The original Seven Essays, in a cool, new outfit ...

Statistics: Experiment Design, Results, and Interpretation

Statistics: Experiment Design, Results, and Interpretation

A completely randomized (CRD) experiment was conducted in a glasshouse to test whether mature sugar beet roots are heavier when using a new fertilizer (A). Twenty pots of equal size were filled with the same compost and randomly placed on the bench in two parallel rows as shown below. To ten of pots the old fertilizer (B) was added, and to the other ten pots the new fertilizer (A) was added. Sugar beet seedlings of the same age and similar size were then planted into these pots.

(i) Use descriptive statistics and graphic display to present the results. Are the parametric assumptions met? With reference to the test assumptions, choose a statistical test to determine whether the new fertilizer produces heavier sugar beet roots than the old fertilizer.

(ii) State the null hypothesis and analyze the data. Interpret the results using 95% confidence limits and the test of significance. Is there evidence that the average weight of roots is increased by using the new fertilizer?

(iii) Suggest an improvement in the design of this experiment? Are there any disadvantages to the new design?