I know this is fundamental. Please forgive me. I am new to R.
I ran the Scheffe post-hoc test after ANCOVA using "scheffe.test" from package "Agricolae". I got the following output:
> # Post-Hoc > library(agricolae) Warning message: package ‘agricolae’ was built under R version 3.4.2 > > # Qstorm ~ LandUse*Rainfall > comparison <- scheffe.test(r3,"LandUse", group=FALSE, console=TRUE, + main="Qstorm among different LandUse*Rainfall") Study: Qstorm among different LandUse*Rainfall Scheffe Test for Qstorm Mean Square Error : 64.70584 LandUse, means Qstorm std r Min Max LF2 9.413656 13.234435 82 0.00000 64.90228 LF3 6.509084 10.008503 123 0.00000 46.30602 OP 4.676709 6.863499 121 0.00019 30.18665 VJR 5.710812 10.682185 66 0.00062 45.31401 Alpha: 0.05 ; DF Error: 384 Critical Value of F: 2.628149 Comparison between treatments means Difference pvalue sig LCL UCL LF2 - LF3 2.9045716 0.0949 . -0.4238827 6.233026 LF2 - OP 4.7369466 0.0009 *** 1.3975073 8.076386 LF2 - VJR 3.7028437 0.0531 . -0.1579546 7.563642 LF3 - OP 1.8323751 0.3682 -1.1569616 4.821712 LF3 - VJR 0.7982721 0.9354 -2.7640364 4.360581 OP - VJR -1.0341029 0.8718 -4.6066775 2.538472
May I ask, how do i interpret this results? More specifically, which are the coefficient of slope and coefficient of intercept to be substituted into the model?
Thank you!
Best Answer
The second table of the results provides the pairwise comparison of the means, for all possible combination of groups. The comparison needs to be done on the 'Difference' column from the second table with the 'Critical value of F', and the p-values are generated based on that. Simply check the p-value column, if it is less than 0.05, it indicates the mean for the two groups are different.
In your example 'LF2 – OP' has a p-value of 0.0009, which says that the mean of groups 'LF2' and 'OP' are significantly different.
Similar Posts:
- Solved – Significant result with one-way anova test but not with Scheffe test
- Solved – Post-Hoc pairwise comparison for slope (interaction term) in R
- Solved – Post-Hoc pairwise comparison for slope (interaction term) in R
- Solved – How to generate a Randomized Complete Block Design
- Solved – How to generate a Randomized Complete Block Design