Chapter 8 Flashcards
Külpes introspection
tried to research problem solving through introspection.
participants focused on their mental experience
stated that some thought are imageless
thorndike’s behavioral approach to problem solving
concentrated on behavior
problem solving is trial and error
operant conditioning
gestalt approach
there is insight to find new solutions;
and reproductive thinking to apply past solutions
insight
The reorganizing or restructuring of the elements of the problem situation in such a way as to provide a solution. Also known as productive thinking.
two string problem
a problem where people must repurpose tools to solve it
functional fixedness
The inability to use an object appropriately in a given situation because of prior experience of using the object in a different way. Is a “block” to problem solution
mental set
A term to describe the rote application of one successful method to solve a problem which makes one ‘blind’ to an alternative and possibly much simpler method.
stages in the information processing approach (IPA)
general problem solver
- representing the problem
- selection of operators
- implementation of the selected operators
- evaluation of the current state
- representing the problem (IPA)
a !problem space! is constructed which includes both the initial state and the goal state, the instructions and the constraints on the problem and all relevant information retrieved from long-term memory. To assist such representation, symbols, lists, matrices, tree diagrams, graphs and visual imagery can all be used. This first stage reflects the assumption that problemsolving can be regarded as a form of search in a space consisting of all possible states of the problem.
- selection of operators (IPA)
operators are actions that will achieve a goal, and are used for transforming the initial state
- implementation of the chosen operators (IPA)
this results in a new current state within the problem space
- evaluation of the current state (IPA)
if it corresponds to the goal, a solution is reached
heuristics
Methods or strategies which often lead to a problem solution but are not guaranteed to succeed.
most problems are solved this way.
have to be distinguished from algorithms
algorithm
a method that will always yield a solution, sooner or later. You might have an algorithm to solve 21x123, a strategy that will always work
state-action tree
A diagram showing all the possible sequences of actions and intermediate states which can be constructed if the problem is well-defined.
problem reduction
An approach to problem solving that converts the problem into a number of sub-problems, each of which can be solved separately.
hill climbing
operators are selected which make the current state of the problem as similar as possible to the goal state, so the problemsolver is likely to choose moves that reduce the distance between the current and goal states
means-ends analysis
A general heuristic where a sub-problem is selected that will reduce the difference between the current state and the goal state.
e.g.: making travel plans
difficulties in applying problem solving strategies
people have trouble making moves that increase the distance between current state and goal state, although they might be necessary
for very complex problems we need to shift strategies
inappropriate use of heuristics
impasse
A sort of mental ‘blank’ experienced when trying to solve a problem, which is accompanied by a subjective feeling of not knowing what to do.
Can be caused by representing the problem wrongly
representational-change explanation
reorganizing and restructuring the elements of a problem. ´The includes “chunk decomposition” and “constraints relaxation”
chunk decomposition
Familiarity with certain objects or events can result in
the abstraction of a pattern or chunk of components or features, and to solve a problem we may need to change the way this information has been encoded. the probability of reencoding a piece of information is an inverse function of how tightly the information is chunked in the current representation
constraint relaxation
We natural assume constraints to the solution of a task, that might not actually be there. constraint relaxation can help overcome an impasse by making the requirements of the task less restrictive than initially assumed.
two process model
- using a representational change
- the selection of the procedure is reinforced when it is used repeatedly, assuming it is successful
When non-insight problems are encountered, the first process is skipped and if an insight problem follows the representational change process will have to be applied first.
Therefore first solving non-insight problems will not
hinder the subsequent solution of insight problems. However, solving insight problems using one particular representational change process will inhibit the solution of subsequent insight problems if these need to be solved via a different form of representational change
insight problem
a problem that requires representational change, i.e. chunk decomposition or constraint relaxation
three phases of analogical mapping
- the problem to be solved must be interpreted
- a useful source of analogy must be retrieved from long term memory
- some similarity between source analogue and
target problem has to be noted and the elements of the source analogue mapped onto the target problem
after you found a fitting analogy
The novel information provided by the source allows inferences to be drawn and transferred.
Alternatively, the mapping may involve matching relationships rather than conveying new knowledge.
We might even induct an abstract schema to be used in the future
real life use of analogies
not an automatic process, requires controlled attempts
people can use higher order analogies if prompted and in arguments, but stay more superficial for practical problem solving.
Is a generic problem solving technique that is equally used by novices and experts (of a specific field)
structural and superficial similarities
structural similarities are “higher order” similarities as they refer to the underlying relations between the problems.
superficial similarities refer to objects and their properties that are shared
effectiveness of analogical reasoning
analogical reasoning is beneficial because it aids schema abstraction, which is helpful solve problems
two functions of analogies
problem solving; uses higher order analogies
illustration: uses more superficial analogies
three criteria for problem solving
- The activities must be goal-directed, i.e. the individual attempts to attain a particular end state.
- The attainment of the goal or solution must involve a sequence of mental processes rather than just one.
- These processes should be discernibly cognitive.
deductive reasoning task
A problem that has a well-defined structure in a system of formal logic where the conclusion is certain.
inductive reasoning task
A problem that has a well-defined structure in a system of formal logic where the conclusion is highly probable but not necessarily true.
why we make formal reasoning errors
because we process massive amounts of information each day and have to make short cuts. We make less mistakes if the formal problem is close to real life situations
confirmation bias
to form our hypothesis we try to confirm our first idea, instead of disconfirming it, which would often be a more correct strategy.
This bias is not a general tendencies but influenced by factors like working memory constraints and encouragement to test multiple hypotheses
human accuracy to basic inference rules (e.g. modus ponens)
modus ponens: no mistakes
modus tollens: 30% mistakes
affirming the consequent and denying the antecedent: often seen as valid although they are not
why do people make deductive logical mistakes?
because thy draw from their existing knowledge to judge the conclusion. Also the form of the promises (tips or warnings) as well as the authority of the source influence reasoning
how do people solve deductive problems?
they use a mix of logic and heuristic, like confirmation or matching bias
mental logic theories
our reasoning is underpinned by mental rules.
Three parts: a set of inference schemas; selection of appropriate schema; pragmatic principles that influence how the problem statements are interpreted
additional rules can be acquired, existing rules can be altered
pragmatic reasoning schemata
Clusters of rules that are highly generalized and abstracted but defined with respect to different types of relationships and goals. These are pragmatic not logical, although they might appear logical.
Has not received much attention in research
mental model
A representation that we construct according to what is described in the premises of a reasoning problem and this representation will depend on how we interpret these premises.
reasoning using mental models
first comprehend the premises, second reason with the models you made. It can be difficult to comprehend a premise (as in a complicated sentence) and it can be difficult to reason with models. Did you find all possible models? The more possible mental models, the more difficult the reasoning
probabilistic approach
our everyday reasoning is probabilistic and we make errors when presented with logical tasks in the laboratory because we draw on the strategies we use in everyday life
dual process account
- fast and intuitive
- slow and deliberate
different brain regions are activated depending on the process
type 1 processing (dual process account)
old and evolved early
is innate or based on experiential learning
operates largely automatically
it functions independently of general intelligence
independent of working memory
type 2 processing (dual process account)
new and distinctively human
permits rule based reasoning
heavily dependent on working memory
