McNemar’s Test
McNemar’s Test is a statistical method used to analyze paired données nominales appariées, specifically when you want to evaluate the differences in responses from the same subjects under two different conditions. This test is particularly useful in situations where you have a binary outcome (yes/no, success/failure) measured before and after an intervention or treatment.
The primary purpose of McNemar’s Test is to determine if there are significant changes in the proportions of categorical responses. It is often applied in clinical trials, surveys, and other studies where participants are assessed twice. For example, it can be used to analyze whether patients’ conditions improved after a certain treatment.
To perform McNemar’s Test, you typically create a 2×2 tableau de contingence qui résume le nombre de résultats dans les catégories suivantes :
- Nombre de sujets positifs avant et positifs après le traitement.
- Nombre de sujets positifs avant et négatifs après le traitement.
- Nombre de sujets négatifs avant et positifs après le traitement.
- Nombre de sujets négatifs avant et après le traitement.
McNemar’s Test specifically focuses on the discordant pairs (those who changed their responses) and uses the formula:
χ² = (b – c)² / (b + c)
where b is the count of subjects who changed from positive to negative, and c is the count of those who changed from negative to positive. The resulting chi-square value can be compared to a critical value from the distribution du chi carré pour déterminer la signification statistique.
In summary, McNemar’s Test is a valuable tool for researchers analyzing the effects of interventions on binary outcomes, helping to clarify whether observed changes in data are statistically significant.