I am currently working on refactoring a summer camp registration system to include some new features and will also be using it as the basis for a new after-school class registration system.

For this new version of the system, I want to start out by writing use case descriptions for the registration process (and ultimately I plan to document all the use cases).

I understand that the purpose of a use case is to accomplish a user's goal, but I'm still not sure how the registration process should be divided into use cases.

Here are the options I've been thinking about:

  1. One big "Register Student(s) For Classes" use case
  2. Divide it into separate use cases: "Enter Student Information", "Select Classes", "Pay for Registrations"
  3. Have separate use cases but also have a "Register Student(s) For Classes" use case which "includes" the more granular use cases

Another factor to consider is timing: parents may want to enter their students' basic information (name, age, etc.) prior to the time when registration begins, but of course some parents will complete the whole registration process in one sitting. Also, payment won't always immediately follow class selection; in some cases parents will be allowed to complete their payment later (within a 48 hour window).

I read some articles about use cases on Alistair Cockburn's website but I'm still confused as to what the guidelines are for dividing up use cases in a case like this.

My inclination is to go with option 3 since the system will have a wizard interface that could potentially guide the user through the registration from start to end, accomplishing the "user goal" of registering his/her student(s) for classes, but the "included" use cases could also be initiated on their own. I was also wondering if I should have a separate "Modify Class Selections" use case for the case when the parent logs back into the system later and changes class selections on behalf of the student, or whether that should just be an alternative scenario of the "Select Classes" use case.

What would you recommend, and why? What guidelines should be used in general when analyzing use case boundaries, beyond the idea of a "user goal"?


I found this quote from the first chapter of Alistair Cockburn's book to be helpful:

Depending on the level of view needed at the time, the writer will choose to describe a multi-sitting or summary goal, a single-sitting or user goal, or a part of a user goal, or subfunction. Communicating which of these is being described is so important that my students have come up with two different gradients to describe them: by height relative to sea level (above sea level, at sea level, underwater), and by color (white, blue, indigo).

It seems to me that "Register Student(s)" is a summary level goal, since it could potentially take multiple sittings to complete. The line is blurred in this case, but I think this idea of a summary use case suggests that my approach #3 above is a reasonable one.

1 Answer 1


In short, there's no real right or wrong answer here: it's one of those things where its up to your work style and work flow. Personally, I like to take all the users that could possibly ever use my system, and then list them out.

  • Parent
  • Student
  • Teacher
  • Administrator

Then, I write ALL the different things those users would want to accomplish for my system

  • Parent
    • Register Child/Children (Could be more than one child)
  • Student
    • Register Themselves
  • Teacher
    • View student's who've registered for their class
  • Administrator
    • Register Students
    • Move student's to different sections

Finally, each one of those behaviors becomes a use case, with main flows, and alternate flows, such as a parent entering an invalid class, or an administrator trying to move a student into a full section. You may find that some of these use cases are very similair and that's fine, feel free to combine them. At least this way you know you aren't missing any!

  • 1
    And make lots of passes through the list. Consider the consequences and constraints of each. How does anyone register for a class, when no one is putting them in the system? What is a section, and who and what can be done with it? How do parents, teachers, and administers become recognized by the system? How do you verify they are who they say they are?
    – JustinC
    Jun 7, 2013 at 3:39
  • I'm aware of the importance of defining Actors but wasn't listing things out in quite this way, which is helpful. It seems that you agree that my approach #3 could work well? My understanding is that sometimes there can be more than one "primary actor" for a use case (or sometimes a "secondary actor") but I will try keeping the registration by admin use case separate from the registration by parent use case and see how it compares. Jun 7, 2013 at 3:43
  • Can anyone point me to any books or articles that talk about use case decomposition in a bit more depth? This answer is quite helpful but I'd still like to understand the general guidelines better. Jun 7, 2013 at 3:45
  • 1
    While resource requests are discouraged here, I will suggest that Alexander, Fowler, Cockburn, and Jacobson all have good books on the subject of use cases/requirements. Some are a bit older and some of the thinking has changed slightly since, but they remain fairly "classical" on the subject.
    – JustinC
    Jun 7, 2013 at 5:20
  • Thank you. I know @Ampt said there's no "right or wrong answer here" but I just wanted to repeat my earlier question: "It seems that you agree that my approach #3 could work well?" Jun 7, 2013 at 13:00

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.