the dark side of information proliferation

climate change example and attention economy

• Less than 20% of the population mention diet (pilot data)

• Most people think littering is more important than diet (Truelove & Parks, 2012)

• A meat based diet is equivalent to driving an extra 20 miles in your car a day (Weber & Matthews, 2008)

• deficit-model of science communication suggests that if you were told this your beh would change, but it doesnt

• inclined not to believe - have lots of info in head, find some counter evidence which disqualifies it

what is the deficit model of science communication

• taught the correct thing to do and you will do it

• reason you do the wrong thing is because of a deficit in understanding - so if given the correct info will believe it and act accordingly

information proliferation

• the capacity to access and contribute to a growing quantity of information.

• “A wealth of information creates a poverty of attention.” Herbert Simon, cog control is ability to control your own actions, have lots of info coming in, makes it difficult to know what to pay attention to

• Some estimates that we produce more information in two days now than in the last five millennia.

• This places information under the influence of an attention economy.

• Lots of info out there, can be described in various ways, has spec effects on attention, makes it hard to know what to pay att to. everyone is competing for attention as there are so many things to pay attention to

cognitive selection

• process for selecting information based on features valued by cognition. This depends on what information is searched for, attended to, comprehended, encoded, and later reproduced.

lifecycle of info in modern world

• people say things, you decide what to look at/listen to

• some of it makes it into memory, later reproduce this info and become the speaker

• listener hears things, encodes some of it, then reproduce some of it

• however it is not necessary to encode info (eg may retweet a tweet)

role of algorithms

• algorithms pay attention to who’s listening and reproduce things that people are more likely to listen to

• Like provocative and argumentary material so more likely to produce that 

• Lots of info, select some stuff, system learns what you pay attention to

• If small bias in system, whole system pays more attention to certain kinds of things

selection process

• lots of things are produced by the speaker, but only some things get listened to and encoded

evolution of american english

• looked at concreteness (how easily you can visualise something) - eg china would be 5 (the top) whereas essentialness would be 1 (lowest)

• women slightly more concrete

• concrete words are more easily recalled in memory tasks and is more interesting and easier to understand. also more readily learned by first and second language learners

• found that lang is getting more concrete

why is lang getting more concrete

• language competition

• More competition for people’s attention means the use of more memorable words are favored, and concrete words are more memorable

• more resistant to be lost from language 

• Speak language, listener listens and encodes, as its easy to remember will reproduce 

four forces of cog selection

• Selection for belief consistent info

• Selection for negative info

• Selection for social info

• Selection for predictive info

does more information make us smarter

• no

• if choose what you listen to, curate what you are hearing so you only hear what you want to hear

• selected info makes you more confident in what you already believe

• algorithm amplifies this bias

confirmation bias

• selection for belief consistent info

• also known as Motivated reasoning

• bias assimilation (the tendency to interpret information in a way that supports the desired conclusion - these are all self serving biases

• so whenever given more info, initial position becomes stronger

confirmation bias and climate change

• corner et al 2012

• gave spetics and non-sceptics of climate change an article about it that was either sceptical or non-sceptical

• Sceptics say irish times (pro cc) not convincing but scotsman (not cc) is.

• Reverse for non-sceptics

• Already have self-serving bias to not believe in those things 

polarisation of capital punishment and vaccines

• Lord, Ross, Lepper, 1979

• presented mixed info supporting both sides

• pps did not become more neutral, but more polarised

• amplified what they already believed

• same found for vaccine (nan and daily, 2015)

problem of self-selected info

• can impair accuracy and understanding

• just gain confidence in what you already know

negativity bias

• selection for negativity info

• people more likely to click on news articles were someone is being hurt - form of selection, drives news/social media in a specific direction

• also loss aversion - fear losses more than we fear relavant gains

what could negativity bias be due to

• evo selecton for survival - if someone hurts themself, want to know how so you know how to avoid it

are you more or less likely to share negative info?

