1: reasoning skills development Flashcards
what is understanding probability
- important cognitive skill
- Frost et al (2019) allowing organisms to learn about regularities in environment
understanding probability in context of sensory input
– understanding probability in the context of sensory input involves recognising how often different events occur
What is statistical learning
- process of detecting and learning patterns or regularities in the environment that are either spatial (physical space) or temporal (time)
What does statistical learning allow
Allows individuals to predict future events based on the frequencies and sequences of stimuli they encounter
What is statistical learning an important mechanism of
An important mechanism of language learning
How do infants use statistical learning to learn language
They detect and internalise patterns and structures in spoken language.
it involves recognizing statistical regularities in sounds and syllables and words
Helps them recognise which syllables are likely to occur together and which are not, allowing them to segment words from speech
Statistical learning and language study
Saffran et al (1996)
- 8m infants could distinguish words from non words in an artificial language after being exposed to a stream of syllables based on statistical properties
What is predictive learning and an example
As children grow, statistical learning allows them to predict the next word or sound based on context
For example: if a child hears “I want to eat” they may predict next word will relate to food
3 aspects to probability understanding
- detecting regularities
- incorporating prior knowledge and new evidence
- generalising prior statistical knowledge to new situations
Xu and Garcia (2008) study
- showed infants expected and unexpected outcomes
- infants looked longer at unexpected outcome
-sensitive to base rates - looked longer when first shown context of box and experimenter drew unlikely sequence of balls
What are base rates
- underlying distribution of a set of outcomes in a population
- statistical info which is overlooked
What are base rates
- underlying distribution of a set of outcomes in a population
- statistical info which is overlooked
- base rate = general probability
Base rate neglect
Cognitive bias where people tend to ignore the general probability (base rate) of an event in favour of specific info
What does base rate neglect lead to
- faulty judgements like overestimating or underestimating likelihood of event
Explain the engineer- lawyer problem
- how people neglect base rates for making judgements
- kahenman & Tversky (1973)
In a group of 100 people: 70 are a lawyer and 30 are an engineer
Presented with a description of a randomly selected individual from the group e.g Dave js very analytic and spends lots of time on his hobbies like maths puzzles. he prefers working alone
Asked a question on what’s the probability that Dave is a lawyer
Most people rely on the specific info about Dave and ignore the base rates
A lot more lawyers than engineers in the sample, should be more inclined to pick lawyer
What assumption did the engineer- lawyer problem challenge
Challenged with the assumption that humans are always rational
Why do some people think base rate neglect occurs? Study to back it up
Reasoning in probabilities is not natural for humans
gigerenenzer and Hoffrage (1995)
- people reason more accurately when probability problems are said in frequencies
- e.g instead of 10% chance, say 10 out of 100 people.
- putting problems into frequencies reduces errors
how to identify base rate neglect in children
to identify specific skill in children, have to isolate crucial ability and remove unnecessary procedural knowledge.
- determine appropriate testing conditions e.g. suitable for age and developmental stage of child.
describe Girotto and Gonzalez (2008)
can 3-10y use base rates to correctly choose which outcome is most likely
- shown box with blue and yellow chips
- more blue chips in box than yellow
asked to predict wether blue or yellow chip would be drawn
5y children could identify more likely outcome
- 5y could use base rates to make correct judgements around 5
why do adults ignore base rates if children use them
representative heuristic
what is representative heuristic
type of initiative rule of thumb where you ignore base rate info and focus on individuating info
Guatheri and Denison (2018)- representative heuristic study
- info in study either pointed to blue (nice) robot or green (bad) robot.
- base rate 80% blue, 20% green
-base rate and specific info revealed- all children performed well - base rate and specific case info conflicted- all groups struggled to use base rate correct
-4y relied more on base rate info- prefer statistical info
- older children nd adults preferred case specific info
- as children age past 4- start to show preferences for specific info rather than base rates
Gualtieri et al (2019) base rates and eye witness
children as young as 4-5 combine base rates with eyewitness testimony accuracy- depending on reliability of eyewitness- showing reasoning skills
what is over hypothesis learning
- it is not ideal for people to ignore base rates and focus on testimony
- allows learners to make predictions about general situations
- allows abstract knowledge to be formed and applied to new circumstances
-avoids trial and error in each situation
Frost et al (2019) over hypothesis learning
- relying on prior info is helpful cue
- we come to situations having already learned a lot about the world
what is induction
- making generalisations from specific examples
- generalising info from specific instances to form general rules, e.g. learning words, casual reasoning
- instead of 2 different situations, needing 2 separate rules, more efficient to form general abstract rules that can be applied to multiple situations
Dewar and Xu (2010) and how does it demonstrate inductive reasoning
- infants observe sequences where certain patterns were followed by outcomes which were expected or unexpected
- experimental condition- object drawn from box which breaks the pattern- reaction measured by how long they looked at outcome
-longer gaze suggests these detect the difference - control condition= outcome expected based on prior patterns- infants attention less drawn- shows they find outcome predictive
demonstrates how infants use inductive reasoning to form expectations about what should happen next. showed surprise when they had no built expectation on what would happen next
over hypothesis learning study
Felsche et al (2019)
- is it a evolutionary primitive skill shared with other species or is it recently evolved and only found in humans
- 80 children aged 4-5 and 11 monkeys (6y)
- tested over hypothesis formation for rewards from buckets
- children would learn from previous draws and were more likely than chance to chose high value reward from new buckets
- monkeys always at chance as to which reward to pick
when do infants start to develop over hypothesis
1st year of life
even in infancy- children are rational constructivist learners
