Hi All,
1. We are aware that your quiz grades and feedback are not yet visible. We are working with course admin to correct that, and we appreciate everyone’s patience. This is a new feature of the CLE website used this year, and any feedback for improving it for next year is welcome.
2. Lab 2 had an optional STATA question, with an answer below.
Your STATA test output should yield the following data:
|
| IQ | |
Hyperbili | N | Mean | Std Dev |
Yes | 81 | 105.9 | 14.1 |
No | 162 | 104.0 | 12.5 |
Total | 243 | 104.6 | 14.1 |
The correct p-value is the middle option: p=0.35
For the ttest function, STATA gives 3 pvalues: the left and right pvalues measure the probabilities that chance alone can find that the "test" mean is less than (left value) or greater than (right value) the "control" mean. These are one-tailed hypotheses. The center pvalue measures the probability that the means are not equal, allowing for the possibility of the test mean to be either greater than or less than the control mean. This is a two-tailed hypothesis. By convention, the two-sided (or "two-tailed") result is the most robust option and what is reported. There are rare cases where a one-sided pvalue is what is reported (such as in a non-inferiority trial), but there will be more discussion on this in the Fall Quarter classes.
- Epi218 TAs