Interpretare un intervallo di confidenza

Inferenza per dati categorici in R

Andrew Bray

Assistant Professor of Statistics at Reed College

Intervalli di confidenza

Conclusione: la vera proporzione di americani felici è tra 0,705 e 0,841.

Cosa intendiamo per “confidenti”?

Dataset 1

ds1 <- filter(gss, year == 2016)

p_hat <- ds1 %>%
  summarize(mean(happy == "HAPPY")) %>% 
  pull()

SE <- ds1 %>%
  specify(response = happy, 
          success = "HAPPY") %>%
  generate(reps = 500, 
           type = "bootstrap") %>%
  calculate(stat = "prop") %>%
  summarize(sd(stat)) %>% 
  pull()

c(p_hat - 2 * SE, p_hat + 2 * SE)

0.7073114 0.8393553

confidence-interval

Dataset 2

ds2 <- filter(gss, year == 2014)

p_hat <- ds1 %>%
  summarize(mean(happy == "HAPPY")) %>% 
  pull()

SE <- ds1 %>%
  specify(response = happy, 
          success = "HAPPY") %>%
  generate(reps = 500, 
           type = "bootstrap") %>%
  calculate(stat = "prop") %>%
  summarize(sd(stat)) %>% 
  pull()

c(p_hat - 2 * SE, p_hat + 2 * SE)

0.8348831 0.9384503

Screenshot 2019-02-21 18.05.03.png

Dataset 3

ds3 <- filter(gss, year == 2012)

p_hat <- ds1 %>%
  summarize(mean(happy == "HAPPY")) %>% 
  pull()

SE <- ds1 %>%
  specify(response = happy, 
          success = "HAPPY") %>%
  generate(reps = 500, 
           type = "bootstrap") %>%
  calculate(stat = "prop") %>%
  summarize(sd(stat)) %>% 
  pull()

c(p_hat - 2 * SE, p_hat + 2 * SE)

0.7626359 0.8906974

Dataset 3

ds3 <- filter(gss, year == 2012)
p_hat <- ds3 %>%
  summarize(mean(happy == "HAPPY")) %>%
  pull()
SE <- ds3 %>%
  specify(response = happy, 
          success = "HAPPY") %>%
  generate(reps = 500, 
           type = "bootstrap") %>%
  calculate(stat = "prop") %>%
  summarize(sd(stat)) %>%
  pull()

c(p_hat - 2 * SE, p_hat + 2 * SE)

0.7626359 0.8906974

Dataset 3

ds3 <- filter(gss, year == 2012)
p_hat <- ds3 %>%
  summarize(mean(happy == "HAPPY")) %>%
  pull()
SE <- ds3 %>%
  specify(response = happy, 
          success = "HAPPY") %>%
  generate(reps = 500, 
           type = "bootstrap") %>%
  calculate(stat = "prop") %>%
  summarize(sd(stat)) %>%
  pull()

c(p_hat - 2 * SE, p_hat + 2 * SE)

0.7626359 0.8906974

Dataset 3

ds3 <- filter(gss, year == 2012)
p_hat <- ds3 %>%
  summarize(mean(happy == "HAPPY")) %>%
  pull()
SE <- ds3 %>%
  specify(response = happy, 
          success = "HAPPY") %>%
  generate(reps = 500, 
           type = "bootstrap") %>%
  calculate(stat = "prop") %>%
  summarize(sd(stat)) %>%
  pull()

c(p_hat - 2 * SE, p_hat + 2 * SE)

0.7626359 0.8906974

Dataset 3

ds3 <- filter(gss, year == 2012)
p_hat <- ds3 %>%
  summarize(mean(happy == "HAPPY")) %>%
  pull()
SE <- ds3 %>%
  specify(response = happy, 
          success = "HAPPY") %>%
  generate(reps = 500, 
           type = "bootstrap") %>%
  calculate(stat = "prop") %>%
  summarize(sd(stat)) %>%
  pull()

c(p_hat - 2 * SE, p_hat + 2 * SE)

0.7626359 0.8906974

Dataset 3

ds3 <- filter(gss, year == 2012)
p_hat <- ds3 %>%
  summarize(mean(happy == "HAPPY")) %>%
  pull()
SE <- ds3 %>%
  specify(response = happy, 
          success = "HAPPY") %>%
  generate(reps = 500, 
           type = "bootstrap") %>%
  calculate(stat = "prop") %>%
  summarize(sd(stat)) %>%
  pull()

c(p_hat - 2 * SE, p_hat + 2 * SE)

0.7626359 0.8906974

Intervalli di confidenza

Interpretazione: “Siamo al 95% sicuri che la vera proporzione di americani felici sia tra 0,705 e 0,841.”

L’ampiezza dell’intervallo dipende da

n
livello di confidenza
p

Ayo berlatih!

Inferenza per dati categorici in R