An agricultural scientist wanted to compare the effect of a new fertilizer to that of three older fertilizers—X, Y, and Z—on the
growth of vegetables typically grown in small gardens. Two hundred green bean seedlings were individually planted in identical pots and randomly assigned to one of four groups of 50 each. Seedlings in one group were given the new fertilizer, and the three remaining groups of seedlings were given fertilizers X, Y, or Z, respectively. At the end of four weeks, all seedlings were dried and weighed. The scientist found that the mean weight of the seedlings in the group given the new fertilizer was significantly greater than the mean weights of seedlings in the other three groups. The scientist concluded that the new fertilizer was more effective than the other fertilizers for all vegetables. Why is the scientist’s conclusion not appropriate?
The experiment lacked a control group that did not use any fertilizer.
Step-by-step explanation:
A control group is a group that is identical in every way to all experimental groups except that the control group is not subjected to any treatment. This is done in order to have a standard against which all experimental groups are compared. When no control group is included in an experiment, the scientist cannot know whether the treatment actually had an effect on the experimental groups.
The scientist’s conclusion is not appropriate because it is based on a subjective appreciation (“significantly greater than”).
<em>How “greater” is “significantly greater”?
</em>
A more objective and technical approach would be fixing a level of significance, let's say 0.05 or 0.01 for example, measuring the samples means and standard deviations and then proceeding with an appropriate hypothesis testing in order to see if the differences in means are really significant.
