Another (*very* similar) framework is item response theory. The model you’re using would correspond to what’s called a 2PL model, named because it has two parameters: difficulty and discrimination. IRT difficulty measures the difficulty of an item and corresponds to difficulty in your model; IRT discrimination measures how sharply an item distinguishes between individuals near a given ability level and corresponds to the slope in your model. In the 2PL model, the probability of a person answering an item correctly is , very similar to your model.
They also have something called the item information curve, which graphs how precisely an item measures a trait across different levels of that trait. Though they graph the inverse squared standard error (information), rather than the standard error itself like you do. (I think their way of graphing it makes prettier graphs, though.)
Another (*very* similar) framework is item response theory. The model you’re using would correspond to what’s called a 2PL model, named because it has two parameters: difficulty and discrimination. IRT difficulty measures the difficulty of an item and corresponds to difficulty in your model; IRT discrimination measures how sharply an item distinguishes between individuals near a given ability level and corresponds to the slope in your model. In the 2PL model, the probability of a person answering an item correctly is
They also have something called the item information curve, which graphs how precisely an item measures a trait across different levels of that trait. Though they graph the inverse squared standard error (information), rather than the standard error itself like you do. (I think their way of graphing it makes prettier graphs, though.)