Sampling

Based on Chapter 7 of ModernDive. Code for Quiz 11.

  1. Load the R packages we will use.
  1. Quiz Questions

Question: 7.2.4 in Modern Dive with different sample sizes and repetitions

Modify the code for comparing different sample sizes from the virtual bowl

Segment 1: sample size = SEE QUIZ

1.a) Take 1150 samples of size of 28 instead of 1000 replicates of size 25 from the bowl dataset. Assign the output to virtual_samples_28

virtual_samples_28  <- bowl  %>% 
rep_sample_n(size = 28, reps = 1150)

1.b) Compute resulting 1150 replicates of proportion red

virtual_prop_red_28 <- virtual_samples_28 %>% 
  group_by(replicate) %>% 
  summarize(red = sum(color == "red")) %>% 
  mutate(prop_red = red / 28)

1.c) Plot distribution of virtual_prop_red_28 via a histogram

use labs to

ggplot(virtual_prop_red_28, aes(x = prop_red)) +
  geom_histogram(binwidth = 0.05, boundary = 0.4, color = "white") +
  labs(x = "Proportion of 28 balls that were red", title = "30") 

Segment 2: sample size = 53

2.a) Take 1150 samples of size of 53 instead of 1000 replicates of size 50. Assign the output to virtual_samples_53

virtual_samples_53  <- bowl  %>% 
rep_sample_n(size = 53, reps = 1150)

2.b) Compute resulting 1150 replicates of proportion red

virtual_prop_red_53 <- virtual_samples_53 %>% 
  group_by(replicate) %>% 
  summarize(red = sum(color == "red")) %>% 
  mutate(prop_red = red / 53)

2.c) Plot distribution of virtual_prop_red_53 via a histogram

use labs to

ggplot(virtual_prop_red_53, aes(x = prop_red)) +
  geom_histogram(binwidth = 0.05, boundary = 0.4, color = "white") +
  labs(x = "Proportion of 53 balls that were red", title = "53") 

Segment 3: sample size = 118

3.a) Take 1150 samples of size of 118 instead of 1000 replicates of size 50. Assign the output to virtual_samples_118

virtual_samples_118  <- bowl  %>% 
rep_sample_n(size = 118, reps = 1150)

3.b) Compute resulting 1150 replicates of proportion red

virtual_prop_red_118 <- virtual_samples_118 %>% 
  group_by(replicate) %>% 
  summarize(red = sum(color == "red")) %>% 
  mutate(prop_red = red / 118)

3.c) Plot distribution of virtual_prop_red_118 via a histogram

use labs to

ggplot(virtual_prop_red_118, aes(x = prop_red)) +
  geom_histogram(binwidth = 0.05, boundary = 0.4, color = "white") +
  labs(x = "Proportion of 118 balls that were red", title = "118")

ggsave(filename = "preview.png", 
       path = here::here("_posts", "2022-04-09-sampling"))

Calculate the standard deviations for your three sets of 1150 values of prop_red using the standard deviation

n = 28

virtual_prop_red_28 %>% 
  summarize(sd = sd(prop_red))
# A tibble: 1 × 1
      sd
   <dbl>
1 0.0880

n = 53

virtual_prop_red_53 %>% 
  summarize(sd = sd(prop_red))
# A tibble: 1 × 1
      sd
   <dbl>
1 0.0666

n = 118

virtual_prop_red_118 %>% 
  summarize(sd = sd(prop_red))
# A tibble: 1 × 1
      sd
   <dbl>
1 0.0443

The distribution with sample size, n = 118, has the smallest standard deviation (spread) around the estimated proportion of red balls.