Criteria: DCP Aspect 3

This aspect is all about the graph so if you are going to use a practical for assessment of DCP then it better have a graph in it. Furthermore the graph should be a straight line graph, the IB doesn't actually specify this but it's much simpler if it is so I only assess straight line graphs.The criteria states that there should be errors and uncertainties where relevant, it's best to forget the "where relevant" since they always are. This is for the other sciences where they don't have uncertainties. There is also no mention that there should be a straight line graph but if you read the clarifications you find that to be awarded a 2 then the student must "draw lines of minimum and maximum gradients and determine the uncertainty in the best straight-line gradient" and you can only do that with a straight line graph. There are many cases where it's not possible to linearise a function to plot a straight line but for IB assessment its best to stick to the ones you can.

Here is an example of how we would like the graph to look:

The problem is that the data doesn't always turn out so nicely but if you select your practicals carefully you can get a nice spread of points with medium sized error bars with best fit lines that pass through them

 

 

Checklist: DCP Aspect 3

Processed data should be presented in a graph.  This graph should be linearised. The graph should be drawn using Logger Pro.  If not possible to linearise the function then a curve can be plotted, however this can not gain a complete score.

I am again simplifying things here by saying that I only accept straight line graphs. one could argue that a curve is more appropriate but the moderator might disagree so I decided to cut out the grey area. I say must use Logger Pro here because that's the programme I have bought for my students, graphs can be on paper but I would not accept a paper graph. I like to think that I am doing more than teaching physics and want to encourage appropriate use of technology.

The graph must have a heading, axis labels and units.

I'm not so strict about the heading but any missing axis label or units will prevent a graph scoring 2. I seems a bit harsh but its quite important and not difficult to remember.

Independent variable should be on the x axis

If everything else was perfect I would probably let this one go but generally the independent should be on the x.  There are exceptions like time for example which is always on the x axis even though it is often the dependent.

Graph must include error bars

And the error bars must be the same as the errors in the table. This is worth checking and that is much easier if you mark digitally since you can simply copy their data and draw your own graph. If the error bars don't match then it can't be given 2 marks. The most common mistake is to calculate the errors in the table from the spread of data but use the instrument error in the graph. Whatever graphing programme you use it must be able to different sized error bars for each point.

If the uncertainties were calculated incorrectly but then used correctly to plot the error bars then the student will lose marks in aspect 2 but not aspect 3. This follows the principle that errors carried forward are not penalized. However if the error bars are incorrect it is going to mess up the evaluation when they try to justify an error that wasn't there.

A best fit line should be plotted automatically

Again this means straight line. I always tell students to plot a linear fit rather than proportional since forcing the line through (0,0) will hide potential information about systematic errors. The best fir line should be plotted automatically but sometimes the auto one misses an error bar or two. If this is the case then a manual line might be better, details of how to do the manual line are in the introduction to graph plotting section.

The equation of the line must be displayed (y=mx+c).

All graphing programmes will display the equation of the line but you may have to tick a box for it to be visible.

Manually fit the steepest and least steep lines that fit the error bars

The reason for plotting the steepest and least steep lines is to find the uncertainty in the gradient. There doesn't seem any point in finding the gradient and its uncertainty unless it represents something. If all you want to know is if the relationship is linear then you don't need the gradient for that reason I always organise my practicals so that there is some known value to be found. Loggerpro is very good for fitting line by eye, again check out the graph plotting section.

The gradient and should be quoted.

The values will be displayed on the graph but its worth repeating them underneath. Here unknown value can be calculated too.

The uncertainty in the gradient should be calculated from the steepest and least steep lines.

The steepest and least steep lines are used to find the maximum and minimum value of the unknown, the uncertainty is then approximated to be ½(max-min).