About cookies

The NCETM site uses cookies. Read more about our privacy policy

Please agree to accept our cookies. If you continue to use the site, we'll assume you're happy to accept them.


Personal Learning Login

Sign Up | Forgotten password?
Register with the NCETM

Bowland Maths Forum

Only logged in users can post to this forum.

Case Study: Crash test

ncetm_administrator 16 June 2008 14:19
Portal Administrator

Pupils use computer software to explore the impact of car crashes under varying conditions and the effects on a simulated dummy. Pupils can select a car, a crash point and a speed, then watch an animation of the crash and see the results as physical impact on the dummy and as numerical data.

In Crash test, there are 3 stages to the software: researching, testing and presentation. In the research centre, pupils decide which data to collect, conjecture, collect and analyse data and test hypotheses. In the test laboratory, pupils focus on hypothesis testing; they are able to test up to 14 pre-defined hypotheses and choose 3 ‘test packages’ in each case. In the presentation suite, there are tools such as varieties of graph paper and also help to prepare a presentation of results to share with others.

Link to this Post   Alert us about this post

garystars 25 June 2008 21:37
Done this task with two groups.

Once with a high ability year 8 group as a very open task and they really enjoyed and did very well. Setting their own hypothesis and testing them themselves.

The other time with a very weak year 7 group in a quite directed fashion I choose the hypothesis for them but did leave it to the pupils how to test it. They also got a lot out of the task.

A great resource!
Link to this Post   Alert us about this post

ajonion 17 February 2009 15:44
kumers writes:
I have trialled the Crash Test.  It was brilliant , encouraged pupil participation and generated great discussion from year 7 lower ability pupils up to year 9 pupils who have started their data handling module.  A great intro to hypothesis, experimentation and testing and conclusions.  The material itself cannot be dragged on for longer than 2 or maybe 3 lessons at a stretch.  It must be used with caution as the novelty could wear off pretty fast.
Link to this Post   Alert us about this post

sathomas 01 July 2009 09:17
I'm trying out Crash Test at the moment with low ability year 8 and year 9.

They've had one lesson so far and seem to be enjoying it.  Some have already stared to test their hypotheses.

I shall see how they get on with the graphs and presentations.  Any tips would be useful.
Link to this Post   Alert us about this post

maths_sr 27 November 2011 23:27
I'm due to be trying this out with a group of year 8 & 9 students, but I can't seem to work out how to get data the students can use for anything other than bar charts.
Please can someone explain to me how the students go about collecting data that is suitable for pie charts & scatter diagrams. Our school is new to using these projects & noone else I've spoken to from my department has been able to answer this question either. We haven't had much time to look at it collaboratively, so pointers in the right direction would be very much appreciated. Thanks! :o)
Link to this Post   Alert us about this post

ajonion 28 November 2011 16:55
Dear maths_sr, although I don't have the answer to this myself, I will make sure a teacher who has used it gets this message and I will also contact the developer (those who wrote it) for an answer. If anyone else can get onto this more quickly please do. Alice
Link to this Post   Alert us about this post

nrmoore33 28 November 2011 20:03
 Hello there maths_sr,
In terms of correlation, you can plot speed against average dummy damage or car damage for a bit of a crude but hopefully enlightening scatter diagram.
Pie charts are a little tricky as there isn't really any data collected which is a given quantity split between certain variables but you can do dual or triple bar charts (comparing cars, impacts or speeds acorss all control factors) quite simply.
If you have very able pupils you can start analysing the different hypotheses and make pie charts of what percentage of people were correct/incorrect/inconclusive.
Further to this, you can really extend the task by setting percentage parameters for 'serious', 'moderate' and 'minor' damage and see what different types of collision cause each different level of severity (this discrete model also allows more basic data handling).
The best thing I've found about this task is pushing the pupils to set realistic and useful hypotheses such as trying to predict the percentage damage difference across collisions at 30mph vs. 40mph (is the tv campaign about 80% correct?).
Hope this all helps, lemme know if I'm not making any sense!
Link to this Post   Alert us about this post

maths_sr 29 November 2011 00:12
Hi Nick & Alice,
Thanks for your swift replies. From what I can see there will only be 3 points to plot, unless you are suggesting plotting all 3 types of car & 2 impact points (giving 18 points to plot) on the same graph, which seems pointless to me, given we are trying to teach them to control their variables. (I can't see that re-running the experiment gives different results each time either, which would allow this to be possible.) Also, we should be teaching them that scatter diagrams are only suitable for continuous data, but the average dummy data isn't continuous- it's a rating out of 10.
I'm glad I'm not alone in thinking the data collected isn't suitable for pie charts either. My personal problem is I have able students (levels 6c - 7a) who will spot they're being told to fix the variables & tell me I'm wrong directing them to put all the data on 1 chart. They are also likely to let me know they're insulted being asked to draw bar charts.
I like the idea of asking them to set percentage parameters & investigate which types of collision cause each level of severity- hopefully it should get them discussing that "serious" means different things to different people & such terminology shouldn't be used when collecting data. This is something I think I can work with, although this has all students working on the same thing & will make presentations incredibly similar & tedious after the first few have taken place.
Any further help with this would be superb.
Link to this Post   Alert us about this post