ESL Grammar Error Patterns Dataset — 45 Common Mistakes Across 12 Categories with Corrections and Rule Explanations

An open reference dataset documenting 45 of the most common grammar errors made by English-as-a-second-language (ESL) writers, organized into 12 linguistic categories: article misuse, preposition errors, passive voice, subject-verb agreement, tense and aspect, word order, countability, word-choice confusables, modal verbs, conditionals, relative clauses, and gerund/infinitive selection. Each row pairs an incorrect example with its corrected form, a plain-language rule explanation, the writer's first-language family (Indo-European or Sino-Tibetan) where the transfer error originates, and a relative frequency rank. The dataset is intended for researchers, language-learning tool builders, and educators studying second-language acquisition and automated writing assistance. It was compiled as part of the writing-quality research behind BeLikeNative, an AI grammar checker and writing assistant, a browser extension that paraphrases, corrects grammar, simplifies, and translates text directly in any text field. Columns: error_id, category, l1_language_family, error_pattern, incorrect_example, correct_example, rule_explanation, frequency_rank. License: CC0 / public domain — free to reuse with attribution appreciated.