# Rough Theory

Theory In The Rough

## To Die For

So L Magee assisted with one of my planning theory tutorial sessions earlier today, and I returned the favour this evening by dropping in on one of LM’s quantitative methods sessions. I’m certain LM found my presence supportive and ever-helpful. LM never tells me such things, of course, but I’m sure that’s due to some personal reticence about expressing deep emotion.

At any rate, during the session, LM decided to answer a student’s question about expected values and the chi-square test with an example that started, “Suppose you roll a 5-sided die…” LM tried valiantly to progress beyond this starting point, but the example was just… distracting.

“I’m sorry,” one student objected, “but I have to visualise these things. What does a 5-sided die look like?”

LM tried to deflect the question – it’s a hypothetical; Dungeons and Dragons must have all kinds of strange dice; etc. But the student really wanted to know. You could see everyone in the room trying – and failing – to visualise such a thing. The hapless visual learner finally gave up and proposed that we substitute some sort of random spinning dial in our hypothetical example. The tute moved on, but I just couldn’t. I kept flashing back to this discussion, bursting into erratic, poorly-muffled giggling fits at unpredictable intervals while LM was trying to explain other statistical concepts. As I said, I’m sure LM finds me supportive and ever-helpful.

And since I’m supportive and ever-helpful, a gift! LM, the next time this happens in your tutorial, you can show your students this:

### 4 responses to “To Die For”

1. lmagee May 14, 2007 at 10:25 pm

“I’m certain LM found my presence supportive and ever-helpful. LM never tells me such things, of course, but I’m sure that’s due to some personal reticence about expressing deep emotion.”

Such surety amid a discussion of probability! There’s a chance that there are other options… I hazard a guess that there are at least four other equi-probable outcomes to this dilemma:

1. No “deep emotion” was experienced.
2. “Deep emotion” was experienced, but it was merely correlated with, rather than caused by, the afore-mentioned presence.
3. The “deep emotion”, such as it was, was negatively rather than positively correlated with the phenomenon in question.
4. Such “presences”, “tellings” and “expressings” are observations made by observing observers, who make distinctions in the autopoetic system of communication, constituted by selections of information, utterances and understanding, which remains systemicly closed to the 6 billion psychological systems who in turn are inscrutable and uninterpretable as part of the environmental conditions of the communicative system, thus rendering incoherent ascriptions of communicative acts (“expressing[s]”) to psychological states (“deep emotion”) – since, clearly, the observer can only know what (s)he knows, see what (s)he sees, and cannot see that (s)he cannot know that (s)he cannot see what (s)he cannot know. QED.

At least under a Bayesian approach, the last option actually must assume a much greater so-kalled (sic) ‘prior probability’.

2. E May 16, 2007 at 10:58 pm

I have strange dice leftover from my uber nerd days, but I never had a five sided dice! Waitasecond. I think I am still a uber nerd, so I take that back. 😛 And being an uber nerd I would just take a 10 sided dice and count 1&2 as 1; 3&4 as 2; 5&6 as 3; 7&8 as 4; and 9&10 as 5. I also think that dice looks like it has greater or less chance of some numbers coming up due to its shape, I am suspicious about how random it may actually be.

3. N Pepperell May 16, 2007 at 11:32 pm

The person claims that the die is “Statistically accurate!” And who am I to mistrust information from the internet?

4. E May 18, 2007 at 2:40 pm

LOL
Well if you read it on the interweb it must be true!

Why on earth I am concerned about probability of that die being statistically accurate…? I am being such a positivist in my last comment!
Accurate of what exactly again?

😛