Lecture 1
Behavioural data science: a multidisciplinary scientific field that aims to facilitate
understanding, prediction and change of human behaviour through the analysis of
behaviourally defined variables (gathered with modern digital technology).
understanding: constructing theories to explain behaviour.
prediction: applicating statistical models to predict behaviour.
change: developing intervention to change behaviour.
- data: representation of an observation. Data is usually structured in rows and columns.
- phenomena: patterns in data, researchers are interested in these phenomena. For example
the pattern you find in general IQ scores.
- theory: an explanatory theory is a set of principles that explains phenomena. It describes a
world in which the phenomena would follow as a matter of course.
The lexical decision task: deciding whether or not a string of letters is a real or fake word.
measures the ease with which lexical representations are activated from your
memory.
participants should try to do this as quickly as possible
dependent variables are response time and accuracy of the right recognized words
general slowing theory: all cognitive processes slow down with aging. A diffusion model
research found that drift rate doesn’t differ with age. Older adults are just way more careful,
they had bigger boundary separation, so more accurate but slower in reaction time.
4 key parameters of social contagion models
- degree centrality: speed and number of connections
- eigenvector centrality: amount of well-connected others, social status
- diffusion centrality: how well-positioned an individual is to spread and hear information
- betweenness centrality (bridging): social chameleons who connect otherwise disparate.
3 types of people in the SIR model by Kermack and McKendrink
1. susceptible
2. infected
3. recovered
balloon effect: drug trafficking spreading to new area due to new rules and laws.
cockroach effect: routes change, like cockroaches they will find their way.
Watts and Strogatz made clear that real-world networks aren’t totally ordered (not always
clear rankings between nodes) nor random (all nodes have unequal probabilities of being
connected to others).
Toothbrush problem: psychologists don’t want to use each other’s theories like its their
toothbrushes.
, Psychological theories should be constructed in a different way, there needs to be a
standard method that helps with constructing a theory.
Method:
1. identify a set of phenomena you are trying to explain
2. come up with a proto-theory
3. formalize the phenomena and proto-theory, so work out what
would happen if the theory were true.
4. evaluate how well the proto-theory explains the phenomena
5. evaluate the whole theory and maybe predict new phenomena
induction: make rules from one smaller sample.
deduction: generating predictions from general theories.
abduction: generate theories to make sense of phenomena.
Conspiracy theories are so easy to believe because they explain the
phenomena too well, they are perfectly matched to events. But, they
don’t always predict!
Theories relate to data indirectly, primarily via the empirical phenomena from which the
theories are set up.
Behavioural data science: a multidisciplinary scientific field that aims to facilitate
understanding, prediction and change of human behaviour through the analysis of
behaviourally defined variables (gathered with modern digital technology).
understanding: constructing theories to explain behaviour.
prediction: applicating statistical models to predict behaviour.
change: developing intervention to change behaviour.
- data: representation of an observation. Data is usually structured in rows and columns.
- phenomena: patterns in data, researchers are interested in these phenomena. For example
the pattern you find in general IQ scores.
- theory: an explanatory theory is a set of principles that explains phenomena. It describes a
world in which the phenomena would follow as a matter of course.
The lexical decision task: deciding whether or not a string of letters is a real or fake word.
measures the ease with which lexical representations are activated from your
memory.
participants should try to do this as quickly as possible
dependent variables are response time and accuracy of the right recognized words
general slowing theory: all cognitive processes slow down with aging. A diffusion model
research found that drift rate doesn’t differ with age. Older adults are just way more careful,
they had bigger boundary separation, so more accurate but slower in reaction time.
4 key parameters of social contagion models
- degree centrality: speed and number of connections
- eigenvector centrality: amount of well-connected others, social status
- diffusion centrality: how well-positioned an individual is to spread and hear information
- betweenness centrality (bridging): social chameleons who connect otherwise disparate.
3 types of people in the SIR model by Kermack and McKendrink
1. susceptible
2. infected
3. recovered
balloon effect: drug trafficking spreading to new area due to new rules and laws.
cockroach effect: routes change, like cockroaches they will find their way.
Watts and Strogatz made clear that real-world networks aren’t totally ordered (not always
clear rankings between nodes) nor random (all nodes have unequal probabilities of being
connected to others).
Toothbrush problem: psychologists don’t want to use each other’s theories like its their
toothbrushes.
, Psychological theories should be constructed in a different way, there needs to be a
standard method that helps with constructing a theory.
Method:
1. identify a set of phenomena you are trying to explain
2. come up with a proto-theory
3. formalize the phenomena and proto-theory, so work out what
would happen if the theory were true.
4. evaluate how well the proto-theory explains the phenomena
5. evaluate the whole theory and maybe predict new phenomena
induction: make rules from one smaller sample.
deduction: generating predictions from general theories.
abduction: generate theories to make sense of phenomena.
Conspiracy theories are so easy to believe because they explain the
phenomena too well, they are perfectly matched to events. But, they
don’t always predict!
Theories relate to data indirectly, primarily via the empirical phenomena from which the
theories are set up.
