In August of 2012, news outlets ranging from the Washington Post to the Huffington Post ran a story about the rise of atheism in America. The source for the story was a poll that asked people, “Irrespective of whether you attend a place of worship or not, would you say you are a religious person, not a religious person or a convinced atheist?” This type of question, which asks people to classify themselves in one way or another, is common in polling and generates categorical data. In this lab we take a look at the atheism survey and explore what’s at play when making inference about population proportions using categorical data.
To access the press release for the poll, conducted by WIN-Gallup International, click on the following link:
Take a moment to review the report then address the following questions.
While we know the findings are derived from sample statistics, it appears the claims apply to the overall population. This is supported by sentence like “the global average is 23%”. There is no mention of sample or sample size which could be misleading to the general population.
We must assume the following that the samples are unbiased, randomly chosen and independent from one another. This is reasonable to assume that the samples are independent and randomly chosen. However, the claim that the samples are unbiased should be looked into further as certain demographics within a country such as age or income group could significantly skew the results.
Turn your attention to Table 6 (pages 15 and 16), which reports the sample size and response percentages for all 57 countries. While this is a useful format to summarize the data, we will base our analysis on the original data set of individual responses to the survey. Load this data set into R with the following command.
atheism correspond to?Each row of Table 6 corresponds to summarized data from a particular country. Each row of atheism corresponds to an individual’s response from a particular country and from a particular year.
To investigate the link between these two ways of organizing this data, take a look at the estimated proportion of atheists in the United States. Towards the bottom of Table 6, we see that this is 5%. We should be able to come to the same number using the atheism data.
us12 that contains only the rows in atheism associated with respondents to the 2012 survey from the United States. Next, calculate the proportion of atheist responses. Does it agree with the percentage in Table 6? If not, why?## # A tibble: 2 x 2
## response n
## <fct> <int>
## 1 atheist 50
## 2 non-atheist 952
The calculated proportion of atheists is nearly 5% which matches the percentage of Table 6. However, the us12 data only shows “atheist” or “non-atheist” responses while the table has data for “religious person”, “not a religious person”, “convinced atheist” and “don’t know/no response”.
## [1] 4.99002
As was hinted at in Exercise 1, Table 6 provides statistics, that is, calculations made from the sample of 51,927 people. What we’d like, though, is insight into the population parameters. You answer the question, “What proportion of people in your sample reported being atheists?” with a statistic; while the question “What proportion of people on earth would report being atheists” is answered with an estimate of the parameter.
The inferential tools for estimating population proportion are analogous to those used for means in the last chapter: the confidence interval and the hypothesis test.
Conditions:
- the data must come from a random sample or a randomized experiment
- the sampling distribution of \(p\) needs to be approximately normal (as least 10 expected success and 10 expected failures)
- individual observations need to be independent (the sample size should be < 10% of the population if sampling without replacement)
All of the conditions are met.
If the conditions for inference are reasonable, we can either calculate the standard error and construct the interval by hand, or allow the inference function to do it for us.
inference(us12$response, est = "proportion", type = "ci", method = "theoretical",
success = "atheist")## Single proportion -- success: atheist
## Summary statistics:
## p_hat = 0.0499 ; n = 1002
## Check conditions: number of successes = 50 ; number of failures = 952
## Standard error = 0.0069
## 95 % Confidence interval = ( 0.0364 , 0.0634 )
Note that since the goal is to construct an interval estimate for a proportion, it’s necessary to specify what constitutes a “success”, which here is a response of "atheist".
Although formal confidence intervals and hypothesis tests don’t show up in the report, suggestions of inference appear at the bottom of page 7: “In general, the error margin for surveys of this kind is \(\pm\) 3-5% at 95% confidence”.
The margin of error is 0.0135.
## [1] 0.0135
## [1] 0.013524
inference function, calculate confidence intervals for the proportion of atheists in 2012 in two other countries of your choice, and report the associated margins of error. Be sure to note whether the conditions for inference are met. It may be helpful to create new data sets for each of the two countries first, and then use these data sets in the inference function to construct the confidence intervals.france12 <- subset(atheism, nationality == "France" & year == "2012")
japan12 <- subset(atheism, nationality == "Japan" & year == "2012")In France, the confidence interval is: ( 0.2657 , 0.3089 ); and the margin of error is: 0.02156
inference(france12$response, est = "proportion", type = "ci", method = "theoretical",
success = "atheist")## Single proportion -- success: atheist
## Summary statistics:
## p_hat = 0.2873 ; n = 1688
## Check conditions: number of successes = 485 ; number of failures = 1203
## Standard error = 0.011
## 95 % Confidence interval = ( 0.2657 , 0.3089 )
## [1] 0.02156
In Japan, the confidence interval is: ( 0.281 , 0.3329 ); and the margin of error is: 0.025872
inference(japan12$response, est = "proportion", type = "ci", method = "theoretical",
success = "atheist")## Single proportion -- success: atheist
## Summary statistics:
## p_hat = 0.3069 ; n = 1212
## Check conditions: number of successes = 372 ; number of failures = 840
## Standard error = 0.0132
## 95 % Confidence interval = ( 0.281 , 0.3329 )
## [1] 0.025872
Imagine you’ve set out to survey 1000 people on two questions: are you female? and are you left-handed? Since both of these sample proportions were calculated from the same sample size, they should have the same margin of error, right? Wrong! While the margin of error does change with sample size, it is also affected by the proportion.