• more likely

social risk amplification

• Humans distort info by communicating it

• Hazards interact with human biases to amplify public hysteria

phd student hills and jagiello about social risk

• gave people a doc to read which had different facts of antibac agents

• then retold to the next person

• found that when it was told over and over again it became more distorted and negative

• amplified the negative aspect of info

jagiello and hills nuclear power vs food additives

• same thing done but with food additives and nuclear power

• then half recieve reintroduction of balanced info

• this has no effect on the info reproduced

• people become hyperofcused on neg consequences

• when share, more likely to repeat it - neg bias built into retelling of info

examples in real world

• hurricanes

• ebola

• covid

social bias

• selection for social info

• social risk ampliciation comes from other people, amplifies negativity

• slt - imitate others, just watching another person led to similiar beh

social monitoring

• paying attention to what other people think/do

• phones promote hyper natural sm

• evo bias - eg want to know what other people in the tribe are doing

• many people reject what they know when confronted with people who feel differently (asch)

• will also cling onto fringe beliefs when at least on other person shares those beliefs, makes them more confident in their beliefs

• people turn off executive (pre-frontal) control when they get advice from other people - dont engage in effortful cog control

social herding

• pps picked music and listened to a sample online

• could see what other people chose (social influence) or couldnt (independent)

• more unequal in independent condition - less likely to just download what other people are listening to

influence of another classroom

• social influence groups less like other social influence groups

• independent groups more like other independent groups

• so social influece amplifies noise - noise is started with whoever is in that room

should we work separately or independently to solve hard problems

• Mason, Jones, & Goldstone (2008)

• too much connectivity makes harder problems harder

• People are connected together in different configurations(see neighbors’ performance). They are trying to choose the best guess (a number) to get the highest score

• When the problem is easy, more neighbors is good, can all talk and share info. Eg where is the toilet

• When the problem is hard, fewer neighbors is better. Split up look in dif directions to find solution. Preserve independence in search, with other, if see where they look then you will also try and look there.

pattern bias

• Selection for predictive information 

• Superstition and overfitting

replication crisis

• scientific crisis that refers to the difficulty or impossibility of reproducing the results of many scientific studies.

• it's a growing concern that the scientific literature is accumulating erroneous studies.

why is the replication crisis happening

• product of info proliferation

• if probability of success is p then probability of at least one success approaches 1 as n goes to infinity.

• When a researcher tests multiple hypotheses, they correct for this using Bonferonni corrections.

• But if multiple researchers test multiple hypotheses, there is no correction.

• Ioannidis (2005), more researchers testing more hypotheses means an increase in the absolute number of type I (false positive) errors

• Any selection bias in publishing will amplify the proportional amount of errors

consequences of each of these biases

• Selection for belief consistent information—> Extremism

• Selection for negative information —> Fear/Anxiety

• Selection for social information —> Herding

• Selection for predictive information —> Replication crisis, Risk seeking

how does info proliferation influence the evo of information

• by increasing competition for attention

• reducing information’s generation time - the time it takes for information to move from one mind to another.

why does misinformation have an advantage in competitive environments

• it is freed from the constraints of being truthful, allowing it to adapt to cognition’s biases for distinctive and emotionally appealing information

• lies proliferate faster than the truth

what is the tendency to select like-minded individuals in decision making associated with?

• groupthink

what is social risk amplification particularly prominent for

• dread risks - unpredictable, catastrophic, and indiscriminant risks to life and limb such as plane crashes, nuclear disasters, epidemics, and terrorism

precautionary principle

•  a propensity to base decisions about new technology on their potential downside risks without considering their potential benefits

mindspace article

Messenger - perceived authority of the source

Incentive - monertary usually

Reference points matter - care about final outcome?

Losses loom larger than gains

Overweight small probabilities - we are prone to overestimate the probability of unlikely but easy to imagine or recall events, such as winning the lottery.

Allocate money to discrete mental amounts - label particular benefits eg winter fuel payment. People view same money differently

Inconsistently live for today at the expense of tomorrow - prefer smaller more immediate payoffs

Norms - social and cultural

Defaults - option if no active choice is made

Salience - beh influence by what our attention is drawn to

Priming - activation of knowledge in memory makes it more accessible and therefore more influential in processing new stimuli

Affect - act of experiencing emotion

Commitment - procrastinate and delay taking decisions that are likely to be in our long-term interests

Ego - behave in a way that supports the impression of a positive and consistent self-image.