Is it a good idea to publish a complete open-access solutions manual to accompany my open-access textbook?

Question

It will likely be several years (probably more) before I write a textbook, but it's something I've been finding myself pondering in the back of my mind recently.

I would like to distribute my book à la Lienhard; that is, freely available on a website. Additionally, I would like to distribute complete solutions to the exercises therein through the same medium. However, I wonder if

1. educators might not adopt the book because the exercise solutions are easily found, and/or
2. students might not get the most out of the book because the exercise solutions are easily found.

If I distribute the exercises in .tex format, like LeVeque, the first point could be mitigated by allowing instructors to customize them more easily. To the second point, I did about as well in classes where solutions were provided with the problems (and the problems weren't graded) as I did in ones with solutions provided after the due date. But I'm sure that won't be true for everyone, and I know that there are other factors.

So I pose to the educators and students on this site whether it would be a good idea to freely provide complete solutions to my freely provided textbook.

Note: I see this question as related, but it doesn't address the specific case of having a one-stop shop for solutions for entire sets of problems that is updated and cross-referenced correctly (i.e., no web searching required). Perhaps this question is even closer, but it doesn't address open-access resources.

Answer 1

Consider a compromise: Even numbered problems have solutions provided, odd numbered problems do not. Make sure each important point and degree of difficulty is covered by both.

An instructor setting homework would base it on odd numbered problems.

An independent student, or one that wants more practice e.g. before an exam, would use the even numbered problems.

Answer 2

I note, from my experience, that students who have access to the solutions will look at them as soon as the solution becomes difficult instead of concentrating and focusing on working through to the solution. Some students will look at the prepared solution and “feel” they know how to solve it , but when it comes to doing a question “cold” then they find they don’t know.

Based on the comments from students I give the problems (which have the final answer so they can check accuracy) and a week or so later I give access to the worked solution - most seem happy with this.

Answer 3

Personally, as the author of some open-source freshman physics textbooks, I've kept the solutions secret and only given electronic access to faculty who adopted the book. I hand out the solutions to students on paper and ask others not to distribute solutions electronically to their students. I'm not under the illusion that this has kept solutions from getting onto the web, but I suspect that what is on the web is probably fragmentary and inconvenient to access.

I know at least one author of a very nice math textbook (Hefferon, linear algebra) who has included all solutions in the book.

Educationally, it seems that either way can be made to work. What I do works for me, and I assume that Hefferon chose what he was doing because it worked well for him.

In terms of getting adoptions, I also don't see any clear evidence. Both Hefferon and I have had our books adopted by other people.

I do have the impression that at most schools in the US, making solutions freely available is considered unusual. Putting solutions on the internet for free, in convenient form, is not something you can change your mind about later if you decide it was a bad idea.