Ethics and Data Science ACTUAL
QUESTIONS AND CORRECT ANSWERS
What are ethics (laws, morals and ethics) - Correct answers✔✔Laws: structures what can and
cannot be done
Morals: explaines what someone may or may not do
Ethics: the debate about what should or should not been done
other frameworks (norms and rules) - Correct answers✔✔Rules: conditioned preserved to
achieve behavior
Norms: agreed frameworks to shape behavior
Issues with data science? - Correct answers✔✔- Data not collected directly from human
subjects: complicates transparency/consent
- People are often unaware their data has been collected: complicates access/participation
- Uses of data are often secondary to original purpose: complicates purpose limitation
- Data mining is not hypothesis-based: complicates legitimate purpose. Previously, hypotheses
were formed before doing research. In data science, often hypotheses are not formed.
What's wrong with data science? - Correct answers✔✔- Predictive analytics use proxies to creat
calculated publics
- Privilege bias
- Function creep
Classic ethical theories - Correct answers✔✔1. Deontologist (Batman)
- kantianism
- rawlianism
, - moral rights
- Cosmopolitanism
2. consquentalist
- utilitarianism
3. communiterian
- communiterianism
- republicanism
- virtue ethics
- anti-theorist
General data protection right (GDPR) problems these days - Correct answers✔✔1. no interest for
locations
2. no awareness for data analytics
3. no interest for personal identification/ anonymity
Relation between data and privacy - Correct answers✔✔data enables 1-1 identification,
problems with the samaritan radar - Correct answers✔✔- bad algorithm
- they did not want response from alghoritms
- creepy actions
- only public available data
The problems with algorithmic identification - Correct answers✔✔- groups are (un)consciously
formed
- Data analytics produces unconscious grouping
QUESTIONS AND CORRECT ANSWERS
What are ethics (laws, morals and ethics) - Correct answers✔✔Laws: structures what can and
cannot be done
Morals: explaines what someone may or may not do
Ethics: the debate about what should or should not been done
other frameworks (norms and rules) - Correct answers✔✔Rules: conditioned preserved to
achieve behavior
Norms: agreed frameworks to shape behavior
Issues with data science? - Correct answers✔✔- Data not collected directly from human
subjects: complicates transparency/consent
- People are often unaware their data has been collected: complicates access/participation
- Uses of data are often secondary to original purpose: complicates purpose limitation
- Data mining is not hypothesis-based: complicates legitimate purpose. Previously, hypotheses
were formed before doing research. In data science, often hypotheses are not formed.
What's wrong with data science? - Correct answers✔✔- Predictive analytics use proxies to creat
calculated publics
- Privilege bias
- Function creep
Classic ethical theories - Correct answers✔✔1. Deontologist (Batman)
- kantianism
- rawlianism
, - moral rights
- Cosmopolitanism
2. consquentalist
- utilitarianism
3. communiterian
- communiterianism
- republicanism
- virtue ethics
- anti-theorist
General data protection right (GDPR) problems these days - Correct answers✔✔1. no interest for
locations
2. no awareness for data analytics
3. no interest for personal identification/ anonymity
Relation between data and privacy - Correct answers✔✔data enables 1-1 identification,
problems with the samaritan radar - Correct answers✔✔- bad algorithm
- they did not want response from alghoritms
- creepy actions
- only public available data
The problems with algorithmic identification - Correct answers✔✔- groups are (un)consciously
formed
- Data analytics produces unconscious grouping
