Experimental Design FRQ

‘Open-ended’ experiment questions;

‘Indicate equipment you are to use’; if the student is given total freedom to select equipment then anything reasonable that could be successfully used to achieve the aim is expected (common sensors, video cameras, light from the sun, etc. are fine). Sometimes students are limited to selecting equipment from a given  list (this list is always comprised of simple inexpensive equipment that most schools have, and that students ‘should’ have had some familiarity with through demonstration or experiment). This part usually carries no credit… it’s to get the student’s creative juices flowing and to concentrate their minds (and to eliminate silly answers)… the exam readers typically check to see if the student has used all of the listed equipment in the ‘how I would do the experiment’ section of the question… if the equipment that has been chosen doesn’t correspond to that used in the experiment, then points are lost.

‘Draw a labeled diagram’; Many students forget to label  the diagram (and so lose at least one point) and forget to draw those things that they think of as obvious, such as thermometers and stopwatches (some go to the other extreme and label the air, the experimenter, etc… no credit lost for this, we just giggle). We are expecting a simple 2D diagram that could be used by another AP student from another school as a useful guide to set up the experiment (perspective, shading, color, etc , are fine but don’t get any extra credit). Not drawing equipment that was in the ‘Indicate the equipment you are to use’ section usually loses a point or two. 

‘Describe how you would use your equipment to conduct the experiment’; Again, if all the equipment is not mentioned here in some relevant way, points are lost. Although we would love to see this done in the third person, past tense etc, it is not necessary. A nice, brief, organized description that guides the reader through the major steps of the experiment is what we want. This description would include initial states (i.e. ‘The cart is released from rest’ rather than ‘The cart is released’), final states (‘…until the cart reaches the end of the ramp’), how many times the experiment is repeated, ways to reduce error. A bulleted list is usually fine (although some readers have wanted to penalize students for this in the past, so far unsuccessfully). I find the most useful advice is to write this description much like a recipe that is to be followed by someone smart that they have never met, and that doesn’t quite understand the point of the experiment.

‘Name one source of error’; here the students will not get any points for naming ‘Human error’… air resistance, error in reading the stopwatch, measuring the height etc. We have very rarely asked for anything to do with percentage error. Very occasionally we ask about the effect of error (‘if the mass of the string were taken into account, what would the effect be on the acceleration of the cart’ (I made this up) ).

Analyze-data questions…

The student is provided with a set of data from an experiment that has already been completed.

Sometimes the student is asked to adapt the data in a way that will make the interpretation of the resulting graph easier. The student is usually given a very big hint about what to do… the result is (always?) the linearization of the data. 

Then the student is usually asked to ‘graph the data’.  Many students seem to think that all graphs should pass through the origin, and that extrapolation of experiment data is perfectly fine. If the data doesn’t pass through the origin, then it is almost certain that there will be a question asking for the significance of the non-zero intercept (such as in photoelectric effect).

Axes… sometimes the axes are labeled for the student, sometimes not. If the axes are not labeled, it has been known for a point to be given for choosing to use the vertical axis for the independent variable. It’s a smart idea to add appropriate units to the axis label (worth at least a point).

Plotting data and sketching a best fit line… (sometimes, we even plot the data for the student!).  Data points should be clear, in the form of a small cross. We expect to see a single smooth line… if it’s a straight line it should look like one (use a ruler). A pencil is the best tool. Sometimes a best fit line will pass directly over one of the student’s data points; if the data point is not very visible, the student may then lose credit for plotting all of the data points. The best fit line should a ‘sketch’ of one… if the student wants to do regression analysis on his calculator, that’s fine, but no extra points. This means that, if the student is asked later to find the slope of the line, that there a range of acceptable scores that will get full credit.

Finding the slope of the line… if it’s a straight line, and the student is asked to find the slope ‘from the graph’ then the student better show that the graph was used (spring constant= slope of graph = ΔF/Δx = (52 – 0) / (13.3-0) = etc. If the data used in finding the slope is only that supplied in the original data (not the best fit line) then things look decidedly fishy. The slope should include appropriate units. 

‘What would happen if’… an example would be a graph related to the gravitational field strength, showing the relationship between the U_s of a particular mass against the mass of a planet… ‘what would happen if the mass were placed on a planet with a greater mass?’ or ‘a larger mass were used in the experiment’...  well, this is where the physics is tested. Apart from investigating the meaning of the slope and area of the graph (dimensional analysis, scouring brain for a relevant formula etc), no helpful advice here.