New Internet-Based Statistical Analysis Software for Analyzing On-Farm Test Results

Please note the information on this page is from February 2003 and AGSTATS is now fully retired.

That said, faculty in the on-farm testing program at the University of Nebraska Lincoln recognized that there was still need for on-farm statistical analysis software and have developed and are supporting a tool named FarmStat.  It can be found at  University of Nebraska-Lincoln CropWatch  Like AGSTATS, FarmStat is free, but one must register to use the program.  We wish you continued success in your on-farm testing work!

Pacific Northwest Conservation Tillage Handbook Series No. 18
Chapter 10 – New Technology Access, Adaptation and Economics, February 2003

Authors: Roger Veseth, WSU / UI Extension Conservation Tillage Specialist; and Russ Karow, OSU Extension Agronomist and Head of OSU Crop and Soil Sciences Dept., Corvallis

Growers and grower-advisers now have Internet access to an easy-to-use, Window’s-based software for statistical analysis of the results of simple on-farm field experiments.

The software was recently completed through a grant from the Pacific Northwest STEEP (Solutions To Environmental and Economic Problems) research and educational program on new technologies for conservation farming systems. It was developed by a group of undergraduate students at the Oregon State University Department of Computer Science, in collaboration with a number of Extension specialists at OSU, Washington State University, and University of Idaho.

Background

While many statistical analysis programs exist, most require significant knowledge about statistics and are not applicable for most growers and growers advisers. The new AGSTATS02 software was developed to replace the simple DOS-based AGSTATS program released in 1990 by PNW Extension specialist involved in a STEEP regional on-farm testing project. Over 450 copies of this earlier software version have been distributed around the world since the program was released. The new software was initiated as a result of a growing interest in expanding the program to a Windows-based environment with more user options in the statistical analysis and data management.

Software Design

The AGSTATS02 program allows the choice of either completely randomized or randomized complete block designs of field experiments. These are the most common and practical designs for field trials, particularly with field-scale equipment. The program can handle up to 16 treatments and 16 replications and allows choice of significance levels (1, 5, 10 or 20% probability) rather than the fixed 5% level used in the earlier AGSTATS program. It uses a spreadsheet-like data entry format, and allows storage, retrieval and printing of data files. Data is analyzed using a standard ANOVA analysis of variance procedure. Least significant difference values are calculated along with probability and coefficient of variation values.

The program is currently designed primarily as a web-based software. Although an option is available to download the program for stand-alone use, it is NOT RECOMMENDED — except for highly experienced computer users, since some systems configuration files on the download computer can be changed during the AGSTATS02 software downloading process, causing problems with running some other software loaded on the computer. Considerable time and effort was spent trying to create a stand-alone program that could be easily downloaded. However, due to the great diversity of computer operating systems, Internet browsers, and software interfaces that exist in today’s world, this was deemed a futile effort given the limited budget and timeframe for this project. So to be on the safe side, we recommend running AGSTATS02 on the Internet.

AGSTATS02 is still in an evolutionary phase. Changes are being made as the glitches are discovered, but as noted earlier, because of the great diversity in computer hardware and operating systems in today’s world, the program will never be foolproof. The old DOS version of this program (AGSTATS) can be downloaded from the Website. Go to the main page toolbar and click on “Download.” Then chose to download the MS-DOS program. This download does not change your computer configuration, as does downloading AGSTATS02, and so is safe to do. Two files must be transferred to your computer and must be co-located in the same directory for the program to function. The old program is far more restrictive than AGSTATS02 but still quite useful.

Internet Procedures

The ATSTATS02 program can be accessed at the URL address listed above. You are encouraged to first click on “On-Farm Testing” on the middle blue bar of the screen for a basic review of the principles behind designing scientifically-valid, on-farm field trials. This reference, “On-Farm Testing — A Scientific Approach to Grower Evaluation of New Technologies,” is a PNW Extension Conservation Tillage Handbook Series publication that is linked from the main Website. This publication helps to provide an overview of some of the terminology and options utilized in the AGSTATS02 program.

A second reference section to review before getting starting with the software can be accessed by clicking on “Program Documentation” on the end of the middle blue bar on the screen on the AGSTATS02 Website. Then, under the category “General User Documentation”, review “Introduction” and “Using AGSTATS02.”

To run the AGSTATS02 program on the Internet, click on “RUN IT” on the blue bar. You will be prompted for a login name and password. If you have not used the software, click on “New to AGSTATS02? Register Now!” You will receive an automated e-mail notification with your registration information. Save the e-mail copy for future reference on your “Login” name and “Password” to use AGSTATS02 in the future. If you have problems getting past this dialog box, you may need to download a newer version of your Internet browser. Once in the software program, you can simply follow the directions provided.

We hope the AGSTATS02 software will facilitate new opportunities to develop on-farm trials and more accurately analyze the results.