destiny team mailing list archive

[Zach Elko <zachelko@xxxxxxxxx>]

Yes, but, the size of the question list has no effect on which random number
is selected to be used as an index into the list of questions. That's why
randomizing the list once ahead of time (or randomize any list ahead of
time) is almost always a better solution than randomizing each time you need
to choose something.

Zach J. Elko
Ball State University Unified Technology Support
Certified Hardware & Software Technician


On Thu, Mar 11, 2010 at 1:32 PM, Doug Shook <dsshook@xxxxxxxxx> wrote:

> Doing run throughs with the questions we have now is not an adequate test
> of repetition as even the largest set we currently have is smaller than I
> would anticipate the actual sets being.
>
> But hey, I would stop you from doing it, it certainly won't hurt anything.
> Go nuts.
>
>
> On Thu, Mar 11, 2010 at 1:29 PM, Zach Elko <zachelko@xxxxxxxxx> wrote:
>
>> As-per doing some run-throughs, I'd beg to differ on the chance of
>> repeating being slim.
>>
>> As for repetition, I think that should manifest itself in repeated plays
>> of the same question set (a question set embodies a subject, a category, and
>> a difficulty), as opposed to the same run-through of a level.
>>
>>
>> Zach J. Elko
>> Ball State University Unified Technology Support
>> Certified Hardware & Software Technician
>>
>>
>> On Thu, Mar 11, 2010 at 1:27 PM, Doug Shook <dsshook@xxxxxxxxx> wrote:
>>
>>> I guess I don't see a problem with doing that, but I don't see why its
>>> particularly necessary either.  If the question sets are of any decent
>>> length the chances of them repeating a question is very small to begin
>>> with.  Also, I don't see the harm in repeating a question or two every now
>>> and then.  If they miss questions it would be helpful for them to get
>>> another shot at it to try and correct their mistake.  Practice makes
>>> perfect, right?
>>>
>>>
>>> On Thu, Mar 11, 2010 at 1:10 PM, Zach Elko <zachelko@xxxxxxxxx> wrote:
>>>
>>>> My view on this is as follows:
>>>>
>>>> Coverage is very important. When a teacher assigns homework, their goal
>>>> is that all of those problems are attempted because each one or each type
>>>> illustrates some particular skill.
>>>>
>>>> The way we randomly select questions now hinders this, but there is an
>>>> easy fix:
>>>>
>>>> Randomize the question set at the time it is created, and then select
>>>> questions from it in-order. This way, each run of the game, the order of
>>>> questions for each specific difficulty and set are randomized, but they
>>>> don't need to be selected at random. By selecting them at random, you run
>>>> the risk of duplicates and not getting full-coverage.
>>>>
>>>> Doug, I can implement this change if you'd like, I know you've done a
>>>> lot recently and are busy with Battleship. Just let me know.
>>>>
>>>> Zach J. Elko
>>>> Ball State University Unified Technology Support
>>>> Certified Hardware & Software Technician
>>>>
>>>>
>>>>
>>>> On Thu, Mar 11, 2010 at 12:43 PM, Doug Shook <dsshook@xxxxxxxxx> wrote:
>>>>
>>>>> I agree with what Ian said, but I also wanted to point out that the
>>>>> games draw questions randomly from the pool of available questions.
>>>>> Therefore its very unlikely that they will even see every question in a set
>>>>> when they play the game.  That being said, making the sets a reasonable size
>>>>> is important to ensure they see a variety of questions each play through.
>>>>>
>>>>> The chances that they will repeat questions is entirely random, so
>>>>> don't factor that in when making the sets.  Memorization is not a bad thing
>>>>> either, especially for simpler questions.  You would EXPECT them to memorize
>>>>> answers, that means they're learning.  Since the placement of the choices is
>>>>> also randomized each time a question is asked, there's little to fear about
