Reliability · edition 2026-10-23
IPIP-NEO-120 reliability: alpha for 5 domains and 30 facets
Cronbach’s alpha tells you whether the items on a scale agree with each other enough to be added into one score. These are the values for the IPIP-NEO-120, computed on 3,079 completed results.
Domains
Alpha runs from .797 (Openness) to .903 (Neuroticism). Mean .868.
Facets
Median alpha .774. Range .565 (Liberalism) to .880 (Cautiousness).
Measurement error
The standard error of measurement of a domain score is between 5.1 and 5.5 points on the 0–100 scale.
Five domains, 24 items each
Each domain score adds up 24 items. The chart shows alpha with its 95% confidence interval; the table gives the figures.
| Domain | Alpha | 95% CI | Alpha after screening | SEM (points) |
|---|---|---|---|---|
| Openness | .797 | .785 – .807 | .791 | 5.5 |
| Conscientiousness | .898 | .893 – .903 | .899 | 5.1 |
| Extraversion | .881 | .875 – .888 | .879 | 5.4 |
| Agreeableness | .863 | .855 – .870 | .859 | 5.2 |
| Neuroticism | .903 | .898 – .909 | .904 | 5.4 |
| Mean across domains | .868 | — | .866 | — |
- Alpha and its interval are for all 3,079 results. Intervals are bootstrap percentile intervals over 300 resamples of the respondents with a fixed seed, so they reproduce exactly on a re-run.
- Alpha after screening is alpha once the 257 results flagged by the answer-pattern rules (below) are set aside.
- SEM is the standard error of measurement on the 0–100 score metric, SD × √(1 − α). With an SEM of 5.1, a Conscientiousness score of 60 sits within about 55 to 65 at 68% and about 50 to 70 at 95%.
Thirty facets, 4 items each
A four-item scale cannot be as consistent as a twenty-four-item one, because alpha rises with the number of items. Facet alphas below the domain figures are therefore expected. 27 of the 30 facets are at .70 or above; Liberalism (.565) is the lowest.
| Facet | Domain | Alpha | Mean (0–16) | SD |
|---|---|---|---|---|
| O1 Imagination | Openness | .806 | 11.6 | 3.6 |
| O2 Artistic Interests | Openness | .740 | 11.0 | 3.5 |
| O3 Emotionality | Openness | .660 | 10.9 | 3.4 |
| O4 Adventurousness | Openness | .721 | 8.9 | 3.5 |
| O5 Intellect | Openness | .749 | 11.6 | 3.4 |
| O6 Liberalism | Openness | .565 | 8.6 | 3.3 |
| C1 Self-Efficacy | Conscientiousness | .845 | 11.2 | 3.3 |
| C2 Orderliness | Conscientiousness | .868 | 8.6 | 4.6 |
| C3 Dutifulness | Conscientiousness | .670 | 11.6 | 2.9 |
| C4 Achievement-Striving | Conscientiousness | .771 | 11.0 | 3.5 |
| C5 Self-Discipline | Conscientiousness | .773 | 8.8 | 3.7 |
| C6 Cautiousness | Conscientiousness | .880 | 10.2 | 4.3 |
| E1 Friendliness | Extraversion | .833 | 8.4 | 4.1 |
| E2 Gregariousness | Extraversion | .804 | 6.0 | 4.1 |
| E3 Assertiveness | Extraversion | .819 | 9.7 | 3.7 |
| E4 Activity Level | Extraversion | .700 | 8.0 | 3.5 |
| E5 Excitement-Seeking | Extraversion | .751 | 7.8 | 3.7 |
| E6 Cheerfulness | Extraversion | .809 | 10.1 | 3.6 |
| A1 Trust | Agreeableness | .852 | 8.2 | 3.7 |
| A2 Morality | Agreeableness | .829 | 12.3 | 3.5 |
| A3 Altruism | Agreeableness | .765 | 11.4 | 3.3 |
| A4 Cooperation | Agreeableness | .728 | 12.2 | 3.5 |
| A5 Modesty | Agreeableness | .764 | 8.6 | 3.9 |
| A6 Sympathy | Agreeableness | .775 | 10.4 | 3.6 |
| N1 Anxiety | Neuroticism | .833 | 9.0 | 4.1 |
| N2 Anger | Neuroticism | .829 | 7.0 | 4.0 |
| N3 Depression | Neuroticism | .817 | 6.3 | 4.0 |
| N4 Self-Consciousness | Neuroticism | .752 | 8.1 | 4.0 |
| N5 Immoderation | Neuroticism | .746 | 7.6 | 3.7 |
| N6 Vulnerability | Neuroticism | .804 | 6.1 | 3.9 |
How the five domains relate
Correlations between the five domain scores, computed on all 3,079 results. The two largest are negative: Conscientiousness with Neuroticism (−.52) and Extraversion with Neuroticism (−.49). No pair is large enough to mean that two scales measure the same thing.
| Domain | Openness | Conscientiousness | Extraversion | Agreeableness | Neuroticism |
|---|---|---|---|---|---|
| Openness | — | .07 | .26 | .23 | −.12 |
| Conscientiousness | .07 | — | .30 | .26 | −.52 |
| Extraversion | .26 | .30 | — | .14 | −.49 |
| Agreeableness | .23 | .26 | .14 | — | −.11 |
| Neuroticism | −.12 | −.52 | −.49 | −.11 | — |
Alpha by language
The same 120 items and the same scoring key are used in every language, so domain alphas can be compared across languages. A language is listed when at least 100 results are available. Mean alpha runs from .844 to .873 across the five.
| Language | n | Openness | Conscientiousness | Extraversion | Agreeableness | Neuroticism | Mean alpha |
|---|---|---|---|---|---|---|---|
| English | 1,558 | .800 | .908 | .882 | .869 | .905 | .873 |
| Spanish | 357 | .805 | .886 | .886 | .860 | .896 | .867 |
| French | 192 | .793 | .890 | .843 | .839 | .897 | .852 |
| Chinese | 188 | .774 | .873 | .842 | .880 | .925 | .859 |
| Japanese | 157 | .749 | .871 | .910 | .813 | .879 | .844 |
Answer-pattern screening
Three rules flag results with a long run of identical responses, almost no variation between answers, or a completion time that is too short. They flag 257 of 3,079 results (8.3%): 218 for identical runs, 7 for almost no variation, 62 for speed. A result can trip more than one rule. Nobody is removed from the main table; the screened column repeats every alpha without the flagged results.
The median completion time is 8 min 10 s, and the middle 80% of respondents finish between 5 min 19 s and 16 min 37 s. The median longest run of identical answers is 6 items; at the 95th percentile it is 10.
Sample and method
The analysis set is every completed IPIP-NEO-120 result held on 2026-10-23 that passes the integrity checks: 3,214 rows read, 3,079 analysed. Each result is re-scored from its 120 answers, and any result whose stored domain scores differ from the fresh scoring is left out, as is a repeat submission of the same answers within two hours. Results routed through an employer or applicant-tracking invitation, and results taken through a school class link, are not part of the sample.
The sample is people who chose to take a personality test on a consumer website. Alpha is computed from the item covariance matrix.