As the time for fall and winter producer meetings approaches, it seems appropriate to provide a few reminders to consider as you digest all the information that will be presented to you.

An article in the Oct. 13 2015 Southeast Farm Press Daily entitled Don’t be fooled: How farm research results can be misleading by Dr. Eric Prostko, Univ. of Georgia Extension Weed Specialist, is a must read on this subject. Below are my summaries of his major points in the article.

• Research and extension scientists have the responsibility to provide growers with reliable, research-based information. This information can be based on research studies conducted by both themselves and their colleagues.


• Research studies discussed by the presenter or writer are assumed to have been conducted using accepted scientific practices that include replications across time and space. The presented results should be compiled using accepted statistical tools and analysis procedures.


• Recommendations to producers should be made only after research on a particular topic has been conducted across an appropriate number of years and/or locations.


• Results should be presented in a way that will not lead to misinterpretation. To underline this point, pay particular attention to the two graphs shown in the article and Dr. Prostko’s narrative about how the same data presented in the two different graphs can appear to show different results.


• It is important for producers to look closely at research results to determine their validity and application before making major production changes. Be wary of results based on limited data.


• Be cautious of relying on nonreplicated data or data that have not been properly analyzed. Be especially mindful of whether or not differences are merely numerical or are in fact statistical differences.

Additional points that should be considered when presented with information at a meeting or from the various other venues follow.

• In today’s world of multiple media outlets, all disseminated information should be prominently identified as either an opinion or a fact-based presentation. Credible fact-based presentations will provide conclusions that are based on valid evidence.


• Writers and presenters who dispense information should clearly identify where and when this information was generated so the viewer or reader can decide if the location that produced the data is applicable to their operation, and if the data are recent.


• Writers and presenters should not assume or convey cause and effect when items are correlated, or when there may be multiple causes of or reasons for an occurrence.


• A claim or effect is stronger when presented data and results are corroborated by more than one source rather than when supported by a single experiment or example.


• The peer-review process should never be skipped in academic research. That is the final necessary step in validating the conduct of and results from that research. Remember that results from experiments that are presented in meetings are often not reviewed by other persons and/or may be preliminary in nature.


• Remember that an oral presentation can be slanted in any direction the presenter wants. That is not to say the presentation is tainted; rather, it just hasn’t been reviewed by colleagues for quality of content and validity of the interpretation and application of the results.


• Statistical differences among treatment or input effects in a research project should be viewed from both a magnitude and economic standpoint. The magnitude of differences that are statistically significant may not be sufficient to result in an economic difference. Always ask for this clarification from the presenter.

I encourage you to keep the above points in mind as you attend meetings and view presentations, or as you read and file results from recent research.

Composed by Larry G. Heatherly, Oct. 2015, larryheatherly@bellsouth.net