Question
Your cat strutted across your keyboard and deleted some data points by accident! For 10 points each, describe some ways you deal with your missing data.
[10m] Your analysis could ignore all rows with missing data, at the cost of likely introducing this property to the results. Estimators with an expected value differing from the parameter they’re estimating have this property, which names a tradeoff with variance.
ANSWER: bias [or being biased]
[10e] You can alternatively try to fill in your missing data by using this statistic, which involves adding up all the values in a column and dividing by the number of values.
ANSWER: arithmetic mean [accept cell mean imputation; accept column mean; prompt on average]
[10h] You settle on using this more sophisticated imputation strategy that reduces bias, championed by Donald Rubin. A popular imputation method does this strategy “by Chained Equations.”
ANSWER: multiple imputation [or MICE or Multiple Imputation by Chained Equations]
<IZ, Other Science: Math>
Summary
| 2023 ARCADIA at UC Berkeley | Premiere | Y | 1 | 30.00 | 100% | 100% | 100% |
| 2023 ARCADIA at Claremont Colleges | Premiere | Y | 1 | 20.00 | 100% | 100% | 0% |
| 2023 ARCADIA at Indiana | Premiere | Y | 3 | 16.67 | 100% | 67% | 0% |
| 2023 ARCADIA at RIT | Premiere | Y | 1 | 20.00 | 100% | 100% | 0% |
Data
| Cornell MATLAB | Cornell R | 10 | 10 | 0 | 20 |