Resources
A university institutional review board (IRB) audits an international multi-center clinical trial evaluating in vivo AAV-delivered CRISPR-Cas9 base editors for familial hypercholesterolemia. Across three demographic trial tiers, researchers surveyed enrolled patients regarding their subjective comprehension of germline versus somatic modification risks during the informed consent process. The observed counts of patients demonstrating 'Adequate Comprehension' versus 'Inadequate Comprehension' were recorded:
Using a Chi-Square test of independence at significance level : (a) Calculate the expected count for Tier 1 with Adequate Comprehension and for Tier 3 with Inadequate Comprehension. (b) Compute the test statistic across all cells, identify the degrees of freedom , and determine whether researchers reject or fail to reject the null hypothesis that informed consent comprehension is independent of healthcare trial tier (Critical value for : ).
Answer and steps
Final answer
(a) , ; (b) ; with , since (), we reject the null hypothesis and conclude that comprehension levels differ significantly across trial tiers.
Steps
Part (a): Calculating Expected Frequencies
Under the null hypothesis (), comprehension level is independent of the healthcare tier. The expected frequency for any cell in row and column is calculated using the formula:
- Expected frequency for Tier 1 with Adequate Comprehension ():
- Expected frequency for Tier 3 with Inadequate Comprehension ():
(Note: Because all three tiers have equal row totals of , the expected count for Adequate Comprehension is and for Inadequate Comprehension is across all tiers).
Part (b): Computing the Chi-Square Statistic and Hypothesis Testing
The Chi-Square test statistic measures the divergence between observed counts () and expected counts ():
Evaluating each of the six cells:
- Tier 1, Adequate:
- Tier 1, Inadequate:
- Tier 2, Adequate:
- Tier 2, Inadequate:
- Tier 3, Adequate:
- Tier 3, Inadequate:
Summing these individual cell contributions:
Degrees of Freedom and Decision Rule:
- Degrees of freedom ():
- Comparison: At significance level , the critical value is .
- Conclusion: Because the calculated test statistic is significantly greater than (), we reject the null hypothesis .
This demonstrates a statistically significant disparity in informed consent comprehension among patients across different healthcare tiers, highlighting key bioethical concerns regarding health literacy and equitable access during gene-editing clinical trials.
Typical errors
01 The student divides by the observed frequency instead of the expected frequency in the chi-square denominator, yielding an incorrect test statistic.
Sample Student Work
Reviewed by ACEpath TeacherStep-by-Step Review
Walk Through It with Your Teacher- Step 1
- Step 2Correct
You placed the observed counts () in the denominator of each cell term instead of dividing by the expected counts ().
- Step 3Correct
; since , reject .
You determined the degrees of freedom and the decision rule correctly, but your comparison used the test statistic computed with the incorrect denominator.
WHAT YOU DID WELL
Compliments from Your TeacherYou correctly stated and calculated the expected counts using row and column totals.
You found the correct squared deviations for all six table cells.
You calculated the degrees of freedom accurately and identified the correct rejection criterion.
THE ERROR
Where Your ACEpath Teacher Spotted the MistakeYou divided each squared difference by the observed count rather than the expected count .
WHY IT HAPPENED
ACEpath Teacher's DiagnosisYou confused the baseline reference in Pearson's chi-square test statistic. Under the null hypothesis, deviations are standardized relative to what you expect under independence (), not what was empirically observed ().
REASONING PATH
Your ACEpath teacher’s suggested approach- Denominator Standardization
Pearson's chi-square statistic standardizes squared discrepancies against expected counts: .
Whenever you set up each term in a chi-square test, check that the value in the denominator matches the null expectation , never the sample count .
CORRECTED PATH
ACEpath Teacher's DemonstrationSet up each cell term dividing by the expected count: .
Evaluate the fractions: .
Compare with the critical threshold: , so you still reject at .
(a) , ; (b) , ; since , reject .
