Big Questions in Big Data

Today

1. Big Data and Psychology

2. The Promise of Big Data

3. The Peril of Big Data

Big Data and Psychology

Question 1. Are classical psychological research methods outdated?

- random sampling

- laboratory experiments

- self-report questionnaire

:

- random sampling : not necessary if researchers can capture all of everyone’s data; likely not necessary in big data research

- laboratory experiments : can show causation but may lack external validity; big data uses A/B testing

- self-reportquestionnaires : can be biased ( Ps don’t know the answer or don’t tell the truth ) ; when accurate, the most straightforward way at getting at a person’s own knowledge, insight, perspective; used to test how effective algorithms are

- ex. targeted ads based on big 5

Are Classical Psychological Research Methods Outdated ?

- Mayer-Schonberger & Cukier, p. 30-31

- Stevens-Davidowitz, p. 278

Psychological Research

If big data research becomes as ubiquitous as experts suggest that it will

1. Research psychologists may no longer exist.

2. Research psychologists may change from the classical psychological approach to a big data approach.

Big Data and Psychology

- Question 2a. Is theory dying ?

- 2b. If so, is that a bad thing ?

Is theory Dying ?

1. Research psychologists already take untestable theories less seriously than testable ones.

2. Theory is not necessary for finding which foods to stock before hurricanes or choosing which campaign ad will be the most effective.

Is theory is dying, Is it a bad thing ?

We, as human, have a deep longing to understand not only what is happening but why it is happening... but

1. our theories may be incorrect.

2. there may be no reason why certain things happen.

The promise of Big Data

Big Data has the power to Dispel myths from popular culture _ Mayer-Schonbreger & Cukier

- Danish cell phone owners from 1987 to 1995, n = 358,403

- Nationwide registry of cancer patients, n = 10,729

- Investigated whether owning a cell phone (vs. not) was associated with increased cancer risk and whether amount of time owning a cell phone was correlated with risk.

 : no increase in cancer risk associated w/ cell phones

Big Data has the power to address climate change

: Research in the journal Science uses big data to estimate how much of the Earth could be planted with trees and discusses the implications of planting trees for the amount of carbon in the Earth’s atmosphere.

Big Data saves lives _ Mayer-Schonberger & Cukier 

: analyzing data from premature babies for precursors to infection can help infections be treated sooner and with less medical intervention

ex. Very constant vital signs (heart rate, respiration, blood pressure, and blood oxygen) can signal that a major infection is likely to occur soon.

The Peril of Big Data

Consent

: because big data analysis often involves data exhaust ( and researchers do not always know how they will use the data when it is originally collected ), it does not always involve consent

Our Data may be used to manipulate us

https://www.youtube.com/watch?v=mrnXv-g4yKU

Our data may be used to convict us of crimes we have not yet committed _ Mayer-Schonberger & Cukier

ex. movie : Minority Report

https://www.youtube.com/watch?v=1ysXMu73SYo