Comments on: SHould I be using probability to design this interactive?
http://ask.metafilter.com/158199/SHould-I-be-using-probability-to-design-this-interactive/
Comments on Ask MetaFilter post SHould I be using probability to design this interactive?Wed, 30 Jun 2010 06:14:05 -0800Wed, 30 Jun 2010 06:14:05 -0800en-ushttp://blogs.law.harvard.edu/tech/rss60Question: SHould I be using probability to design this interactive?
http://ask.metafilter.com/158199/SHould-I-be-using-probability-to-design-this-interactive
I'm scriptwriting an online interactive interview for a curriculum resource. There are ten questions, and students will be able to choose six of these to 'ask', and then see a video clip of each reply. Ideally, I want students to hear four strong answers and two weak ones. Based on this criterion, should I be using probabilty to work out how many of the ten answers should be weak or strong? If so, how? I think I need 6 or 7 strong answers just by applying the ratio.post:ask.metafilter.com,2010:site.158199Wed, 30 Jun 2010 06:02:22 -0800dowcragprobabilityBy: adamrice
http://ask.metafilter.com/158199/SHould-I-be-using-probability-to-design-this-interactive#2267466
If you want each student to get 4 strong/2 weak, this won't work. You'll need to create a strong answer and a weak answer for each question, and serve them up in appropriate ratios.<br>
<br>
If it's good enough for the students to get 4 strong/2 weak <em>on average</em>, then you can take a statistical approach.<br>
<br>
Another approach might be "select 4 questions from menu A, and 2 questions from menu B."comment:ask.metafilter.com,2010:site.158199-2267466Wed, 30 Jun 2010 06:14:05 -0800adamriceBy: EndsOfInvention
http://ask.metafilter.com/158199/SHould-I-be-using-probability-to-design-this-interactive#2267473
I'm a little confused as to what you mean. Is it that:<br>
- You have 10 video clips.<br>
- Students will each randomly choose 6 of the 10 clips (well, randomly in that you have no control over which ones they choose).<br>
- You want students to always end up seeing exactly 4 "strong" clips and 2 "weak" clips.<br>
- You want to know how to make the ratio of the 10 clips clips of strong:weak so that students always get 4 strong and 2 weak.<br>
<br>
If that's all correct, then...<br>
It's currently impossible to do exactly that, even if you change the ratio, the number of clips the students select, or change the total number of videos. There will always be a combination of choices that doesn't fit the correct ratio. Examples:<br>
6 strong, 4 weak - student could choose 2 strong , 4 weak.<br>
5 S, 5 W - student could choose 5 S, 1 W.<br>
8 S, 2 W - student could choose 6 S.<br>
etc...<br>
<br>
Possible solutions are: <br>
a) only offer 6 videos in total (in which case the students have no choice)<br>
b) you design it so that the questions are picked randomly by the interactive system (instead of picked by the student) to ensure they get the right ratio (more variation between students, but the individual students still get no choice).<br>
c) design the system so that the students are forced into choosing the right ratio, e.g. If their first four choices are strong, the system automatically removes the remaining strong questions and leaves just the weak ones, etc. This preserves the ratio but gives the students a little more freedom than option a. In this case the best ratio is probably 6:4 or 7:3 strong:weak.<br>
<br>
On Preview: adamrice also has some good alternatives. I like the one where each video has two versions, and the system decides the right ratio as that gives the students maximum choice over specific questions.comment:ask.metafilter.com,2010:site.158199-2267473Wed, 30 Jun 2010 06:22:26 -0800EndsOfInventionBy: dowcrag
http://ask.metafilter.com/158199/SHould-I-be-using-probability-to-design-this-interactive#2267889
Thank you both for your responses.<br>
<br>
Filming right and wrong versions of each answer would not be possible. I did wonder if this approach would not work since students' selections won't be controlled, and you both have clarified this. I think I will suggest we create five strong and five weak answers and create a bit more variability in who students choose (this whole thing will be repeated for three people).comment:ask.metafilter.com,2010:site.158199-2267889Wed, 30 Jun 2010 10:26:54 -0800dowcrag