Sampling Error
- Review
- Sampling Error vs. Measurement Error
- Non-citizen Voting
Example
- Racial Discrimination in Police Shootings
October 21, 2024
Sampling Error
Example
population: full set of cases (countries, individuals, etc.) we’re interested in describing
sample: a subset of the population that we observe and measure
inference: description of the (unmeasured) population we make based on the (measured) sample
and there is uncertainty about what is true about the population, because we only measure a sample
random sampling: sampling cases from the population in a manner that gives all cases an equal probability of being chosen.
This procedure creates samples that:
The population:
The sample:
The inference:
The difference between the value of the measure for the sample and the true value of the measure for the population
\[\mathrm{Value}_{sample} - \mathrm{Value}_{population} \neq 0 \xrightarrow{then} \mathrm{sampling \ error}\]
To understand random sampling error and sampling bias, it can be useful to understand…
the sampling distribution:
Let’s now imagine that the population is students in class last week who completed the survey on rent as percent of expenditures…
To illustrate different sampling errors: We can simulate taking samples of students in class and plot the sampling distribution
histogram = Sampling distribution (the frequency of sample averages across different samples)
Blue line = Population Average (true in-class average)
Red line = Sampling Distribution Average (average of SAMPLE averages)
We want to know: what fraction of Canadian adults prefer Trudeau as PM?
There are ~31 million Canadians over the age of 18: assuming our sample is random, about how many people (\(n\)) do you think we’d have to survey to come up with sample mean and margin of error of \(\pm 1\) points that includes the population mean with a probability of 99%?
Each dot is the result of a survey of voters during the 2020 US Presidential Election. These surveys suggested that by election day voters preferred Biden to Trump by \(8.4\) percent. Biden actually won by only \(4.5\) points.
Is this sampling error? Is this a random error or a bias?
It depends: if this is going on, then sampling bias
It depends: if this is going on, then sampling bias
It depends: if there are “shy” Trump voters, then measurement bias.
In addition to winning the Electoral College in a landslide, I won the popular vote if you deduct the millions of people who voted illegally
— Donald J. Trump (@realDonaldTrump) November 27, 2016
White House senior advisor doubles down on voter fraud claims: “Voter fraud is a serious problem in this country” pic.twitter.com/DC6lVPQznz
— ABC News (@ABC) February 12, 2017
Claim: Widespread voter fraud: “14% of non-citizens voted”
Richman et al:
Discuss: Do you find this persuasive? Why or why not?
The political scientists who run the CCES survey point out:
measurement error of individuals as citizens/non-citizens, leads Richman et al to sample of “non-citizens” that include citizens and non-citizens:
Nobody who consistently reports being a non-citizen votes.
Measurement Error (of individuals’ citizenship)
\(\Downarrow produces\)
Sampling Error (sample that should be of non-citizens includes citizens)
\(\Downarrow produces\)
Measurement Error (about the population of non-citizens)
\(\Downarrow\)
authors make incorrect inference that hundreds of thousands of non-citizens vote illegally.
We’ve previously compared claims made:
Evidence we have considered suggests (1) is likely untrue.
What about (2)?
Discuss with your neighbors:
Is there racial discrimination in police use of force? \(\xrightarrow{}\)
Fryer (2019) defines racial discrimination as differential treatment of different racial groups attributable to “taste-based” (as opposed to “statistical”) discrimination.
statistical discrimination:
inequality that exists between demographic groups even though economic agents (police, consumers, workers, employers, etc.) are rational and non-prejudiced.
taste-based discrimination:
discrimination based on a preference to treat groups differently based solely on their membership in those groups
A focus on racial bias as taste-based discrimination leads social scientists to conceive of racial discrimination in policing as:
Lily Hu questions whether it makes sense to think of racial discrimination in this way:
Given the legacy of slavery, explicit and sometimes state-sponsored discrimination by race, there are large income and wealth disparities between Black and white Americans.
Does it make sense to say: “People who are poorer are more likely to be killed by the police, regardless of race. Due to racial inequalities in wealth, a greater fraction of Black Americans are poor. Some portion of the gap in police killings of white and Black Americans is explained by these differences in wealth; and they are not ‘racial discrimination’.”?
While we might dispute whether Fryer’s definition of racial discrimination is correct…
… we can still use his concept to test whether there is racial discrimination in police shootings, because it is a transparent and systematic definition.
Is there racial discrimination in police use of force? \(\xrightarrow{}\)
Fryer (2019) uses this variable to capture taste-based discrimination:
racial differences in stop outcome: differences by race in the rate of police violence against people who have been stopped by the police and who are otherwise similar on other attributes (suspect demographics, officer demographics, circumstances of the stop).
Knox et al (2020) raise concerns that these variables lack validity:
Is there racial discrimination in police use of force? \(\xrightarrow{}\)
Data sourced from police records
From police reports
Sample of ‘Stops’:
Compare people shot by police to a sample of people who “could have been shot”.
Coded from sample of police reports: to determine “otherwise similar encounters”
Do we have any concerns about…
Police reports as a source: attributes of police encounter may be misreported
Prejudiced perceptions \(\to\) downward bias in racial prejudice in police shootings: biased toward finding no or “reverse” racial discrimination
Sampling bias:
Insofar as Fryer is interested in inferring racial bias in policing in across the United States, sample of cities may suffer from sampling bias.
Is there racial bias in police use of force? \(\xrightarrow{}\)
Answer?
Knox et al show that addressing validity problem leads to large increase in estimated racial disparities in police use of force (excluding killings); likely true for shootings
Next is… Causality