14 ms·

Finding and fixing standard misconceptions about program behavior

77 points by vector_spaces 2y ago

justin_oaks 2y ago
This is great. It's good to see someone attempting to identify a mistaken mental model and then correcting it. In both formal and informal education, it is rare to see teachers/parents/mentors identifying misunderstandings and correcting them. All too often the one teaching reiterates the lesson material instead of asking the learner about their understanding and correcting misunderstandings.
- 082349872349872 2y ago
  I self-taught basic algebra with a "choose your own adventure" book that advanced to the next topic for correct answers and went through pages explaining the common misconceptions then returned to the question for incorrect. (to be fair, class sizes were ~30 when was in school, so asking the learner about their understanding and correcting misunderstandings would not have been likely: assuming 40 minutes lecture time for a 50 minute class, that's 20 seconds per student, and at most ~2 minutes if no lecture time whatsoever)
Turing_Machine 2y ago
Nice, but needs some more attention to variant (but equivalent) responses. For example, for one of the questions in the first module I selected "Other", then typed "Error" (in the free-form response box). The answer it was looking for was "error" (lower-case e). Either it should accept both, or if you're really insistent on distinguishing the two, it should be made more clear at the beginning that system messages (rather than explicit results) are going to be case-sensitive. Counting "ABC" wrong when the expected answer is "abc" would be fair game. Counting "Error" wrong for a pseudo-language that hasn't even been formally defined is not, or so I see it. Looking at some of the languages on which this is claims to be based: In JavaScript, typing abc when it's undefined produces: Uncaught ReferenceError: abc is not defined at <anonymous>:1:1 (anonymous) @ VM33:1 (note capital E in ReferenceError) In Python3 you get: NameError: name 'abc' is not defined. Did you mean: 'abs'? (again, note capital E in NameError) In Racket 8.9 you get: abc: undefined; cannot reference an identifier before its definition (no 'error' of any kind!) I didn't install OCaml, because I need another "package manager" like I need a hole in the head. Java: does not compile, although to be fair the compiler error message does have an 'error with lower-case e' in it. C#: Yeah, not gonna install that either. This may seem extremely nit-picky (and it is), but when you're teaching people who've never programmed before, and who may have never even encountered the concept of case sensitivity before ("But Google doesn't care even if I type in all caps!"), nit-picky is the way it needs to be. As I said, I do like this, but it needs more attention paid to parsing free-form responses. The problem is not unique to this system, of course. The free-form response is where most systems of this general type tend to fail.
- echoangle 2y ago
  There was an example directly before that question which shows an example output when dividing by zero. There you can see what format they expect the error to be.
  - Turing_Machine 2y ago
    These are supposed to be beginning students. You can't just assume that they know that there's even any difference between "Error" and "error".
- soegaard 2y ago
  > In Racket 8.9 you get: > abc: undefined; cannot reference an identifier before its definition > (no 'error' of any kind!) But that is an error. If you try this in the terminal, you will see that the message is colored red to show it is an error. If you try it in DrRacket, you will see the message written in red text with icons leading to the stack trace. If you try it in Emacs, you will see the message in red indicating it is an error.
echoangle 2y ago
I get that the syntax options are just to adapt to different language users but it’s a little bit confusing that some claims are actually wrong in the specific language. In the first test, they state that adding two bools gives an error, even though this is perfectly valid in Python.
Animats 2y ago
> We also believe that the terms “call-by-value” and “call-by-reference” are so hopelessly muddled at this point (between students, instructors, blogs, the Web…) that finding better terminology overall would be helpful. Maybe that shouldn't even be exposed to the programmer. The programmer-level questions are, is it copyable, is it mutable, and is it alias-free? Whether it's passed as a copy or a pointer is really an issue for the compiler. If you're passed a read-only copy of something guaranteed to not be aliased, you can't tell the difference from a reference. Some Modula compilers made that decision automatically, based on object size. Rust compilers have the info to do this. I'm sometimes asking myself whether I should pass, say, an array of 3 32-bit floats in graphics code by reference or by value. The compiler knows better than the programmer what the hardware can copy fast, and that may differ with the platform.
- erik_seaberg 2y ago
  Deep value equality can also be expensive to check, so some languages default to reference equality even for immutable values.
  - mrkeen 2y ago
    I used to think having both == and .equals in Java was specifically dumb, but the more I think about it, the more I dislike this kind of sneaky reference equality in the general case. Comparing references isn't even an optimised version of comparing values, because when the references don't match but the values do, equality will be false. I struggle to think of when reference equality would be useful to me. Like if I were writing a standard for-loop over an array, it's like I'd be comparing 'i' to 'i'. If == gives me a fast 'yes' response, I feel like I (or my algorithm) should have already known they were going to be the same. If == gives me a fast 'no' response, then do I not actually care about the value?
    - erik_seaberg 2y ago
      I think Python got this right, where "x is y" is identity (not often used but still concise) and "x == y" is value equality (a method you can override).
acbart 2y ago
Oooh, this is awesome. I love the formalizationd presented here, and the formalization of a notional machine into something more powerful and executable. Can't wait to hear more about this. Great name too! SMoL!