Think back to the formula for the standard error: \(SE = \sqrt{p(1-p)/n}\). This is then used in the formula for the margin of error for a 95% confidence interval: \(ME = 1.96\times SE = 1.96\times\sqrt{p(1-p)/n}\). Since the population proportion \(p\) is in this \(ME\) formula, it should make sense that the margin of error is in some way dependent on the population proportion. We can visualize this relationship by creating a plot of \(ME\) vs. \(p\).
The first step is to make a vector p that is a sequence from 0 to 1 with each number separated by 0.01. We can then create a vector of the margin of error (me) associated with each of these values of p using the familiar approximate formula (\(ME = 2 \times SE\)). Lastly, we plot the two vectors against each other to reveal their relationship.
n <- 1000
p <- seq(0, 1, 0.01)
me <- 2 * sqrt(p * (1 - p)/n)
plot(me ~ p, ylab = "Margin of Error", xlab = "Population Proportion")p and me.The margin of error reaches a maximum when p = 0.5 and decreases quadratically on either side of the maximum to reach 0 at the extremities p=0 and p=1.0.
The textbook emphasizes that you must always check conditions before making inference. For inference on proportions, the sample proportion can be assumed to be nearly normal if it is based upon a random sample of independent observations and if both \(np \geq 10\) and \(n(1 - p) \geq 10\). This rule of thumb is easy enough to follow, but it makes one wonder: what’s so special about the number 10?
The short answer is: nothing. You could argue that we would be fine with 9 or that we really should be using 11. What is the “best” value for such a rule of thumb is, at least to some degree, arbitrary. However, when \(np\) and \(n(1-p)\) reaches 10 the sampling distribution is sufficiently normal to use confidence intervals and hypothesis tests that are based on that approximation.
We can investigate the interplay between \(n\) and \(p\) and the shape of the sampling distribution by using simulations. To start off, we simulate the process of drawing 5000 samples of size 1040 from a population with a true atheist proportion of 0.1. For each of the 5000 samples we compute \(\hat{p}\) and then plot a histogram to visualize their distribution.
p <- 0.1
n <- 1040
p_hats <- rep(0, 5000)
for(i in 1:5000){
samp <- sample(c("atheist", "non_atheist"), n, replace = TRUE, prob = c(p, 1-p))
p_hats[i] <- sum(samp == "atheist")/n
}
hist(p_hats, main = "p = 0.1, n = 1040", xlim = c(0, 0.18))These commands build up the sampling distribution of \(\hat{p}\) using the familiar for loop. You can read the sampling procedure for the first line of code inside the for loop as, “take a sample of size \(n\) with replacement from the choices of atheist and non-atheist with probabilities \(p\) and \(1 - p\), respectively.” The second line in the loop says, “calculate the proportion of atheists in this sample and record this value.” The loop allows us to repeat this process 5,000 times to build a good representation of the sampling distribution.
mean to calculate summary statistics.The distribution appears normal and it is fairly concentrated with no long tails. The center is at 0.09969 with a \(\sigma\) = 0.009287382.
