Answer and explanation:
To address this question, we could design an experiment where the effect of soil and gravel can be tested independently. One approach is to test this via <u>Completely Random Design (CRD)</u>. Here, the <u>treatments would be assigned to experimental units completely at random</u> and each treatment will be tested in <u>replication</u> (e.g. triplicates).
The treatment levels would be (1) <u>Grass growing in the soil (triplicates)</u> and (2) <u>grass growing in the gravel (triplicates)</u>. There will be total 6 experimental units and they will be placed randomly. The growth parameters of grass would be then tested after growing for a certain period of time, let's say 90 days.
After this period, the <u>grass will be harvested</u> and the <u>biomass, root length and shoot length</u> will be measured. Because each treatment has three replicate, we will have a clear estimation of soil and gravel on plant growth. The standard deviation would help estimate the effect of natural variations in different treatments. However, they would be consistent because of same environment given for the growth.
The <u>independent variables</u> would be "<u>soil and gravel</u>" and <u>dependent</u> variables would be "<u>growth parameters of the grass (plant root and shoot length and biomass)</u>".