Monday, 29 June 2020

I understand material but cant solve problems. How to change it?



As someone who teaches computing, I see this issue a lot. Overall, I agree with Jeff Erickson's Answer to I understand material but cant solve problems
I will add some more context.
Many students don’t really know what learning means. They think memorizing facts or “looking over notes” is enough. Often, because they have done well in earlier classes that way. They think that if an assignment says to write a poker game, then submitting a working poker game means they understood the assignment, even if they submit their roommate’s program, find 90% of it on the web, or get TAs to dictate most of the program to them. They focus on getting a grade or a working program by any means they can, even if they don’t actually learn to do the things that are asked of them. I see this every semester. Students come to office to ask about the exams and say “I do so well on the assignments.” So I ask them to describe their work process, and I know where we are when I hear “When an assignment comes out, I read it over [good!] and then go to office hours, because I know I’ll need help [bad].” What’s bad? Going to office hours to get TAs to tell you stuff before you’ve actually tried anything yourself. Getting help AFTER you’ve tried a few times and thought about things deeply is good. Going before you’ve even thought it through or drawn or a picture of a data structure or listed subtasks is bad, because you are robbing yourself of learning.
I often remind students that the world doesn’t need 200 more poker programs (or search tool, or RPN calculator, or whatever). The point of the assignment isn’t to get a working poker program: it’s to develop the skills that, it turns out, let you write something like a poker problem. If you actually engage the exercise authentically, you will develop the skills and you’ll get the program as a side effect of that. Figuring out how to get the result without developing the skills undercuts the assignment.
Any course, not just computing or other STEM courses, should give you new abilities, skills, not just information and a grade. History classes should ultimately teach you how to analyze evidence and historical processes in a way that allows you to interpret, for example, current events. This is MUCH more than memorizing.
One way to understand this situation is to consider a version of Bloom’s Taxonomy of learning (What Is Bloom's Taxonomy? A Definition For Teachers -

)
Bloom classified skills into different levels. There are a few versions around, but let’s pick this one: memorization, understanding, applying, analyzing, synthesizing, evaluating, and some add creating (the one at the link above leaves this out, but I think it’s important).
I was lucky to learn this in high school. I had a high school teacher who would classify her assignments into these categories, and she said memorizing was important, but if you only did the memorization assignments, you got a D. That’s the lowest, minimal level of cognitive skill. And yet most students come to college believing that’s all there is! To get a C, you had to show understanding (usually by being able to summarize things in your own words) and apply information to novel situations. To get a B, you had to analyze situations (solutions) and synthesize novel approaches from the different things you had studied. To get an A, you had do think critically and evaluate things and create something new (for her, this was always a creative presentation to the rest of the class). I had her for two classes. She drove me nuts, and she made me realize what learning really means. Thank you, Ms. Evans! After her class, I not only learned to set my sights higher, but I developed some skill at evaluating my own understanding. “Does this make sense?” That means I might be up to a D level of understanding. Can I recognize when to apply this? Now we’re getting somewhere! I find many students are unable to evaluate their own understanding, and they suffer the self-delusion of assuming that, if they turned in a working program dictated by the TA, and the grade was good, they must have been successful.
Honestly, most exams, even in STEM classes, have most of the questions at the lower levels. But, in computing, we routinely ask you to understand programs on the exam (without a roommate or TA there to explain things to you or a computer you can use to run it), and we ask you to solve a problem, devise an algorithm, and write code. That gets at application, which many students are not used to doing. (Math is the same: can you see when to apply the chain rule, for example.) We will sometimes get at the higher levels, though not so often creativity. For example, we may ask for an algorithm/function to do something, then ask about its runtime or memory complexity, then ask about alternatives and what situations would be best for which alternatives (analysis and evaluation). As a working programmer, this last step, the ability to weigh alternatives based on the current need, is vital.
So, the key to doing better on exams is not to think about how to get better exam grades. That’s trying to get the effect without working on the cause. You need to practice. Do practice problems. Do lab problems if there are any. Then, you and your friends can test yourself by proposing variations on lab/practice problems and solve them. When you read, read actively: do the exercises and work through the examples on your own and make sure you are able to derive everything yourself. It can be very effective to have a study group where you pick problems, work on them separately, and them compare your solutions. You’ll find each other’s mistakes, and you’ll have the experience of solving a problem yourself, and then seeing and evaluating a few other solutions.
Good skill!


-Mark Sheldon, PhD Computer Science, Massachusetts Institute of Technology(MIT) (1995)

No comments:

Post a Comment

My Life Story: 5000 rupees to 500 crores (Last Part)

Read the first part here before proceeding below :  First Part A fter running the coaching center in Guntur for one year, I had to shut it d...