One-sample

proportion tests
HYPOTHESIS TESTING IN R

Richie Cotton
Data Evangelist at DataCamp
Chapter 1 recap
Is a claim about an unknown population proportion feasible?
Standard error of sample statistic calculated using bootstrap distribution.

This was used to compute a standardized test statistic, ...

which was used to calculate a p-value, ...

which was used to decide which hypothesis made most sense.

Here, we'll calculate the test statistic without using the bootstrap distribution.

HYPOTHESIS TESTING IN R
Standardized test statistic for proportions
p: population proportion (unknown population parameter)

p^: sample proportion (sample statistic)

p0 : hypothesized population proportion

p^ − mean( p^) p^ − p
z= =
standard error( p^) standard error( p^)
Assuming H0 is true, p = p0 , so
p^ − p0
z=
standard error( p^)

HYPOTHESIS TESTING IN R
Easier standard error calculations
SE(x̄child − x̄adult ) ≈ √
s2child s2adult
+
nchild nadult

SE p^ = √
p0 ∗ (1 − p0 )
n
Assuming H0 is true,

p^ − p0
z=
√
p0 ∗ (1 − p0 )
n
^ and n) and the hypothesized parameter (p0 ).
This only uses sample information ( p

HYPOTHESIS TESTING IN R
Why z instead of t?
(x̄child − x̄adult )
t=
√
s2child s2adult
+
nchild nadult

s is calculated from x̄, so x̄ is used to estimate the population mean and to estimate the
population standard deviation.

This increases uncertainty in our estimate of the population parameter.

t-distribution has fatter tails than a normal distribution.

This gives an extra level of caution.

p^ only appears in the numerator, so z-scores are fine.

HYPOTHESIS TESTING IN R
Stack Overflow age categories
H0 : The proportion of SO users under thirty is equal to 0.5.

HA : The proportion of SO users under thirty is not equal to 0.5.

alpha <- 0.01

stack_overflow %>%
count(age_cat)

# A tibble: 2 x 2
age_cat n
<chr> <int>
1 At least 30 1050
2 Under 30 1216

HYPOTHESIS TESTING IN R
Variables for z
p_hat <- stack_overflow %>%
summarize(prop_under_30 = mean(age_cat == "Under 30")) %>%
pull(prop_under_30)

0.5366

p_0 <- 0.50

n <- nrow(stack_overflow)

2266

HYPOTHESIS TESTING IN R
Calculating the z-score
p^ − p0
z=
√
p0 ∗ (1 − p0 )
n

numerator <- p_hat - p_0

denominator <- sqrt(p_0 * (1 - p_0) / n)
z_score <- numerator / denominator

3.487

HYPOTHESIS TESTING IN R
Calculating the p-value
Two-tailed ("not equal")

p_value <- pnorm(z_score) +

pnorm(z_score, lower.tail = FALSE)

p_value <- 2 * pnorm(z_score)

Left-tailed ("less than")
0.000244
p_value <- pnorm(z_score)

p_value <= alpha

Right-tailed ("greater than")

TRUE
p_value <- pnorm(z_score, lower.tail = FALSE)

HYPOTHESIS TESTING IN R
Let's practice!
HYPOTHESIS TESTING IN R
 Two-sample
proportion tests
HYPOTHESIS TESTING IN R

Richie Cotton
Data Evangelist at DataCamp
Comparing two proportions
H0 : The proportion of SO users who are hobbyists is the same for those under thirty as those
at least thirty.

H0 : p≥30 − p<30 = 0

HA : The proportion of SO users who are hobbyists is different for those under thirty as those
at least thirty.

HA : p≥30 − p<30 ≠ 0

alpha <- 0.05

HYPOTHESIS TESTING IN R
Calculating the z-score
( p^≥30 − p^<30 ) − 0
z=
SE( p^≥30 − p^<30 )

SE( p^≥30 − p^<30 ) = √

p^ × (1 − p^) p^ × (1 − p^)
+
n≥30 n<30

p^ is a pooled estimate for p (common unknown proportion of successes).

n≥30 × p^≥30 + n<30 × p^<30
p^ =
n≥30 + n<30
We only need to calculate 4 numbers: p^≥30 , p^<30 , n≥30 , n<30 .

