Examkin

Question

Difficulty: HardInferential Statistical Tests

Investigators at a tertiary care center are assessing the utility of a new functional MRI (fMRI) protocol compared to standard electroencephalography (EEG) for identifying seizure foci in adult patients with drug-resistant epilepsy. The study enrolls 6060 patients. Each patient is evaluated with both the standard EEG and the new fMRI protocol. The interpreting clinicians classify each patient's results as either 'focal abnormality identified' or 'no focal abnormality identified' for both modalities. The researchers plan to analyze the data to determine if there is a statistically significant difference in the proportion of focal abnormalities detected by the new fMRI protocol compared to the standard EEG. Which of the following is the most appropriate statistical test to analyze these study results?

  1. A
    Chi-square test of independence
  2. B
    Fisher exact test
  3. C
    Paired Student t-test
  4. McNemar testAnswer
  5. E
    Wilcoxon signed-rank test

Answer

The most appropriate statistical test is the McNemar test, which is specifically used to analyze paired categorical (dichotomous) data.
The most appropriate statistical test for this scenario is the McNemar test. Selecting a statistical test depends on two primary factors: the type of data being analyzed and the relationship between the samples. In this study, the outcome variable is categorical and dichotomous (presence vs. absence of a focal abnormality). Furthermore, because both diagnostic modalities (EEG and fMRI) are performed on the same cohort of patients, the data are dependent (paired). The McNemar test is uniquely designed to analyze paired categorical data, typically structured in a 2x2 contingency table, to determine if there is a significant difference in marginal frequencies (e.g., whether one test is significantly more likely to detect an abnormality than the other).

Step-by-Step Solution

1
Identify the relationship between the study samples.
The two groups of results (EEG and fMRI) are derived from the exact same cohort of 6060 patients. Therefore, the data are paired (dependent).
Determining whether samples are independent or paired is the first critical step in selecting any inferential statistical test.
2
Determine the scale of the outcome variable.
The outcome is classified as either 'focal abnormality identified' or 'no focal abnormality identified'. This is a dichotomous categorical variable.
The type of data (continuous, ordinal, or categorical) further narrows down the appropriate statistical method.
3
Select the statistical test that matches paired, categorical data.
The McNemar test is specifically designed to assess differences in proportions for paired categorical data.
Matching the paired design with categorical data eliminates tests meant for continuous data (like the paired t-test) or independent data (like the Chi-square test).

Key Concept

Selection of statistical tests based on variable type and sample dependence.
Rate this question

Topics

Inferential Statistical Tests

More questions from this topic