IPIP-NEO-120 · research data by language · ro
https://bigfive-test.com/research/ro

DATA BASIS
Calculated: 2026-10-24T05:30:04.281Z
Results on this version since 23 October 2026: 5
Valid: 5
In the analysis sample: 5
English analysis sample: 1364

METHOD
Each day, up to 20000 of the latest valid 120-item results across all languages are used. The sample is grouped by the actual question language. Only complete responses to the current translation and item bank are included, so the analysis sample may be smaller than the number of valid results.
Quality filters exclude development and test responses, very fast responses and responses with very little variation. Repeat attempts may be included. Visitors choose to take the test; this is not a representative population sample or language-specific norms.
Item statistics are shown from 30 responses, the comparison with English from 50 in both samples. These are display thresholds, not validation of the translation. The comparison is calculated from rounded item means (two decimals) and is not a test of measurement invariance; differences may reflect wording, culture or who took the test. The English sample consists of other people, not the same respondents in two languages.
Only aggregate statistics are published. Individual responses and background details are not shared.

COLUMNS IN items.csv (UTF-8 with BOM, comma separated, period as decimal mark)
language: Language code the questions were shown in (BCP 47, lower case).
translation_version: Hash of this translation's wording. When the wording changes, statistics start over.
items_version: Hash of the item bank (order and scoring key).
generated_at: Time of the calculation (UTC, ISO 8601).
n_language: Number of complete, valid responses in this language in the analysis sample. Applies to every row.
n_english: The same count for the English analysis sample, which the comparison uses.
item_id: Stable ID of the statement in the item bank.
num: The statement's position in the test, 1–120.
domain: Trait: O, C, E, A or N.
domain_name: Trait name.
facet: Facet number within the trait, 1–6.
facet_name: Facet name.
keyed: plus: "very accurate" scores 5. minus: reverse keyed, "very accurate" scores 1.
text: The statement as shown in this language.
text_english: The English original (Johnson, 2014).
mean: Mean scored response, 1–5, with reverse-keyed items reversed. Empty until the sample reaches 30.
r: Corrected item–facet correlation: Pearson's r between the statement and the sum of the other three in its facet. Empty when undefined.
r_ci_low: Lower bound of the 95% confidence interval for r (Fisher's z, n = n_language).
r_ci_high: Upper bound of the same interval.
english_mean: Mean for the same statement in the English sample.
diff: mean − english_mean.
facet_diff: The mean of diff across the facet's four statements.
beyond: diff − facet_diff: how much the statement differs beyond its facet. Empty for English and until both samples reach 50.

Statements are from the IPIP-NEO-120 (Johnson, 2014), which is in the public domain (IPIP). Translations are credited on the language page.