## Min. 1st Qu. Median Mean 3rd Qu. Max.
## 0.07019 0.09327 0.09904 0.09969 0.10577 0.12981
## [1] 0.009287382
par(mfrow = c(2, 2)) command before creating the histograms. You may need to expand the plot window to accommodate the larger two-by-two plot. Describe the three new sampling distributions. Based on these limited plots, how does \(n\) appear to affect the distribution of \(\hat{p}\)? How does \(p\) affect the sampling distribution?p <- 0.1
n <- 400
p_hats2 <- rep(0, 5000)
for(i in 1:5000){
samp <- sample(c("atheist", "non_atheist"), n, replace = TRUE, prob = c(p, 1-p))
p_hats2[i] <- sum(samp == "atheist")/n
}p <- 0.02
n <- 1040
p_hats3 <- rep(0, 5000)
for(i in 1:5000){
samp <- sample(c("atheist", "non_atheist"), n, replace = TRUE, prob = c(p, 1-p))
p_hats3[i] <- sum(samp == "atheist")/n
}p <- 0.02
n <- 400
p_hats4 <- rep(0, 5000)
for(i in 1:5000){
samp <- sample(c("atheist", "non_atheist"), n, replace = TRUE, prob = c(p, 1-p))
p_hats4[i] <- sum(samp == "atheist")/n
}par(mfrow = c(2, 2))
hist(p_hats, main = "p = 0.1, n = 1040, np = 104", xlim = c(0, 0.18))
hist(p_hats2, main = "p = 0.1, n = 400, np = 40", xlim = c(0, 0.18))
hist(p_hats3, main = "p = 0.02, n = 1040, np = 20.8", xlim = c(0, 0.18))
hist(p_hats4, main = "p = 0.02, n = 400, np = 8", xlim = c(0, 0.18))As seen on the plots above, increasing n decreases the spread of the data about its center. As expected from the Margin of Error vs Population Proportion plot, a smaller p also concentrates the data about its mean due to the decrease in the margin of error.
Once you’re done, you can reset the layout of the plotting window by using the command par(mfrow = c(1, 1)) command or clicking on “Clear All” above the plotting window (if using RStudio). Note that the latter will get rid of all your previous plots.
As seen on the plots above it is sensible in the case of Australia. np = 104 and the sampling distribution is well concrentrated. It is still sensible, although slightly less so in the case of Ecuador. np = 8 so we are actually below the suggested 10. The distribution not too spread out but it appears the left tail is being cut off.
The question of atheism was asked by WIN-Gallup International in a similar survey that was conducted in 2005. (We assume here that sample sizes have remained the same.) Table 4 on page 13 of the report summarizes survey results from 2005 and 2012 for 39 countries.
Answer the following two questions using the inference function. As always, write out the hypotheses for any tests you conduct and outline the status of the conditions for inference.
Ho: \(\mu_{2005} = \mu_{2012}\) Ha: \(\mu_{2005} \neq \mu_{2012}\)
spain2005 <- subset(atheism, nationality == "Spain" & year == "2005")
spain2012 <- subset(atheism, nationality == "Spain" & year == "2012")inference(spain2005$response, est = "proportion", type = "ci", method = "theoretical",
success = "atheist")## Single proportion -- success: atheist
## Summary statistics:
## p_hat = 0.1003 ; n = 1146
## Check conditions: number of successes = 115 ; number of failures = 1031
## Standard error = 0.0089
## 95 % Confidence interval = ( 0.083 , 0.1177 )
inference(spain2012$response, est = "proportion", type = "ci", method = "theoretical",
success = "atheist")## Single proportion -- success: atheist
## Summary statistics:
## p_hat = 0.09 ; n = 1145
## Check conditions: number of successes = 103 ; number of failures = 1042
## Standard error = 0.0085
## 95 % Confidence interval = ( 0.0734 , 0.1065 )
The confidence intervals overlap so there is likely no statistical significance of the difference in proportions.
Ho: \(\mu_{2005} = \mu_{2012}\) Ha: \(\mu_{2005} \neq \mu_{2012}\)
us2005 <- subset(atheism, nationality == "United States" & year == "2005")
us2012 <- subset(atheism, nationality == "United States" & year == "2012")inference(us2005$response, est = "proportion", type = "ci", method = "theoretical",
success = "atheist")## Single proportion -- success: atheist
## Summary statistics:
## p_hat = 0.01 ; n = 1002
## Check conditions: number of successes = 10 ; number of failures = 992
## Standard error = 0.0031
## 95 % Confidence interval = ( 0.0038 , 0.0161 )
inference(us2012$response, est = "proportion", type = "ci", method = "theoretical",
success = "atheist")## Single proportion -- success: atheist
## Summary statistics:
## p_hat = 0.0499 ; n = 1002
## Check conditions: number of successes = 50 ; number of failures = 952
## Standard error = 0.0069
## 95 % Confidence interval = ( 0.0364 , 0.0634 )
In the case of the United States, the confidence intervals do not overlap so there is statistical evidence for rejecting Ho in favor of Ha and inference that there has been a change in atheism between 2005 and 2012.
Type 1 error is rejecting Ho when you shouldn’t have, and the probability of doing so is α (significance level). Out of the the 39 countries in the table, we would expect around 2 to detect a change by chance.
## [1] 2
Let’s use p = 0.5 to reflect the worst case and solve the ME for n when ME = 0.01. As shown below, we would need at least 9604 people in the sample to ensure we are within the guidelines.
## [1] 9604