HYPOTHESIS TESTING IN R
Getting the numbers for the z-score
stack_overflow %>% z_score
group_by(age_cat) %>%
summarize( -4.217
p_hat = mean(hobbyist == "Yes"),
n = n()
)

# A tibble: 2 x 3
age_cat p_hat n
<chr> <dbl> <int>
1 At least 30 0.773 1050
2 Under 30 0.843 1216

HYPOTHESIS TESTING IN R
Proportion tests using prop_test()
library(infer)
stack_overflow %>%
prop_test(
hobbyist ~ age_cat, # proportions ~ categories
order = c("At least 30", "Under 30"), # which p-hat to subtract
success = "Yes", # which response value to count proportions of
alternative = "two-sided", # type of alternative hypothesis
correct = FALSE # should Yates' continuity correction be applied?
)

# A tibble: 1 x 6
statistic chisq_df p_value alternative lower_ci upper_ci
<dbl> <dbl> <dbl> <chr> <dbl> <dbl>
1 17.8 1 0.0000248 two.sided 0.0605 0.165

HYPOTHESIS TESTING IN R
Let's practice!
HYPOTHESIS TESTING IN R
 Chi-square test of
independence
HYPOTHESIS TESTING IN R

Richie Cotton
Data Evangelist at DataCamp
Revisiting the proportion test
library(infer)
stack_overflow %>%
prop_test(
hobbyist ~ age_cat,
order = c("At least 30", "Under 30"),
alternative = "two-sided",
correct = FALSE
)

# A tibble: 1 x 6
statistic chisq_df p_value alternative lower_ci upper_ci
<dbl> <dbl> <dbl> <chr> <dbl> <dbl>
1 17.8 1 0.0000248 two.sided 0.0605 0.165

HYPOTHESIS TESTING IN R
 Independence of variables
Previous hypothesis test result: there is evidence that the hobbyist and age_cat variables
have an association.

If the proportion of successes in the response variable is the same across all categories of the
explanatory variable, the two variables are statistically independent.

1 Response and explanatory variables are defined in "Introduction to Regression in R", Chapter 1.

HYPOTHESIS TESTING IN R
Job satisfaction and age category
stack_overflow %>% stack_overflow %>%
count(age_cat) count(job_sat)

# A tibble: 2 x 2 # A tibble: 5 x 2
age_cat n job_sat n
<chr> <int> <fct> <int>
1 At least 30 1050 1 Very dissatisfied 159
2 Under 30 1211 2 Slightly dissatisfied 342
3 Neither 201
4 Slightly satisfied 680
5 Very satisfied 879

HYPOTHESIS TESTING IN R
Declaring the hypotheses
H0 : Age categories are independent of job satisfaction levels.

HA : Age categories are not independent of job satisfaction levels.

alpha <- 0.1

Test statistic denoted χ2 .

Assuming independence, how far away are the observed results from the expected values?

HYPOTHESIS TESTING IN R
Exploratory visualization: proportional stacked bar plot
ggplot(stack_overflow, aes(job_sat, fill = age_cat)) +
geom_bar(position = "fill") +
ylab("proportion")

HYPOTHESIS TESTING IN R
Chi-square independence test using chisq_test()
library(infer)
stack_overflow %>%
chisq_test(age_cat ~ job_sat)

# A tibble: 1 x 3
statistic chisq_df p_value
<dbl> <int> <dbl>
1 5.55 4 0.235

Degrees of freedom:

(No. of response categories − 1) × (No. of explanatory categories − 1)

