After gathering harvest data, it’s time to analyze the results. Using statistical analysis will give you confidence in your data.

Statistics is a mathematical approach to determine, within pre-set confidence limits, if there are significant differences between treatments.

Significant differences between treatments occur when differences are consistent and great enough that you can be confident that they aren’t due to chance occurrence. If the treatments differ significantly, you can be confident that one treatment will be better than the other treatment under similar environmental and management conditions. The Least Significant Difference (LSD) is the minimum difference between treatment averages that is still considered significant. We have often mentioned having a level of confidence in the results. The Confidence level is how certain you can be that differences in the data are due to the treatments and not to chance. The confidence level is expressed as a percentage; most researchers use 95% when testing hypotheses with the intent of publication. Let’s say you conduct an experiment 100 times under similar conditions. At a 95% confidence level, you would expect to see similar treatment differences 95% of the time but other results in 5% of the cases. Farm decisions often are made with less certainty and an 80% confidence level may be adequate. You may be wondering why statistics are important. Statistics allow you to take into account that a random event could occur and change the outcome of what would “normally” happen. For example, your favorite college football team is the best team in its conference. This coming Saturday they are to play the worst team in the conference. You are fairly certain that your team is going to win but realize that they could lose, maybe because the star player is injured or the other team plays harder and smarter than usual.

 Statistics can help determine the probability that your favorite team, the conference leader, will not win. If your team and the worst team in the conference play many times, the number of wins and the point margins can be used to determine the risk of losing; we cannot be certain of winning the next game but we may be able to determine that there is just a 5% probability that your team will lose. The same is true with the data you collect with your on-farm research study; it needs to be analyzed to determine the probability that you will get similar results in another year or field.