Text analysis
for serious researchers.

TASS is a no-code analytical environment for dictionary-based text analysis. Group comparisons with assumption checks and post-hoc tests, publication-quality visualizations, and APA citation blocks. Built for social scientists, journalists, and domain experts who need rigorous methods without learning to code.

For decades, LIWC has been the default tool for dictionary-based text analysis in the social sciences. The default has compounding limitations. It cannot handle large datasets. Its dictionaries are opaque and proprietary. It produces no built-in statistical comparisons or visualizations. Its academic license now explicitly excludes anyone outside a university setting, which leaves researchers at think tanks, government agencies, nonprofits, and independent consultancies with no clean alternative that does not require R or Python.

The free alternatives, Yoshikoder and psyLex and LIWCalike, all require coding. No no-code tool offers group statistical comparisons, effect sizes, and publication-quality visualizations in one package.

TASS is that tool. It supports 100,000+ entry datasets, ships with six open dictionaries at v1.0, includes 18 T-Lex lexicons in development, and requires zero lines of code.

TASS handles the entire analysis cycle in a native desktop environment. Data Import accepts CSV, TXT, and XLSX with batch processing; a column-mapping wizard lets you designate text, group, and metadata columns, and rows are previewed read-only before analysis begins. The Analysis Engine performs dictionary scoring at word, entry, and document level, with multi-threaded processing across all CPU cores, a real-time progress display, and time estimates for large corpora.

Group Comparisons are defined by column values and produce Welch’s t-test, Mann-Whitney U, one-way ANOVA, Kruskal-Wallis, Tukey HSD post-hoc tests, Shapiro-Wilk normality checks, effect sizes with interpretation labels, and 95% confidence intervals, computed per category, ready to report. Export & Citation produces CSV at the entry level, multi-sheet Excel workbooks, and 300 DPI PNG + SVG visualizations. Every export includes a pre-formatted APA citation block for your methods section.

TASS ships with six open-source dictionaries at v1.0 and is building eighteen original T-Lex lexicons. Each lexicon is independently citable and freely available for academic and non-commercial use, with full documentation and version control.

v1.0, May 2026. Complete analytical environment: dictionary scoring, n-gram support, structured group comparisons with assumption checks and post-hoc tests, correlation matrix, KWIC concordance, seven visualization types, APA table export. Six open dictionaries at launch.

v1.1, Q3 2026. Paired-samples t-test, factorial ANOVA, macOS build, LIWC .dic format import, concordance plot. T-Lex dictionaries 01 through 07 ship.

v2.0, 2027. Multilingual dictionary support, topic modeling (LDA), readability metrics, inter-rater reliability module, API and CLI for pipeline integration. Full eighteen-dictionary T-Lex library complete.

The T-Lex Project is building eighteen original, open-access lexicons for social science text analysis. Each is a citable, independently documented resource designed to fill real measurement gaps in the field. T-Lex dictionaries ship with TASS and are freely available to the research community under open-access licensing.

Collaboration is asynchronous and self-paced. Contributors review and validate AI-generated word lists against published measurement frameworks, re-score or re-categorize items that require domain expertise, supplement seed lists with terms from their specialization, and provide feedback on category definitions and construct validity.

We are looking for social scientists with domain expertise in one or more of the target constructs, graduate students and early-career researchers interested in computational methods and lexicon development, and anyone curious about how LLMs are used in modern corpus construction and validation. Time commitment is variable; you may contribute to a single dictionary or several. No minimum hours are required, and credit is proportional to contribution.

Contributors who make substantive contributions are offered co-authorship on associated dictionary papers and are named in all published dictionary metadata. All TASS-native dictionaries remain freely available for academic and non-commercial research use. Email tass@simdadllc.com to discuss involvement.