A comprehensive research report to the client must include statistical data analysis using appropriate mathematical tests for statistical significance at various levels of confidence that depend on the variability of the sample population. Analysis of Variance, f and T tests, Non-parametric tests, Regression Analysis, and many other specific statistical tests for treatment effects are used to determine if the degree of effect is significant enough to be scientifically valid with respect to testing of the null hypothesis.

Pacific Ag Group scientists are trained in the scientific method and utilize these techniques in every study we conduct with new products such as pesticides, biological control agents, seed types, soil amendments, new designs of farm equipment, and almost any new technology tested against conventional methods. These technologies are used in replicated experiments where data is collected and entered into computer data bases using the latest versions of analytical software specifically designed for the agricultural research industry.

Our  software capabilities include every available software system, including the commonly used Gylling’s Agricultural Research Manager, Field Pro, Astrix, and American Ag’s eFTN. Our staff is also trained in Microsoft Powerpoint, Excel and Access, for custom graphics and clear presentation of results directly in a format that is familiar and accessable by the client for presentations and transcription into internal company documents.

Because a complete statistical evaluation does not stop with mathematical tests for statistical significance, our experimental results reports are full of explanatory graphs and tables that allow for in-depth review of the most complex scientific work by management. Graphs include standard and three dimensional histograms of multiple experimental variables, regression curves with coefficients of variation data displayed, and artistic representations of proportions and other factors represented as pie charts or population curves through time. All of these statistical and graphical representations facilitate the understanding of complex experimental results - even the best work is useless unless it is easily understood by the end user.