77 points by vector_spaces 2y ago

This is great. It's good to see someone attempting to identify a mistaken mental model and then correcting it. In both formal and informal education, it is rare to see teachers/parents/mentors identifying misunderstandings and correcting them. All too often the one teaching reiterates the lesson material instead of asking the learner about their understanding and correcting misunderstandings.

082349872349872 2y ago
I self-taught basic algebra with a "choose your own adventure" book that advanced to the next topic for correct answers and went through pages explaining the common misconceptions then returned to the question for incorrect. (to be fair, class sizes were ~30 when was in school, so asking the learner about their understanding and correcting misunderstandings would not have been likely: assuming 40 minutes lecture time for a 50 minute class, that's 20 seconds per student, and at most ~2 minutes if no lecture time whatsoever)

Turing_Machine 2y ago

Nice, but needs some more attention to variant (but equivalent) responses. For example, for one of the questions in the first module I selected "Other", then typed "Error" (in the free-form response box). The answer it was looking for was "error" (lower-case e). Either it should accept both, or if you're really insistent on distinguishing the two, it should be made more clear at the beginning that system messages (rather than explicit results) are going to be case-sensitive. Counting "ABC" wrong when the expected answer is "abc" would be fair game. Counting "Error" wrong for a pseudo-language that hasn't even been formally defined is not, or so I see it. Looking at some of the languages on which this is claims to be based: In JavaScript, typing abc when it's undefined produces: Uncaught ReferenceError: abc is not defined at <anonymous>:1:1 (anonymous) @ VM33:1 (note capital E in ReferenceError) In Python3 you get: NameError: name 'abc' is not defined. Did you mean: 'abs'? (again, note capital E in NameError) In Racket 8.9 you get: abc: undefined; cannot reference an identifier before its definition (no 'error' of any kind!) I didn't install OCaml, because I need another "package manager" like I need a hole in the head. Java: does not compile, although to be fair the compiler error message does have an 'error with lower-case e' in it. C#: Yeah, not gonna install that either. This may seem extremely nit-picky (and it is), but when you're teaching people who've never programmed before, and who may have never even encountered the concept of case sensitivity before ("But Google doesn't care even if I type in all caps!"), nit-picky is the way it needs to be. As I said, I do like this, but it needs more attention paid to parsing free-form responses. The problem is not unique to this system, of course. The free-form response is where most systems of this general type tend to fail.

echoangle 2y ago
There was an example directly before that question which shows an example output when dividing by zero. There you can see what format they expect the error to be.
- Turing_Machine 2y ago
  These are supposed to be beginning students. You can't just assume that they know that there's even any difference between "Error" and "error".
soegaard 2y ago
> In Racket 8.9 you get: > abc: undefined; cannot reference an identifier before its definition > (no 'error' of any kind!) But that is an error. If you try this in the terminal, you will see that the message is colored red to show it is an error. If you try it in DrRacket, you will see the message written in red text with icons leading to the stack trace. If you try it in Emacs, you will see the message in red indicating it is an error.

echoangle 2y ago

I get that the syntax options are just to adapt to different language users but it’s a little bit confusing that some claims are actually wrong in the specific language. In the first test, they state that adding two bools gives an error, even though this is perfectly valid in Python.

Animats 2y ago

> We also believe that the terms “call-by-value” and “call-by-reference” are so hopelessly muddled at this point (between students, instructors, blogs, the Web…) that finding better terminology overall would be helpful. Maybe that shouldn't even be exposed to the programmer. The programmer-level questions are, is it copyable, is it mutable, and is it alias-free? Whether it's passed as a copy or a pointer is really an issue for the compiler. If you're passed a read-only copy of something guaranteed to not be aliased, you can't tell the difference from a reference. Some Modula compilers made that decision automatically, based on object size. Rust compilers have the info to do this. I'm sometimes asking myself whether I should pass, say, an array of 3 32-bit floats in graphics code by reference or by value. The compiler knows better than the programmer what the hardware can copy fast, and that may differ with the platform.

erik_seaberg 2y ago
Deep value equality can also be expensive to check, so some languages default to reference equality even for immutable values.
- mrkeen 2y ago
  I used to think having both == and .equals in Java was specifically dumb, but the more I think about it, the more I dislike this kind of sneaky reference equality in the general case. Comparing references isn't even an optimised version of comparing values, because when the references don't match but the values do, equality will be false. I struggle to think of when reference equality would be useful to me. Like if I were writing a standard for-loop over an array, it's like I'd be comparing 'i' to 'i'. If == gives me a fast 'yes' response, I feel like I (or my algorithm) should have already known they were going to be the same. If == gives me a fast 'no' response, then do I not actually care about the value?
  - erik_seaberg 2y ago
    I think Python got this right, where "x is y" is identity (not often used but still concise) and "x == y" is value equality (a method you can override).

acbart 2y ago

Oooh, this is awesome. I love the formalizationd presented here, and the formalization of a notional machine into something more powerful and executable. Can't wait to hear more about this. Great name too! SMoL!