TIPS ON INVESTIGATIONS

1. Repeating and averaging is to achieve greater accuracy; it has nothing to do with fair testing.
2. Repeating and averaging is to achieve greater accuracy; it has nothing to do with fair testing.
3. Don't make a ridiculously numerically precise prediction which, to the reader, is obviously based on what you have found at the end of the experiment (:-() - and cannot possibly be justified (e.g. I predict that for every 10 N increase in the weight of the block, the friction will increase by 1.3 N).
4.  If you have a signed observation sheet from the practical lessons, make sure that you keep it in the write-up, however untidy it is.
5. The heading box at the top of each column of a table of results should always have four things in it: (i) the name of the quantity (ii) the symbol for it (iii) its unit (iv) the inaccuracy.
6. Number the pages. If the write-up is handwritten, do the numbers in pencil, so you can change them later.
7. Use a treasury tag to hold all the sheets together ( It is more versatile than a paper clip, stapling, or putting into a plastic sleeve).  Then put the whole set of sheets into a folder (to prevent the pages getting dog-eared).
8. GRAPHS :
   (i) Never join the dots; draw a best-fit line (trend-line).
   (ii) Never plot each original one of your repeated sets of results. Work out the average value, which is your best value, and only plot that.
   (iii) Data is better plotted with a + rather than any other shape. This is because the length of the lines can then be made to show the size of the inaccuracy.
   (iv) You will almost always want an x-y scatter graph.
   (v) You can force the computer (in Excel) to plot a trendline that goes through (0,0), if you want to. (It is in the trendline menu, as an option to set the intercept)
   (vi) Getting the stupid machine to give you sensible curved trend-lines is very hard. Usually you would do better doing it by hand.
   (vii) Always put the dependent variable (the quantity you are investigating in your aim) on the y-axis.
9. If you are given a sheet of teacher comments, attach it to the front of your investigation. Then (i) you can tick them off as you respond to them (ii) your teacher can see that you have been working through them (which is encouraging) (iii) your teacher doesn't need to spend unnecessary time making the same comments again.
10. Remember to correct all numbers, especially those on tables of results (e.g. calculated by computer) to a sensible number of sig. figs. (or decimal places).
11. Evaluation: Random inaccuracy is not the same thing as systematic inaccuracy.  The first makes your data scatter around equally on both sides of the true trend; the second makes your data all tend to be biased in one direction.
    

Mail us with comments, questions, and corrections.