A study was performed on 20 trees to test the effectiveness of a new fertilizer. The results of this specific study is brought into question. What would possibly improve the reliability of the results?