The impossibility theorem of fairness is a foundational result in the algorithmic fairness literature. It states that outside of special cases, one cannot exactly and simultaneously satisfy all three common and intuitive definitions of fairness - demographic parity, equalized odds, and predictive rate parity. This result has driven most works to focus on solutions for one or two of the metrics. Rather than follow suit, in this paper we present a framework that pushes the limits of the impossibility theorem in order to satisfy all three metrics to the best extent possible. We develop an integer-programming based approach that can yield a certifiably optimal post-processing method for simultaneously satisfying multiple fairness criteria under small violations. We show experiments demonstrating that our post-processor can improve fairness across the different definitions simultaneously with minimal model performance reduction. We also discuss applications of our framework for model selection and fairness explainability, thereby attempting to answer the question: Who's the fairest of them all?

Parsons Problems, in which students unscramble provided lines of code, are an exciting innovation for teaching introductory Computer Science. Prior work has shown that students can learn more efficiently by solving Parsons Problems compared to writing code and can learn more effectively by solving Parsons Problems compared to worked examples. Though Parsons Problems have been found effective, there is some evidence that students can use syntactic heuristics to help them solve these problems without full understanding, confirmed to be in use by mature CS1 students in our pilot study. To address this limitation, we propose Faded Parsons Problems, a variation of Parsons Problems where parts of the provided code are incomplete. We built a Flask app that supports Python programming exercises - traditional code writing, standard Parsons Problems, and Faded Parsons Problems - as well as survey and simple comprehension questions (multiple choice and short answer questions). The system can log anonymized data from participants, enforce time limits, and randomize treatment selection. Exercises support an autograder, implemented as a separate worker using RQ that safely executes arbitrary code. Students can run pre-configured tests as frequently as they want to receive detailed output from the test cases: function arguments, expected output, actual output, print statement output, raised exceptions, and custom messages configured by the instructor. Exercise descriptions, provided code lines (including blanks), initial code, and test cases may be configured from YAML files. The system has been used successfully with 50 simultaneous users. A demo can be viewed at https://codestyle.herokuapp.com/sigcse-demo