Search for a command to run...
This dataset provides readability and comprehension metrics for 4 technology guides translated into 5 languages (English, Spanish, French, Portuguese, and Italian), yielding 20 guide-language pairs. Each record includes word count, sentence count, average sentence length, average word length, and six standardized readability indices: Flesch-Kincaid Grade Level, Flesch Reading Ease, Gunning Fog Index, Coleman-Liau Index, SMOG Index, and Dale-Chall Score. Readability formulas were originally developed for English text, and their cross-linguistic validity remains an active area of research. This dataset contributes empirical evidence by applying identical formulas to parallel texts (same content translated by professional translators) across five Romance and Germanic languages. The results reveal systematic cross-linguistic patterns: French translations consistently score 1.5-2.5 grade levels higher than their English source texts, while Portuguese translations remain closest to English readability scores. These differences arise from structural linguistic features (average word length, sentence complexity conventions) rather than content difficulty. Beyond mechanical readability metrics, the dataset includes user-facing outcomes: completion rates (percentage of users who successfully followed all steps), average time to complete, and user satisfaction ratings on a 5-point scale. These empirical measures enable validation of readability formulas against actual comprehension performance in a senior user population. The data shows moderate but imperfect correlation between Flesch Reading Ease and completion rates (r=0.72), suggesting that readability formulas capture approximately half the variance in real-world guide usability. The guides cover four common technology tasks: email setup, video calling, online security, and navigation apps. Each was designed following plain language principles with a target reading level of grade 6-8, appropriate for the senior audience served by Facil.guide.