44 SportsTurf | April 2014 www.sportsturfonline.com
Facility
& Operations
fertilizer
(X-axis equaling increasing rates of nitrogen fertilizer and the
Y-axis equaling turfgrass growth).
However, we also need to ask, "Is nitrogen fertilizer the only factor
that can increase growth?" The answer is obviously "no." External vari-
able such as temperature and rainfall can influence results as well. So
we can see that statistical relationships are not so clear cut and analyses
try to find the best fit (the slope of the line) for this relationship.
ANOVA is used to analyze differences or equality between treat-
ment means. ANOVAs are useful for comparing two or more means
for statistical significance. Significance between means is often deter-
mined by a threshold value such as the Least Significant Difference as
one measure.
Analysis
of data can be very confusing, drawn out and beyond the
scope of this article. Those of us in Plant Sciences often consult with
statisticians to aid in the analysis of large data sets. Let's leave this up
to the experts.
EXAMPLES OF TABLES AND
CHARTS & WHAT TO LOOK FOR
Understanding data tables becomes an easier task now that you
understand some terms like the mean, standard deviation and least
significant difference. The following example comes from the Na-
tional Turfgrass Evaluation Program website. All tables should be ti-
tled; columns labeled and have some indication of significance
between means.
This example shows a data table for weed ratings in some bermuda-
grass cultivars. The numbers listed under TN1 are means of three repli-
cations of percent weed ratings. Several text boxes explain much of the
information on the data table; however, the most important question
to ask, "Are there any differences, significant differences? " You will
noted that the Least Significant Difference (LSD) value is 1.6 If the
differences between means is greater than 1.6, then you will see a dif-
ferent lower case letter adjacent to that mean. It also specifies that the
LSD is an LSD set at 0.05 or a 95% confidence level. Means with the
Continued from page 10