1 ## load the data
   2 load("data/studrecords.rdata")
   3 
   4 ## Test the null hypothesis that the population mean math score
   5 ## is 500 against a two-sided alternative. Would you accept or
   6 ## reject at a 0.05 significance level?
   7 
   8 
   9 
  10 ## Test the null hypothesis that the mean math score is equal
  11 ## to the mean verbal (sat.v) score against a two sided hypothesis.
  12 
  13 
  14 
  15 ## Test the null hypothesis that the verbal and the math score
  16 ## is equal in each student.
  17 
  18 
  19 
  20 ## Wilcoxon signed-rank test
  21 pre.test <- c(17,12,20,12,20,21,23,10,15,17,18,18)
  22 post.test <- c(19,25,18,18,26,19,27,14,20,22,16,18)
  23 
  24 wilcox.test(pre.test,post.test,paired = T)
  25 
  26 ## exercises
  27 ## load the data frame normtemp from the file temperature.rdata;
  28 ## it contains the body temperature of several individuals,
  29 ## the gender and the heart rate
  30 
  31 
  32 ## test if the temperature is different in male (coded as 1)
  33 ## and female (coded as 2), use the appropriate test.
  34 
  35 
  36 ## test again, compare the results of the t test and the wilcoxon.
  37 
  38 
  39 
  40 ## plot the respective boxplots!
  41 
  42 
  43 
  44 
  45 
  46 ## add another layer
  47 ggplot(normtemp, aes(x = factor(gender), y = temperature)) +
  48     geom_boxplot() +
  49     geom_point()
  50 
  51 ggsave("boxplotpoint.png")
  52 
  53 
  54 ggplot(normtemp, aes(x = factor(gender), y = temperature)) +
  55     geom_boxplot() +
  56     geom_jitter()
  57 
  58 
  59 ggsave("boxplotjitter.png")
  60 
  61 
  62 ggplot(normtemp, aes(x = factor(gender), y = temperature)) +
  63     geom_boxplot() +
  64     geom_jitter(position=position_jitter(width=0.1))
  65 
  66 
  67 ggsave("boxplotjitter2.png")
  68 
  69 ggplot(normtemp, aes(x = factor(gender), y = temperature)) +
  70     geom_boxplot() +
  71     geom_point(stat = "summary", fun.y = "mean")
  72 
  73 ggsave("boxplotmean.png")
  74 
  75 
  76 ggplot(normtemp, aes(x = factor(gender), y = temperature)) +
  77     geom_boxplot() +
  78     geom_point(stat = "summary", fun.y = "mean",
  79                size = 5, colour = "red")
  80 
  81 ggsave("boxplotmean2.png")
  82 
  83 
  84 
  85 
  86 ## Exercises
  87 
  88 ## load the data set GaltonFamilies from the package HistData 
  89 
  90 
  91 ## make a boxplot of childHeight dependent on gender
  92 ## add the means using geom_point()
  93 
  94 
  95 
  96 
  97 
  98 ## do the respective t-test: what is the (trivial) null hypothesis,
  99 ## what is appropriate test, what is the test result and what is
 100 ## the conclusion?
 101 
 102 
 103 
 104 ## bivariate plots
 105 ggplot(GaltonFamilies, aes(x=mother,y=father)) +
 106     geom_point() 
 107 
 108 
 109 ggsave("scatter.png")
 110 
 111 
 112 ## adding a trend line
 113 ggplot(GaltonFamilies, aes(x=mother,y=father)) +
 114     geom_point() +
 115     geom_smooth()
 116 
 117 ggsave("scattertrend1.png")
 118 
 119 
 120 ## adding a trend line
 121 ggplot(GaltonFamilies, aes(x=mother,y=father)) +
 122     geom_point() +
 123     geom_smooth(method="lm")
 124 
 125 ggsave("scattertrend2.png")
 126 
 127 
 128 ## Exercises
 129 ## use again the GaltonFamilies data set; produce a scatter plot
 130 ## of childHeight vs midparentHeight
 131 
 132 
 133 
 134 
 135 
 136 ##add a trend line by using geom_smooth() without any
 137 ## arguments. Which method is used?
 138 
 139 
 140 
 141 
 142 
 143 ## add a second trend line, this time a linear one!
 144 
 145 
 146 
 147 
 148 
 149 
 150 ## now use the aesthetic colour in the first line of the plot
 151 ## definition. What happens?
 152