Lecture 16

Models

DATA 8

Fall 2023

Announcements

• HW 6 released yesterday and due Wednesday 10/4 @ 11pm
• Project 1 due today!
• Midterm on Friday, 10/13, 7-9pm PT
• Scope: everything up to and including lecture that day
• Fill out Accommodations Form by Oct 7 at 11:59pm!

Weekly Goals

• Monday
• Array appending and for statements
• Simulation and Chances
• Wednesday
• Methods of sampling
• Distributions of large random samples
• Today
• Models that involve chance
• Assessing the consistency of the data and the model

​

Review: Distributions

Distributions

• Any random quantity has a probability distribution:
• All possible values it can take
• The probability it takes each value

​

• After repeated draws, it has an empirical distribution:
• All observed values it took
• The proportion of times it took each value

​

• After many independent draws, the empirical distribution looks more and more like the probability distribution

​

Probability Distribution of a Statistic

• Values of a statistic vary because random samples vary
• “Sampling distribution” or “probability distribution” of the statistic:
• All possible values of the statistic,
• and all the corresponding probabilities
• Can be hard to calculate
• Either have to do the math
• Or have to generate all possible samples and calculate the statistic based on each sample

Empirical Distribution of a Statistic

• Empirical distribution of the statistic:
• Based on simulated values of the statistic
• Consists of all the observed values of the statistic,
• and the proportion of times each value appeared

​

• Good approximation to the probability distribution of the statistic
• if the number of repetitions in the simulation is large

​

Assessing Models

Models

• A model is a set of assumptions about the data

​

• In data science, many models involve assumptions about processes that involve randomness
• “Chance models”

​

• Key question: does the model fit the data?

Approach to Assessment

• If we can simulate data according to the assumptions of the model, we can learn what the model predicts.

​

• We can then compare the predictions to the data that were observed.

​

• If the data and the model’s predictions are not consistent, that is evidence against the model.

Today’s Examples

Some Goals of Data Science

• Understand the world better
• Help make the world better

​

For example

• Help expose injustice
• Help counter injustice

​

The skills that you have gained empower you to do this.

First Example

• U.S. Constitution grants equal protection under the law
• All defendants have the right to due process

​

We will study a U.S. Supreme Court case in the 1960s

• A Black defendant was denied his Constitutional right to a fair jury
• The Court made incorrect and biased judgments about
• the data in the case
• the legal processes in the defendant’s original trial
• We will discuss errors and racial bias in the Court’s judgment

This case became the foundation of significant reform.

Jury Selection

All eligible jurors in a County

Jury

Panel should be representative of the eligible jurors

Jury panel

Chosen by deliberate inclusion or exclusion

US Constitution:

“right to a speedy and public trial, by an impartial jury”

Supreme Court Case

Robert Swain’s Case

• Robert Swain, a Black man, was convicted in Talladega County, AL
• He appealed to the U.S. Supreme Court
• Main reason: Unfair jury selection in the County’s trials

Eligible jurors: 26% Black

Jury:

0 Black

Panel should be representative of the eligible jurors

Jury panel:

8 out of 100 Black

Chosen by deliberate inclusion or exclusion

Supreme Court Ruling, 1965

• The Court denied Robert Swain’s appeal.

Eligible jurors: 26% Black

Panel should be representative of the eligible jurors

Jury panel:

8 out of 100 Black

About discrepancies like this, the Court wrote:

​

“the overall percentage disparity has been small”

Discussion Question

​

​

• Court’s view: 8/100 is less than 26%, but not different enough to show Black panelists were systematically excluded

​

• Question: Would 8/100 be a realistic outcome if the jury panel selection process were truly unbiased?

Sampling from a Distribution

• Sample at random from a categorical distribution

​

sample_proportions(sample_size, pop_distribution)

​

• Samples at random from the population
• Returns an array containing the empirical distribution of the categories in the sample

(Demo)

Statistical Bias

• Evidence provided by Robert Swain:

“only 10 to 15% of … jury panels drawn from the jury box since 1953 have been [Black], there having been only one case in which the percentage was as high as 23%”

• Percent of Black panelists was always lower than expected under random sampling
• Bias: when errors are systematically in one direction

A Genetic Model

Gregor Mendel, 1822-1884

A Model

• Pea plants of a particular kind
• Each one has either purple flowers or white flowers

​

• Mendel’s model:
• Each plant is purple-flowering with chance 75%,
• regardless of the colors of the other plants

​

• Question:
• Is the model good, or not?

Choosing a Statistic

• Take a sample, see what percent are purple-flowering
• If that percent is much larger or much smaller than 75, that is evidence against the model
• Distance from 75 is the key

​

• Statistic:

| sample percent of purple-flowering plants - 75 |

​

• If the statistic is large, that is evidence against the model

(Demo)

Two Viewpoints

Model and Alternative

• Jury selection:
• Model: The people on the jury panels were selected at random from the eligible population
• Alternative viewpoint: No, they were biased against black men
• Genetics:
• Model: Each plant has a 75% chance of having purple flowers
• Alternative viewpoint: No, it doesn’t

Steps in Assessing a Model

• Choose a statistic to measure discrepancy between model and data
• Simulate the statistic under the model’s assumptions
• Compare the data to the model’s predictions:
• Draw a histogram of simulated values of the statistic
• Compute the observed statistic from the real sample
• If the observed statistic is far from the histogram, that is evidence against the model

Next time