>>>>> them memorizing locations of answers instead of the answers.
>>>>>
>>>>> Bottom line: question sets need to cover the material adequately, as
>>>>> Ian said.  Within reason, more is better.  Levels are completely independent
>>>>> from the sets and the number of levels should be determined by how long we
>>>>> want the game experience to be.
>>>>>
>>>>> Hope this helps.
>>>>>
>>>>>
>>>>> On Thu, Mar 11, 2010 at 10:40 AM, Ian Crotty <archangel114@xxxxxxxxx>wrote:
>>>>>
>>>>>> Mathematics questions would be easy to generate a lot of as we have a
>>>>>> lot of versatility with them.  We also do not have to be limited to x+y=z.
>>>>>>  we can do things such as a+b+c=z or any combination of things to that
>>>>>> nature.  What we need to make sure of is that we are covering the students
>>>>>> curriculum adequately.  Sometimes repetition can be helpful in teaching them
>>>>>> the concepts.  On the other hand repetition just allows them to memorize the
>>>>>> question and answer and blow through it.  BF will more than likely take up a
>>>>>> goodly chunk of time just getting through 10 levels and I imagine these
>>>>>> students will have maybe 30 minutes to play, an hour tops.
>>>>>>
>>>>>> If we want to see them finish a game we may want to cut down on the
>>>>>> levels as BF (and yes i realize we changed the name but its force of habit)
>>>>>> has no save progression abilities.  It may be a good idea to set up another
>>>>>> time with the class to have them test the game more thoroughly with longer
>>>>>> periods of time with the game or have other people we know play the game as
>>>>>> well.  This may be something we should talk to Christy about (the amount of
>>>>>> time they get with each game) and we can go from there.
>>>>>>
>>>>>> Ian Crotty
>>>>>> UTS Hardware Technician
>>>>>> Ball State University
>>>>>>
>>>>>>
>>>>>> On Thu, Mar 11, 2010 at 10:27 AM, Andy Elsten <aelsten86@xxxxxxxxx>wrote:
>>>>>>
>>>>>>>  alright guys after making some levels and question sets I noticed
>>>>>>> that BF runs every level that we have. Keeping that in mind if we want each
>>>>>>> question to be unique then we need at least 4 times the questions than
>>>>>>> levels in each of the question sets. So if we have 20 levels there hast to
>>>>>>> be at least 80 to 100 questions in each set. If this is desirable no big
>>>>>>> deal I can make the question sets bigger.
>>>>>>>
>>>>>>> My main question is, have we given any thought as to how many levels
>>>>>>> they are going to run through each time? I only post this question to get
>>>>>>> feed back on the decision to make the questions sets huge to get more unique
>>>>>>> questions or post a limit on the number of levels for each run to reduce the
>>>>>>> chances to repeat question. Just let me know your thoughts on the topic
>>>>>>>
>>>>>>> _______________________________________________
>>>>>>> Mailing list: https://launchpad.net/~destiny<https://launchpad.net/%7Edestiny>
>>>>>>> Post to     : destiny@xxxxxxxxxxxxxxxxxxx
>>>>>>> Unsubscribe : https://launchpad.net/~destiny<https://launchpad.net/%7Edestiny>
>>>>>>> More help   : https://help.launchpad.net/ListHelp
>>>>>>>
>>>>>>>
>>>>>>
>>>>>> _______________________________________________
>>>>>> Mailing list: https://launchpad.net/~destiny<https://launchpad.net/%7Edestiny>
>>>>>> Post to     : destiny@xxxxxxxxxxxxxxxxxxx
>>>>>> Unsubscribe : https://launchpad.net/~destiny<https://launchpad.net/%7Edestiny>
>>>>>> More help   : https://help.launchpad.net/ListHelp
>>>>>>
>>>>>>
>>>>>
>>>>> _______________________________________________
>>>>> Mailing list: https://launchpad.net/~destiny<https://launchpad.net/%7Edestiny>
>>>>> Post to     : destiny@xxxxxxxxxxxxxxxxxxx
>>>>> Unsubscribe : https://launchpad.net/~destiny<https://launchpad.net/%7Edestiny>
>>>>> More help   : https://help.launchpad.net/ListHelp
>>>>>
>>>>>
>>>>
>>>
>>
>