Science Buddies: "Ask an Expert"

I am doing a project on the effects of nitrogen phosphorus and potassium fertilisers on grass seed growth.I have planted and fertilised the grass samples.I have also began my report. I am wondering whether I should repeat the experiment a number of times?

Anyone!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! Please reply

Hi helenxxx,

You are correct that it is ideal to have multiple replications of your experiment. In the real world there is always some variation as a result of unexpected variables or simply chance. In order to be sure that your results are typical, it is best to repeat your experiment multiple times and average the results. The number of replications you perform is called the sample size (n), and the larger your sample size, the easier it is for you to accurately identify trends. A good example is trying to calculate whether girls or boys are taller in 2nd grade. If you just measure one boy and one girl, you might end up with the shortest girl and the tallest boy, which would probably not be a good test. If instead you measure 10 boys and 10 girls and take the average of their heights, you will be more likely to see a trend. Taking the average height of 50 kids will make your estimate even more accurate, and so on. You can perform as many replications as you want, more is always better as long as it is feasible. This is a particularly good idea if you are planning to do any statistical tests, many of which factor in sample size.

If you would like to increase the sample size of your project, you should repeat the experiment exactly as you did before. Since you would be starting later, make sure you keep everything consistent... for example, you will want to give each sample the same amount of time to grow before taking measurements, so your old samples should be measured before your new samples. If you want a sample size of 3 (n=3), plant 3 samples for each fertilizer type and take the average of the growth at the end. For example, if you're measuring growth by grass height, measure the height of each sample, average them together (ex: [height 1 + height 2 + height 3]/3), then compare the average heights for each fertilizer type.

Hope that helps!
Tricia

Thanks!!!!!!!!!!!!!!
I'm planting more tomorrow.