02 The student calculates degrees of freedom as (r × c) - 1 instead of (r - 1)(c - 1), misidentifying df as 5.
Sample Student Work
Reviewed by ACEpath TeacherStep-by-Step Review
Walk Through It with Your Teacher- Step 1
- Step 2
- Step 3Correct
For a two-way contingency table, degrees of freedom are constrained by both row and column marginal totals, giving . The formula applies only to a one-way goodness-of-fit test.
- Step 4
, reject
WHAT YOU DID WELL
Compliments from Your TeacherYou correctly calculated the expected counts using the marginal totals and grand total.
You grouped and summed the cell chi-square components accurately to find .
You correctly rejected the null hypothesis based on the critical value comparison.
THE ERROR
Where Your ACEpath Teacher Spotted the MistakeNo, this represents a conceptual confusion between the degrees of freedom for a one-way test vs. a two-way contingency table.
WHY IT HAPPENED
ACEpath Teacher's DiagnosisYou treated the contingency table as independent categories with only one grand total constraint (giving ), overlooking that in a test of independence, every row sum and column sum is fixed, which restricts independent cell choices to .
REASONING PATH
Your ACEpath teacher’s suggested approach- Degrees of Freedom in Contingency Tables
A two-way table with rows and columns has .
Count rows and columns , subtract from each dimension, and multiply the differences.
CORRECTED PATH
ACEpath Teacher's DemonstrationIdentify the number of rows (trial tiers) and columns (comprehension levels).
Apply the two-way contingency table formula: .
Compare the calculated against to conclude rejection of .
(a) , ; (b) , , reject .
03 The student omits squaring the differences in the numerator, summing (O - E) / E instead of (O - E)² / E.
Sample Student Work
Reviewed by ACEpath TeacherStep-by-Step Review
Walk Through It with Your Teacher- Step 1
- Step 2Correct
You omitted the squaring of the numerator difference . In the chi-square formula, each deviation must be squared as so that discrepancies do not cancel out algebraically to zero.
- Step 3
- Step 4
, fail to reject
CorrectSince , reject
Because your calculated test statistic was zero due to the missing square in Step 2, your statistical decision contradicted the true result. With , you must reject the null hypothesis.
WHAT YOU DID WELL
Compliments from Your TeacherYou calculated the expected values for the contingency table accurately.
You determined the degrees of freedom using the correct formula .
THE ERROR
Where Your ACEpath Teacher Spotted the MistakeYou wrote the numerator as rather than , forgetting the exponent.
WHY IT HAPPENED
ACEpath Teacher's DiagnosisA misconception or oversight regarding how distance/discrepancy is measured in goodness-of-fit and independence tests: raw differences naturally sum to zero across marginal totals, requiring the deviations to be squared to capture absolute discrepancy.
REASONING PATH
Your ACEpath teacher’s suggested approach- Chi-Square Test Statistic Formula
The chi-square test statistic is always , never .
Always verify that every individual deviation is squared before dividing by .
- Sanity Check on Deviations
Raw deviations sum to across rows and columns by definition.
If your cell calculations cancel out to despite clear observable differences, check whether you forgot to square .
CORRECTED PATH
ACEpath Teacher's DemonstrationSquare the deviation for each cell: , , , , , and .
Sum all cell values: .
Compare the test statistic to the critical value: since , reject .
(a) , ; (b) , , reject .
Lock it in
What is the formula for Pearson's Chi-Square () test statistic, and what value belongs in the denominator?
. The denominator is always the expected frequency (), because discrepancies are standardized against the null model.
Similar question
In a clinical trial survey, 80 patients reported severe adverse events while 100 were expected under . Calculate the individual chi-square component for this cell.
Answer
.
A study evaluates consent comprehension between two clinical trial sites. The observed counts are Site A: 60 adequate, 40 inadequate; Site B: 40 adequate, 60 inadequate. Both sites enroll 100 participants (total , overall 100 adequate, 100 inadequate). Calculate the expected count for each cell and find the total test statistic .
Answer
Each expected count is . The test statistic is .
An IRB reviews genetic trial compliance across three demographic tiers () and three outcome categories (). A researcher calculates . Find the degrees of freedom . If the critical value at is , state whether to reject or fail to reject .
Answer
. Since the test statistic , reject the null hypothesis .