# What statistical method should I use to analyse data with 3 random variables and 2 fixed variables ? Their interaction is also important

I had 7 treatments for my experiment (where each treatment had a specific type of pesticide) including A, B, C, a combination of A&B, a combination of A&C, a combination of B&C, and a combination of A&B&C. Each treatment was repeated 3 times in 2 different units.

For each treatment, I had 30 plants which received 30 different levels of the nutrient. The location of each plant was fixed (meaning that I had 30 spots for plants in each replicate.) The Nutrient level for each treatment is a continues data varies from 100mg to 1000mg.

I am measuring the weight of the plants after all and this is my output( the dependent variable)

So I have the following factors in my experiment: 1) treatment (type of pesticide(A, B, C, AB, AC, BC, ABC))

2)Nutrient Level (numerical between 100 to 1000, ex: 20, 33, 44, 752, 854, ,...)

3)location of plants.(30 spots, ex: 1,2,3,4,5,6,...)

4)Time (treatments were not conducted at the same time)

5)Unit (I randomly set the treatment in one of the 2 units which were supposed to have the same condition but after a while, I noticed that there is a noticeable difference between the result of each unit)

6)weight of plants (results)

My question is how I should analyze the results.

1)I want to compare different nutrient levels in each pesticide type and their interaction. (this should be a 2 way ANOVA followed by a TukeyHSD test?!! or any other suggestions?)

2)I want to compare different type of pesticides (given that I had 30 different nutrient levels on each type, probably it should compare their interaction?!)

3)I want to check and confirm that the unit effect, location effect or time effect exists or not.