Russell et al (2017) studied the influence of natural and synthetic caffeine on the heart rate of 48 basketball players who competed at the college level. Participants were randomly assigned to three treatment groups where they ingested either naturally-extracted caffeine, synthetic caffeine of a maltodexrin placebo as control. Heart rate measurements during standardised exercise were then collected 30 minustes or 60 minutes after ingestion of the supplements. A two-factor ANOVA was performed with source (natural, synthetic, control) and time (30 minc t30, 60 min: t60) as independent variables and heart rate expressed as a percentage of maximum (\%HRMax) as the dependent variable. glimpse(caff) HA Rows: 48 Ant Columns: 3 HA 5 Source ifct> control, control, control, control, control, control, control, Hin S. Time efct>t30,+30,t30,t30,t30,t30,t30,t30,t30,t30,+30,t30,+30. HA 5 HR obls 25,6,35,6,43.9,22,7,42,9,40.5,41.3,56.7,31.7,69.0,38. caff sse ggplot ()+ aes (x= Time, y HR, fill = Source) + geom boxplot (cosf=10)+labs(y= "HR (x HrMax ) ") Source Time A \# A tibble: 65 * Groups: Time [2] Nin Time Sounce mean sd in If sfct> eint> fit =aov(HR - Source Time, data =caff) autoplot(fit, which =1:2 ) tibrary (emmeans) p1 \& emimip(fit. Source Time) + themeGleachendosition = "tori") tibrary(emmeans) p1 - emmip(fit, Source - Tine) + theme(legend, position = "top") p2 c-emmip(fit, Time - Source) + theme(legend, position = "top") gridi arrange (p1,p2, nrow =1) # * R Results are averaged over some or all of the levels of: Time. Source \#\# P value adjustment: bonferroni method for 4 tests With reference to the post-hoc test output, provide a short summary of the key findings for this experiment. In your answer you should refer to thi estimated pairwise differences and whether or not the difference is significant