More randomness
I've been looking for more examples of randomness in measurement and how to show that finding the mean reduces random error. It's no good simply measuring a line with a ruler since you just get the same value every time but I thought of drawing lines that are supposed to be the same length. I first did this on paper then tried it using paint. I drew one line at the top of the page then tried to draw a line of equal length at the bottom. Each time I had a new attempt I scrolled down to hide the previous. My attempts are shown here. As you can see some are too long and some too short but I tend to over estimate more than underestimate. The mean value was 211.4 pixels and the actual length was 207 pixels. I tried plotting the distribution of values but didn't get anything sensible at all. Seems I was quite lucky with the plasticine spheres example. Maybe I'd get a better spread if different people made the estimations.