-value, and the degrees of freedom.

If the data is found, it is extracted exactly as is. If it's not found, the parser fails and returns an error. However, especially if the input data corresponds to unsupported test categories, or if software updates have changed the formatting output, the parser might extract nearby values. Therefore, it is always the user's responsibility to verify the exact correspondence of the results.

Note: The specific logic relies on English output keywords. Ensure your software (SPSS, JASP, etc.) is configured to display tables and results in English. We actively verify our parser against SPSS 31.0.1 and JASP 0.95.4.0, though it should also work with earlier versions.

Zero-Knowledge Privacy

Because the system runs locally, it operates entirely Client-Side. Your data never leaves your browser, ensuring absolute data sovereignty. No processing occurs on external servers. If you require an even higher level of privacy presentation, you can sanitize your input before pasting.

Sanitization Example

Since the engine is local and deterministic, it doesn't need to 'understand' your variable names to generate perfect prose. You can rename sensitive variables (e.g., 'Depression_Score' → 'Var_A') in your software before copying the output.

Detailed Input
t.test(Depression_Score ~ Drug_Treatment)
Sanitized Input
t.test(Var_A ~ Group_B)
Result: "An independent samples t-test revealed a significant difference in Var_A..."

Important: if you encounter a parsing error and decide to share your input via a Bug Report, we strongly advise you to replace sensitive variable names and values with dummy data beforehand. This allows us to debug the structure without needing to see your actual research data.

Supported Tests & Functions

R

R Statistics

Analysis TypeFunctionExtracted ParametersStatusSupport
One-Sample t-testt.test(mu = 0)
t-statisticdfp-valueMeanSD95% CI
ActiveStable
Independent Samples t-testt.test()
t-statisticdfp-valueCohen's d95% CIGroup MeansSD
ActiveStable
Paired Samples t-testt.test(paired=T)
t-statisticdfp-valueCohen's d95% CIMean DifferenceSD
ActiveStable
One-Way ANOVAaov() / summary()
F-statisticdfp-valueη² (partial)Post-hoc (Tukey)Group Means
ActiveStable
Factorial ANOVAaov() / emmeans()
F-statisticdfp-valueInteractionsη² (partial)Marginal MeansContrasts
ActiveStable
Repeated Measures ANOVArstatix::anova_test()
F-statisticdfp-valueη² (gen)Sphericity
ActiveStable
Kruskal-Wallis Testkruskal.test()
H-statisticdfp-valueη² (effect size)
ActiveStable
Wilcoxon / Mann-Whitneywilcox.test()
W/U-statisticp-valueRank Biserial
ActiveStable
Linear Regressionlm() / summary()
F-statisticAdj. R²Coefficients (β)SEt-valuesp-values
ActiveStable
Logistic Regressionglm(family=binomial)
Odds Ratiosz-valuesAICCoefficients (b)SEp-values
ActiveStable
Linear Mixed Modelslmer()
Fixed EffectsRandom Effectst-valuesp-values (lmerTest)Confidence IntervalsR² (Marginal/Cond)
ActiveStable
Pearson Correlationcor.test()
rdfp-value95% CIN
ActiveStable
Spearman Correlationcor.test(method="s")
rhop-valueN
ActiveStable
Chi-Square Testchisq.test()
dfp-valueN
ActiveStable
Descriptive Tablepsych::describe() / describeBy()
MeanSDMedianMin/MaxSkewnessKurtosisSEGrouped Stats
ActiveStable
JASP

JASP

Supported TableExtracted ParametersInstructionsStatusSupport
Descriptive StatisticsMeans, SD, Min/Max, SE, Distribution params, Split groupsRight-click analysis > Copy as LaTeX.ActiveActive
Independent Samples T-Testt, df, p, Cohen's d, SE diffRight-click > Copy as LaTeX. Check Effect Size.ActiveActive
Paired Samples T-Testt, df, p, mean diff, CIRight-click > Copy as LaTeX.ActiveActive
Chi-Square TestX², df, p-valueRight-click > Copy as LaTeX.ActiveActive
One-Way / Factorial ANOVAF, df, p, partial η²Right-click "ANOVA" > Copy as LaTeX. Check Effect Size.ActiveActive
ANCOVAF, df, p, η², η²p, Marginal Means, Simple ContrastsCopy "ANCOVA", "Marginal Means", and "Simple Contrasts" as LaTeX.
Note: Levene's test is automatically detected and handled as an assumption check.
ActiveActive
Repeated Measures ANOVAF, df, p, η², Sphericity corr.Right-click "Within Subjects" > Copy as LaTeX.ActiveActive
Correlation MatrixPearson/Spearman coeff, p-valuesRight-click > Copy as LaTeX.ActiveActive
Linear RegressionCoefficients (B, β, t, p), R², F, Model Fit, Scientific Notation supportCopy "Model Summary" & "Coefficients" as LaTeX. Supports scientific notation (e.g., 10⁻⁴).ActiveActive
Generalized Linear Mixed ModelFixed Effects, Est. Means, ContrastsCopy tables as LaTeX.ActiveActive
SPSS

SPSS

Required: Export your SPSS Output as Text - UTF8 (*.txt). Go to File → Export, select "Text - UTF8 (*.txt)", then upload the file. See the User Manual for step-by-step instructions.

Supported TableInstructionsStatusSupport
Descriptive Statistics / FrequenciesFile → Export → Text (*.txt). Upload the content using the dashboard button.ActiveStable
Independent Samples t-testFile → Export → Text (*.txt). Include "Group Statistics" and "Independent Samples Test" tables.ActiveStable
Paired Samples t-testFile → Export → Text (*.txt). Include "Paired Samples Statistics" and "Paired Samples Test" tables.ActiveStable
One-Sample t-testFile → Export → Text (*.txt). Include "One-Sample Statistics" and "One-Sample Test" tables.ActiveStable
One-Way ANOVAFile → Export → Text (*.txt). Export the "ANOVA" table results.ActiveStable
Factorial ANOVA / ANCOVAFile → Export → Text (*.txt). "Tests of Between-Subjects Effects" table (Univariate GLM).ActiveStable
Repeated Measures ANOVAFile → Export → Text (*.txt). "Tests of Within-Subjects Effects" table.ActiveStable
Linear RegressionFile → Export → Text (*.txt). Include "Coefficients" and "Model Summary".ActiveStable
Logistic RegressionFile → Export → Text (*.txt). "Variables in the Equation" and "Model Summary" tables.ActiveStable
Mixed Models (LMM)File → Export → Text (*.txt). "Type III Tests of Fixed Effects" table.ActiveStable
Non-Parametric TestsFile → Export → Text (*.txt). Mann-Whitney, Wilcoxon, Kruskal-Wallis, Chi-Square, Pearson/Spearman Correlations.ActiveStable

Looking for step-by-step instructions? Read the User Manual

Need help? Contact Support