Assignment task: There are 4 questions in total please give full descriptions and full answers finish part of assignment only after you have worked your way through Unit 1 and read the paper, Warning Signs in Experimental Design and Interpretation, by Peter Norvig.
1: Define the term "independent variable."
2: Astronomers recently discovered several planets around distant stars outside our solar system. Of the categories of research mentioned in the textbook, which would this be?
3: A new THC-based analgesic (painkiller) was a spectacular success, based on a double-blind study with 4500 patients. You randomly select the case file of one (1) of these patients without knowing the outcome of the entire study, i.e. without knowing about the effectiveness of the medication. It turns out that this patient had received the medication being studied (rather than the placebo). Also, that particular patient showed no improvement, that is, the painkiller did not work for her.
Based on that one file you studied, you decide that the medication has no effect as a painkiller. Is this a Type I or a Type Il error?
4: You decide to test a new brand of liquid fertilizer for tomatoes. It is claimed to improve plant growth within two days after a single application. You use a total of 1000 plants, so you are confident that you will test enough plants. All plants are in a greenhouse under bright light and in the same pots with the same type of soil. Also, you measure the temperature randomly in several parts of the greenhouse and find no variation. You are now confident that all physical/climatic parameters are uniform for all plants. You decide to use 50 of the 1000 plants as controls and make sure these are randomly spaced. Finally, you prepare the fertilizer solution and confirm that each of your 950 experimental plants receives exactly the same volume (200 ml) of the fertilizer solution. You take care NOT to apply the fertilizer solution to the 50 controls. You then lock the greenhouse and come back after three days. It seems like the experimental plants do indeed show more growth when you compare them to the controls, so you get excited and decide to measure them all for statistical confirmation of the difference. There is one problem with your experimental design. What did you forget?