(2 − 1) ∗ (5 − 1) = 4

HYPOTHESIS TESTING IN R
Swapping the variables?
ggplot(stack_overflow, aes(age_cat, fill = job_sat)) +
geom_bar(position = "fill") +
ylab("proportion")

HYPOTHESIS TESTING IN R
chi-square both ways
library(infer) library(infer)
stack_overflow %>% stack_overflow %>%
chisq_test(age_cat ~ job_sat) chisq_test(job_sat ~ age_cat)

# A tibble: 1 x 3 # A tibble: 1 x 3
statistic chisq_df p_value statistic chisq_df p_value
<dbl> <int> <dbl> <dbl> <int> <dbl>
1 5.55 4 0.235 1 5.55 4 0.235

Ask Not

Are the variables X and Y independent? Is variable X independent from variable Y?

HYPOTHESIS TESTING IN R
 What about direction and tails?
args(chisq_test)

function (x, formula, response = NULL, explanatory = NULL, ...)

Observed and expected counts squared must be non-negative.

chi-square tests are almost always right-tailed. 1

1Left-tailed chi-square tests are used in statistical forensics to detect is a fit is suspiciously good because the
data was fabricated. Chi-square tests of variance can be two-tailed. These are niche uses though.

HYPOTHESIS TESTING IN R
Let's practice!
HYPOTHESIS TESTING IN R
 Chi-square
goodness of fit tests
HYPOTHESIS TESTING IN R

Richie Cotton
Data Evangelist at DataCamp
Purple links
You search for a coding solution online and the first result link is purple because you
already visited it. How do you feel?

purple_link_counts <- stack_overflow %>%

count(purple_link)

# A tibble: 4 x 2
purple_link n
<fct> <int>
1 Hello, old friend 1330
2 Amused 409
3 Indifferent 426
4 Annoyed 290

HYPOTHESIS TESTING IN R
 Declaring the hypotheses
hypothesized <- tribble( # A tibble: 4 x 2
~ purple_link, ~ prop, purple_link prop
"Hello, old friend", 1 / 2, <chr> <dbl>
"Amused" , 1 / 6, 1 Hello, old friend 0.5
"Indifferent" , 1 / 6, 2 Amused 0.167
"Annoyed" , 1 / 6 3 Indifferent 0.167
) 4 Annoyed 0.167

2
H0 : The sample matches with the The test statistic, χ , measures how far
hypothesized distribution. observed results are from expectations in
each group.
HA : The sample does not match with the
hypothesized distribution. alpha <- 0.01

1 tribble is short for "row-wise tibble"; not to be confused with the alien species from Star Trek

HYPOTHESIS TESTING IN R
Hypothesized counts by category
n_total <- nrow(stack_overflow) # A tibble: 4 x 3
hypothesized <- tribble( purple_link prop n
~ purple_link, ~ prop, <chr> <dbl> <dbl>
"Hello, old friend", 1 / 2, 1 Hello, old friend 0.5 1228.
"Amused" , 1 / 6, 2 Amused 0.167 409.
"Indifferent" , 1 / 6, 3 Indifferent 0.167 409.
"Annoyed" , 1 / 6 4 Annoyed 0.167 409.
) %>%
mutate(n = prop * n_total)

HYPOTHESIS TESTING IN R
Visualizing counts
ggplot(purple_link_counts, aes(purple_link, n)) +
geom_col() +
geom_point(data = hypothesized, color = "purple")

HYPOTHESIS TESTING IN R
chi-square goodness of fit test using chisq_test()
hypothesized_props <- c( library(infer)
"Hello, old friend" = 1 / 2, stack_overflow %>%
Amused = 1 / 6, chisq_test(
Indifferent = 1 / 6, response = purple_link,
Annoyed = 1 / 6 p = hypothesized_props
) )

# A tibble: 1 x 3
statistic chisq_df p_value
<dbl> <dbl> <dbl>
1 44.0 3 0.00000000154

HYPOTHESIS TESTING IN R
Let's practice!
HYPOTHESIS TESTING IN R