Whenever possible, it is best to replicate one’s experiment— that is, to stage multiple tests of the same comparison. Our scientist could perform a replicated experiment on, say, 10 pairs of ponds, adding fertilizer to one of each pair. Scientists record data, or information, from their studies. Researchers particularly value quantitative data (information expressed using numbers), because numbers provide precision and are easy to compare.
The scientist conducting the fertilization experiment, for instance, might quantify the area of water surface covered by algae in each pond or might measure the dry weight of algae in a certain volume of water taken from each. It is vital, however, to collect data that are representative. Because it is impractical to measure a pond’s total algal growth, our researcher might instead sample from multiple areas of each pond. These areas must be selected in a random manner; choosing areas with the most growth or the least growth, or areas most convenient to sample, would not provide a representative sample.
To summarize and present the data they obtain, scientists often use graphs. Graphs help to make patterns and trends in the data visually apparent and easy to understand. Even with the precision that numerical data provide, experimental results may not be clear-cut. Data from treatments and controls may vary only slightly, or replicates may yield different results.