Lawshe CVR & Critical Value Calculator
Enter the expert-panel size and the number who rated one item essential. The calculator reports the Lawshe CVR and the exact-binomial agreement threshold for the selected significance level.
Content Validity Ratio0.8
- Exact one-tailed probability
- 1.07%
- Critical CVR
- 0.8
- Minimum essential ratings
- 9
- Votes above or below threshold
- 0
- Criterion result
- Meets the selected criterion with 0 votes above the minimum.
Uses an exact one-tailed binomial calculation with chance probability fixed at 0.5.
The threshold is panel-size specific, so the same CVR can pass for one panel size and fail for another.
How to use this calculator
- Enter the total number of subject-matter experts who rated the item.
- Enter how many experts marked the item as essential.
- Choose the one-tailed significance level you want to use for the decision rule.
- Compare the observed CVR and essential-rating count with the critical values.
How the Lawshe CVR is calculated
Lawshe's content validity ratio summarizes how strongly an expert panel agrees that one item is essential. Let N be the number of panelists and E be the number who rated the item essential. The observed ratio is:
CVR = (E - N / 2) / (N / 2) = 2E / N - 1
The CVR ranges from -1 to 1. A value of -1 means no expert rated the item essential. A value of 0 means exactly half did. A value of 1 means unanimous essential ratings.
How the critical value is found
The critical-value part asks a separate question: if each expert were no more likely than chance to call the item essential, how unusual would this much agreement be? The calculator uses an exact one-tailed binomial model with chance probability 0.5:
p(E, N) = sum from j = E to N of C(N, j) x 0.5^N
For the selected alpha, the calculator searches the whole-number essential-rating counts and finds the smallest K whose upper-tail probability is no greater than alpha. It then converts that vote count back to a CVR:
critical CVR = 2K / N - 1
An item reaches the selected criterion when E >= K. Equality counts, because the threshold is defined as the smallest vote count that satisfies the exact-binomial rule.
What moves the result
The expert count matters because the vote count is discrete. With a small panel, one additional essential rating can move the CVR a large distance, and for very small panels even unanimous agreement may not reach a strict alpha. With larger panels, the critical CVR can be lower because a moderate proportion above one-half may already be unlikely under the chance model.
The selected significance level changes the minimum agreement required. A 10% rule is less strict than a 5% rule, and a 1% rule often requires several more essential ratings. The observed CVR itself does not change when alpha changes; only the decision threshold does.
What this calculator leaves out
This calculation evaluates one item's essentiality ratings. It does not judge the full assessment, the quality of the expert panel, the clarity of the rating prompt, the coverage of the content domain, or whether qualitative comments support keeping the item. It also does not adjust for testing many items at once, so a separate multiple-comparison procedure may be appropriate in a large item review.
Worked example
Suppose 10 experts review an item and 9 rate it essential. The observed ratio is 2 x 9 / 10 - 1 = 0.80. The exact one-tailed probability is (C(10, 9) + C(10, 10)) / 2^10 = 11 / 1024 = 0.010742, or 1.07%.
At alpha 0.05, 8 essential ratings are not enough because 56 / 1024 = 0.054688. The minimum count is 9, so the critical CVR is 2 x 9 / 10 - 1 = 0.80. The item meets the criterion exactly, with zero votes above the minimum.
Common questions
What is Lawshe's content validity ratio?
Lawshe's CVR is a single-item index of how many experts judged an item essential. It rescales the essential-rating count so that -1 means no agreement, 0 means half the panel rated the item essential, and 1 means unanimous essential ratings.
What CVR is acceptable for my number of experts?
The acceptable CVR depends on the panel size and the decision rule. This calculator uses the exact-binomial approach associated with Ayre and Scally, then reports the smallest CVR attainable for your panel size at the selected alpha.
How many experts must rate an item essential?
Use the minimum essential ratings row. It is the smallest whole-number count whose exact one-tailed binomial probability is at or below the selected significance level. If your observed count equals or exceeds that number, the item reaches the selected criterion.
Why does the critical CVR change with panel size?
CVR is based on whole people, not a continuous measurement. Adding one essential rating has a larger effect in a panel of 5 than in a panel of 50, so the exact-binomial cutoff changes as the panel size changes.
What is the difference between CVR and a content validity index?
CVR usually refers to the essentiality decision for one item. A content validity index often summarizes relevance ratings across items or experts using a different rating scale and decision rule. The two statistics answer related but not identical questions.
Should an item below the critical CVR always be removed?
Not automatically. A low CVR is evidence that the expert panel did not show enough essentiality agreement under the selected rule, but item decisions should also consider comments, blueprint coverage, item quality, and the consequences of removing the content.