2017 ASHS Annual Conference

Data collection is a crucial part of the scientific process: without data, plant science cannot progress. However, it is only a small part of this process. It is crucial to start with a well-developed hypothesis and careful consideration of which data to collect to test your hypotheses. And since plants are unpredictable, it is necessary to maintain flexibility: if unexpected responses occur, hypotheses may need to be revisited and different data may need to be collected. If unexpected effects are observed, think about how to quantify those effects. This requires an open mind and careful observation during the study. After data collection is finished, careful interpretation of the collected data is also crucial: this is the part of the scientific process where data is converted into information. And presenting this information in a larger, scientific context helps to produce new knowledge, which is in many cases the ultimately goal of the scientific process. When presenting your data, it is important to do so in a context that addresses your hypotheses. And if your study took an unexpected turn, it is perfectly OK to change your hypotheses halfway through or even after data collection has been finished. That flexibility allows scientists to present the most interesting information possible to their audience. Use your data to tell a story that helps your audience understand your reasoning. That is easier said than done. Perhaps the most painful part is to decide which data NOT to present. After spending much time collecting data, it is tempting to present everything. However, if some of your data do not help you address your hypotheses, including these data will likely confuse, rather than enlighten, your audience. In many papers, data interpretation and presentation is limited to determining how different treatments affected the measured variables. Although that approach can be used to describe results, it rarely helps to explain the observed effects. If the parameters to be measured during a study are carefully selected, it may be possible to develop testable hypotheses to explore the mechanism behind the observed effects. That often involves looking for relationships among measured variables, rather than looking at treatment effects. Such relationships can support mechanistical hypotheses of potential causes of observed treatment effects. The process of how to go from data to information to knowledge will be illustrated using a case study.