Ethical Issues in Data Science Coursera Quiz Answers

by

Coursera was launched in 2012 by Daphne Koller and Andrew Ng with the goal of giving life-changing learning experiences to students all around the world. In the modern day, Coursera is a worldwide online learning platform that provides anybody, anywhere with access to online courses and degrees from top institutions and corporations.

Join Now

Week 01: Ethical Issues in Data Science Coursera Quiz Answers

Ethical Foundations

Maximizing happiness is a defining element of which ethical framework:

  • Kantianism / Deontology
  • Virtue Ethics
  • Utilitarianism

The concept of not using others as a means to an end is a key part of which ethical framework:

  • Kantianism / Deontology
  • Virtue Ethics
  • Utilitarianism

The principle of doing what a morally virtuous person would do is a key part of which ethical framework:

  • Kantianism / Deontology
  • Virtue Ethics
  • Utilitarianism

The application of which ethical framework can include attaching numerical values to various outcomes:

  • Kantianism / Deontology
  • Virtue Ethics
  • Utilitarianism

One family in your town has more bread than it possibly can consume before it goes bad. Another family has no food and is starving, due to circumstances beyond their control. You steal a loaf of bread from the first family and give it to the second. Under which ethical framework would this clearly be unethical:

  • Kantianism / Deontology
  • Virtue Ethics
  • Utilitarianism;

In the same scenario as in #5, under which ethical framework would this most likely be ethical? (Select all that apply)

  • Kantianism / Deontology
  • Virtue Ethics
  • Utilitarianism;

Week 02: Ethical Issues in Data Science Coursera Quiz Answers

Security and Security Breaches

Which of the following is the simplest way to obtain sensitive information?

  • Obtain entry to a user’s system by guessing commonly used passwords
  • Plant a virus in the user’s computer and use it to access sensitive information
  • Break into a server room and attach a monitoring device to the network

What is a second, very simple way to obtain sensitive information?

  • Execute a denial of service attack
  • Send a phishing email that asks the user to supply sensitive information to correct an urgent problem
  • Break into a system through a misconfigured firewall

What sequence of events caused the Equifax data breach?

  • A vulnerability was discovered, and a patch quickly sent out and applied within two days as instructed, but the patch turned out to be insufficient to keep the system from being broken into
  • A vulnerability was discovered, and a patch quickly sent out and applied within two days as instructed, but the vulnerability had already been exploited by then
  • A vulnerability was discovered, and a patch quickly sent out, but the patch was not applied in a timely manner

Internet of Things (IOT) devices are particularly susceptible to security problems because:;

  • Users often don’t pay as much attention to secure passwords on these devices as on larger computers
  • Software developed for IOT systems is less likely to give careful consideration to security issues than software developed for large computer systems
  • Both A and B

The ransomware attacks on some smaller US municipalities that we read about in this lesson are said to have succeeded because:

  • The perpetrators were able to breach the firewall of the vendor who supplies software used by many municipalities
  • Staff in the municipalities clicked on a link or opened an attachment that resulted in a virus being planted in the computer system
  • Both A and B;

Week 03: Ethical Issues in Data Science Coursera Quiz Answers

Professional Ethics

The ASA Ethical Guidelines for Statistical Practice includes the statements that the ethical statistician:

  • Identifies and mitigates any preferences on the part of the investigators or data providers that might predetermine or influence the analyses/results
  • Employs selection or sampling methods and analytic approaches appropriate and valid for the specific question to be addressed
  • Does not reveal the limitations of statistical inference or possible sources of error
  • Both A and B
  • All of A, B and C

The ACM code of professional ethics indicates that computing professionals should:

  • Perform work only in areas of competence
  • Foster public awareness and understanding of computing and its consequences
  • Ensure that the public good is the central concern during all professional computing work
  • Both A and B
  • All of A, B and C

In her remarks after winning a prominent award, Dr. danah boyd described the phrase “Move Fast and Break Things” as abhorrent. This phrase has been a motto used in:

  • NASCAR stock car racing
  • The tech entrepreneurial world
  • The computer gaming industry

A study found that GitHub code created by women is downloaded more frequently than GitHub code created by men if:

  • The people doing the downloading are women
  • The people doing the downloading are men
  • The gender of the code creator is unknown
  • The topic of the code is user interface design;

According to former Uber engineer Susan Fowler, when she initially reported misconduct by her manager to humans resources (HR), no significant action was taken because:

  • She didn’t present sufficient evidence
  • There was no specific company policy that the manager had violated
  • The manager was a top performer and HR said this was the first report of misconduct

According to the article from MIT Technology Review about a draft paper by Dr. Timnit Gebru and co-authors about ethical issues in large language models, those issues include:

  • Bias in training data can lead to bias in results
  • There is no proof that the algorithms are correct
  • The financial and energy costs of running the models are very high
  • Both A and B
  • Both A and C
  • All of A, B and C

Week 04: Ethical Issues in Data Science Coursera Quiz Answers

Algorithmic Bias and Facial Recognition

The primary causes of algorithmic bias from algorithms constructed using machine learning include:

  • The training data set reflects human biases
  • The training data set includes an insufficiently diverse sample across dimensions such as gender and race
  • The training data is improperly encoded
  • Both A and B
  • Both A and C
  • All of A, B and C

Which of these is a correct conclusion that was reported in the article that we read about the analysis of COMPAS criminal sentencing algorithm:

  • The categories produced by the algorithm about how likely a person is to re-offend predict close to equally well for black people and for white people
  • The percentage of people who are given high estimates of reoffending but who don’t reoffend is roughly equal for black people and for white people
  • Both A and B are true
  • Neither A and B are true

An article that we read said that if one removes explicit knowledge of the gender of a job applicant from the information provided to a hiring screening algorithm, the algorithm may still be able to correlate gender strongly with:

  • The information that the applicant provides about their family
  • Certain words that the applicant uses, such as “captures” or “executes”
  • The length of the cover letter

Predictive policing based on computer algorithms has been banned in certain municipalities. Predictive policing involves using a computer algorithm to:

  • Estimate the likelihood that a police officer will arrest a person in a particular situation dependent on the race of the suspect
  • Determine whether a police officer should request additional police officers to come to a suspected crime scene
  • Determine which neighborhoods one should place fewer or more police officers based upon criminal histories in those neighborhoods;

The article we read about European Union consideration of facial recognition discussed consideration of a policy that would:

  • Make it illegal to collect facial recognition data
  • Make it illegal to collect facial recognition data without the consent of the person whose data it is;
  • Make it legal to collect facial recognition data in all circumstances

In articles that we read, reasons cited to prefer algorithmic decision making to human decision making include:

  • Research has shown that algorithms are less biased than humans in a variety of tasks including loan decisions and hiring screening
  • Humans are more inclined than algorithms to make decisions based on what they are comfortable with personally, such as hiring people like them
  • Algorithms are better at perceiving intangible qualities such as how well a person would work with colleagues than humans are
  • Both A and B
  • Both A and C
  • All of A, B and C

Review:

Based on our knowledge, we urge you to enroll in this course so you can pick up new skills from specialists. It will be worthwhile, we trust.

Leave a Comment