"rep""tillage""herbicide""yield" "1""Spring""None"65 "1""Fall""None"116 "1""None""None"137 "1""Spring""Applied"84 "1""Fall""Applied"132 "1""None""Applied"146...

I need you to code the problems in R output. I have attached the data tables in the files as well as the problem set


"rep""tillage""herbicide""yield" "1""Spring""None"65 "1""Fall""None"116 "1""None""None"137 "1""Spring""Applied"84 "1""Fall""Applied"132 "1""None""Applied"146 "2""Spring""None"67 "2""Fall""None"139 "2""None""None"124 "2""Spring""Applied"81 "2""Fall""Applied"126 "2""None""Applied"103 "3""Spring""None"88 "3""Fall""None"127 "3""None""None"155 "3""Spring""Applied"103 "3""Fall""Applied"138 "3""None""Applied"138 NitrogenVarietyRep1Rep2Rep3 013.84.34.9 025.14.94.2 037.46.55.6 4013.64.85.3 4026.45.95.2 4038.67.67.2 8015.86.15.3 8026.46.56.6 8038.78.58.5 12015.35.75.3 12027.07.96.2 12039.19.07.8 16015.44.35.9 16026.97.07.4 16039.29.79.3 EPP 8272 Empirical Research in Theory and Practice EPP 8272 Empirical Research in Theory and Practice Problem Set 3 3/15/2021 1 Introduction This week will be the last week of in-class R sessions. This problem set is due THREE weeks from Wednesday on Wednesday, April 7. We are going to work with two different data sets. One of the studies used a randomized complete block design and the other used a split-plot design. The details of each study are provided below. If you so choose, you are welcome to work in groups of 2-4 on this assignment. Only one person per group should submit the assignment, but the name of each group member should be clearly listed. All group members will receive the same grade. You will want to ensure that you use Type III sums of squares when conducting ANOVAs (i.e., use marginal and not sequential fits). If that sounds totally foreign, please check out the tutorial on regression (search “marginal” in that document). You will need the following packages to complete this problem set: library(tidyverse) library(car) library(lme4) library(lmerTest) library(emmeans) 1 2 Randomized complete block A turfgrass management study was initiated in to determine if tillage plan and herbicide applications to control quackgrass (a noxious weed) influenced seed production in a single, commonly grown variety of perennial ryegrass. The experiment was set up as a 3 x 2 factorial with 3 replications in a randomized complete block design (ignore interactions for this problem set, but note that factorial designs are often implemented to evaluate the interactive effect of two treatments on a single response variable). The data are in the “EPP_seed.txt” data file. • Factor 1: Tillage, 3 kinds – Spring = spring tillage/spring seeding – Fall = fall tillage/fall seeding – None = no-tillage/fall seeding • Factor 2: Herbicide, 2 levels – None = no herbicide treatment – Applied = herbicide treatment (to control quackgrass) 1. Load in the data. 2. Graph the data using a boxplot. In the plot, group the data by tillage treatment on the x-axis and then color each box by herbicide treatment. 3. Conduct an analysis of variance (ANOVA) using the lm() and Anova() (from the car package) commands (i.e., assess if tillage and herbicide explain variation in yield) 4. Conduct the same ANOVA as you did in the previous step but use the lmer() command from the lme4 package and the anova() command from the lmerTest package 5. Are there any variables you would like to explore further using pairwise comparisons? Why or why not? If so, conduct such comparisons using the lmer() model. 6. Write 4-5 sentences comparing the conclusions from the two approaches (i.e., using lm() vs. lmer()) including conclusions drawn from any pairwise comparisons you conducted. At least 1-2 of your sentences should include a conclusion written in “biologically meaningful” terms. 2 3 Split-plot design An experiment was designed to assess the effect of nitrogen fertilizer on yield (grams) from three varieties of wheat. The experimental design was a split plot with replicates in three blocks. Five rates of nitrogen fertilizer were applied to whole plots at rates of 0, 40, 80, 120 and 160 kg/ha, and the three varieties were planted in sub plots. The data are in the “EPP_yield.txt” data file. 1. Load in the data. 2. You might notice the data do not have one observation per row (which R expects when fitting linear models). Convert the data from “wide” to “long” format using R. 3. Graph the data using a boxplot. In the plot, group the data by the whole plot factor (one panel for each level of Nitrogen) and display the sub plot factor on the x-axis. 4. Conduct an analysis of variance (ANOVA) using the aov() command (i.e., assess if Nitrogen and Variety explain variation in yield). Treat Nitrogen as a factor for all analyses. 5. Conduct the same ANOVA as you did in the previous step but use the lmer() command from the lme4 package and the anova() command from the lmerTest package. Note: if you get an error that says boundary (singular) fit: see ?isSingular, ignore it. 6. Are there any variables you would like to explore further using pairwise comparisons? Why or why not? If so, conduct such comparisons using the lmer() model. 7. Write 4-5 sentences comparing the conclusions from the two approaches (i.e., using aov() vs. lmer()) including conclusions drawn from any pairwise comparisons you conducted. At least 1-2 of your sentences should include a conclusion written in “biologically meaningful” terms. 3 Introduction Randomized complete block Split-plot design