Business Context
DataCorp, a data analytics firm, has implemented a new exploratory data analysis (EDA) process to enhance model accuracy. They want to determine if this new EDA approach leads to statistically significant improvements in predictive accuracy compared to their previous method.
Given Data
| Group | Sample Size | Accuracy (%) |
|---|
| Old EDA | 50 | 75.0 |
| New EDA | 50 | 80.0 |
Task
Assess whether the observed increase in accuracy from the old EDA method to the new EDA method is statistically significant.
Requirements
- State the null and alternative hypotheses.
- Calculate the pooled accuracy and standard error.
- Compute the z-test statistic for the difference in proportions.
- Find the p-value for the test.
- Determine if the null hypothesis can be rejected at α = 0.05.
- Provide a business interpretation of the results.
Assumptions
- The samples are independent.
- The sample sizes are adequate for normal approximation.
- The accuracy percentages are based on binary classification outcomes.
