Kuder-Richardson 20 (KR-20)
Kuder–Richardson Formula 20 estimates the internal consistency of a test made of dichotomously scored items from the item proportions correct and the variance of total scores.
KR-20은 정답·오답(0/1)으로 채점된 문항으로 이루어진 검사의 내적 일관성을 문항 정답률과 총점 분산으로 추정하는 신뢰도 계수입니다.
What is the reliability of my 25-question true/false quiz?
25문항 OX 퀴즈의 신뢰도는 얼마인가요?
KR-20 equals k/(k − 1) times one minus the sum of item variances pq divided by the total-score variance, and it is identical to Cronbach's alpha for 0/1 items. A Feldt confidence interval conveys the precision of the estimate.
When to use Kuder-Richardson 20 (KR-20)
- All items are scored 0/1
- One administration of a test is available
- Reporting conventional reliability for an achievement or knowledge test
When to use something else
- Items have more than two score points, where Cronbach's alpha applies
- Items measure several distinct constructs, where reliability should be computed per subscale or with McDonald's omega
Assumptions and requirements
- Items are binary and scored in the same direction
- Items measure a single construct
- Errors are uncorrelated across items
What the analysis reports
- KR-20 with Feldt 95% CI
- KR-21 for comparison
- Item proportions correct and pq
- Standard error of measurement
How to report it
Internal consistency of the [value] dichotomous items was KR-20 = [value], 95% CI [[value], [value]] (N = [value]).
Related methods
Run Kuder-Richardson 20 (KR-20) on Quorant
- Data you need
- a table (CSV or Excel) with item columns (e.g. questionnaire items)
- Price
- 150 credits for one analysis up to 10,000 rows (complexity level 1); larger or wider data costs more. Optional methodologist review: 900 credits.
- Checking
- Checked by property tests in every deployment
- Before you pay
- A free pre-analysis diagnosis checks your data and gives an exact quote; nothing is deducted until you approve it.