Standard Deviation from Inspection

[StumbleBum]

Sorry I meant 68%, I'm fairly delirious.

Haha, gets a bit like that after the monotony of Further exams.

[Daenerys Targaryen]

Yeah, if you're given only 68% of the data then you divide by 2, 95% divide by 4, and so on.
But chances are they're gonna give a full set rather than a fraction of it.

[StumbleBum]

Yeah, if you're given only 68% of the data then you divide by 2, 95% divide by 4, and so on.
But chances are they're gonna give a full set rather than a fraction of it.

Mmm don't think throughout the whole year, including all the practise exams, i've ever seen a question where they haven't used a full set... Would be a nice little differentiating question in a VCAA exam though.

[Daenerys Targaryen]

Yeah exactly, it would rather be ridicluous not to give a full set, otherwise the unknown figures can lie pretty much anywhere, unless they tell us the data is symmetrical

[Stick]

I happened to ask my teacher about this the other day - his answer really clears up a lot of the confusion expressed here.

If the question states the data is normally distributed, you need to divide the range by six. However, if it does not state the nature of the data's distribution, then you divide the range by four. Both are only estimations at best, but it explains why there are two formulae floating around.