Since beta testing a mansion is not the same as playing a mansion, here are some guidelines for beta testers.  Experienced testers may want to add to this list.

1. Check overall playability.  Can the player get all stuff in the room (treasures, keys, etc.) and go between rooms without any complications regardless of whether it is an easy, normal or hard mansion?

2. Check for Jack-stucks, half Jack-stucks and cheap deaths.

a) A Jack-stuck (JS) occurs when Jack cannot go anywhere and cannot commit a suicide.  The player has to restart the mansion or from last saving point.
b) A half Jack-stuck (HJS) happens when the player must sacrifice one life in order to continue.
c) A cheap death (cd) when Jack is killed by a monster or a zapper beam without any control of the situation.  Examples are when monsters are too close to an entrance to a room, a moving platform goes right through a zapper beam, a monter appears out of the blue too close to Jack to react upon etc.

3. Make any kind of weird moves you can think about to check whether they can lead to situations not intended by the builder.  Like what happens if Jack does NOT use a key to open a (optional) door (will it lead to JS later on), can Jack jump on the top of a door and thus skip the key, unintended ladderjumps, etc.

4. Check overall layout, expecially the use of layer 6.  Is it enough or perhaps too much?  Also whether a wrong layer is used unintentionally.

Thx for these beta-testing suggestions, brell.  While your 4 points give very good and practical information, your introduction is maybe even more important.

Being volunteer to beta-test a mansion is not a trick to play a new custom mansion earlier than other forummembers. And at the end of the mansion only telling that it was a tough one and that you missed 17 % of the secrets. Or that a jump in room X was difficult, while you knew in advance that you were testing a hard mansion. Serious beta-testing is IMO pleasant, but hard work. With your eyes wide open for all kinds of possible complications.

When beta-testing I work in exactly the same way as ryos: windowed mode (left) and email and/or texteditor (right). Just want to add that for some types of mansion at least one test with full screen is needed. When there are "slow" rooms in a mansion, there is a remarkable difference in frames per second. Full screen makes Jack even more slackening. 

This is a wonderful idea for a thread, with some great information.  

I tend to play full screen when testing because you can miss some important graphical glitches otherwise.  So this means that I have a pad of paper and a pen handy, and I try to jot down room numbers and general notes which I flesh out when typing the feedback to give the designer.  I also will sometimes circle a room number, which is a note to myself to go back and look at the room in the editor, either because there is something really cool and I want to find out how it was done, or because I just want to have another look at something.

On a general note, Jack-Stucks are really tough things to spot in some cases.  In particular, anytime you have doors closing, zapper beams turning on, or gold/silver bricks changing, you have to remember to try and find creative ways to die after making the change so you can test whether Jack "teleporting" back to the start of the room can cause a Jack-Stuck.  I found a Jack-Stuck the other day in Falcon Manor Hard near the end because I died in a room while testing in a spot where I had never died before, and found myself on the wrong side of a door I could no longer open.  Thankfully, it was easy to fix.