Mean and standard deviation by stealth

February 13th, 2010

Formula for sample standard deviation using the square root of the sum of the squares of the deviations from the mean divided by one less than the sample size

Download a set of instructions written in English about how to calculate the mean and standard deviation of a data set.

This has been set to a group of students as homework (old fashioned term, but that is what it was) and I will evaluate the success of the mission in 10 days or so it being half term next week. The numerical answers are mean 101.9 and standard deviation 6.5.

I decided to turn the calculation into an exercise in following instructions as the particular student group was familiar with lab protocols. I then follow up with some explanation. There is a spreadsheet formula reference card in production, and I will be evangelising on the merits of symbolic notation soon enough. I do NOT teach the alternative algorithm as it is numerically unstable with data sets that have low variation.

OpenOffice note: The maths code used to produce the standard deviation formula was

SD = sqrt{{sum{(x-bar{x})^2}}over{N-1}}

I much prefer the OpenOffice formula editor to the more mouse directed one provided with a well known commercial package…

Ten Data Sets

February 7th, 2010

R generated histogram showing heights of a simulated normal sample of 100 army recruits with mean height 173 and standard deviation of 7.5 cm

Download a PDF file with 10 data sets that can be used to illustrate a variety of statistical techniques

Useful pages:

A spreadsheet with all the data sets is on its way so I can demonstrate how to analyse the data once we have decided what methods are appropriate. I like having a dialogue and a spreadsheet instead of Yet Another PowerPoint!

February 2010 desktop

February 6th, 2010

February 2010 desktop

Click on the thumbnail above for a 1280 by 1024 pixel desktop image with a February 2010 calendar. Its on my desktop computer for what is left of this month. James Watt (with slide rule and plan) on the left, so a tenuous link to Maths.

Five number summary

February 5th, 2010

Just 10 minutes on the five number summary. I’ve not produced a screencast for some time, and this topic presented itself as potentially useful to a range of students but not ‘mainstream’ on any of their courses.

I used the NCH screencam software. This YouTube is a ‘single take’ without script walking through the PowerPoint I usually use for this topic. You may notice that I ‘loose’ it in the middle slightly.

Chocolate mushroom package

January 30th, 2010

Japanese chocolate and biscuit mushrooms

Ruth gets chocolate from an international sweet shop in Birmingham. This Japanese pack seems to contain mock mushrooms with chocolate caps and biscuit stalks. I love the typography on the back of the package…

back of Japanese chocolate mushroom package

Some of the details appear to be telling a story (aimed at children?)

chocolate mushroom man loosing his hat

And the teacher figure appears in red on the back – I wonder if this is some kind of food labelling thing?

Red callout on back of the package with teacher-cat

I rather like the teacher-cat, so I used the levels tool in GIMP to remove the colours and exported him (her?) as a GIF cartoon. I might add this as a logo to some of my worksheets to keep the atmosphere light. I can always use the Inkscape trace tool to tidy the edges up.

teacher cat or bear cartoon

Five number summary

January 28th, 2010

gnumeric is a bosting spreadsheet for statistics graphs

Download a two sided handout on finding the five number summary for a set of data

The five numbers are the maximum, the minimum, the median and the upper and lower quartiles. This set of numbers can tell you about the central tendency of your data, the spread, the extreme values, and provide low order information about the shape of the distribution. The blood pressure data sets came from Dr Bradstreet’s Favorite Datasets in Early and Late phases in Drug Research.

Box and whisker plot next, so I can explain how to interpret the five number summary. This series is part of a few sessions I’m planning for some degree students. The box and whisker illustration here was produced using the Gnumeric spreadsheet – it has a useful selection of statistical graphs and can export graphs as SVG or in a variety of bitmaps. Version 1.9.9 is in the Ubuntu 9.10 repositories.


January 27th, 2010

Alan Kays Dynabook sketch

...that have solidified his believe that the PC industry needs to move away from just selling hardware and towards a service-based model that could be used to establish an educational infrastructure. “It’s all about long-term, sustaining relationships,” he told me, something that mobile phone companies have been practicing for years.

Alan Kay on the Apple Tablet and its alleged included phone plan. I suppose a move from rent a pipe to pay per megabyte is inevitable. Via daringfireball.