Problem Solving Techniques PDF

Week 10: problem solving - Features -\> initial state, goal state, obstacles, Problem solving is; goal directed, deliberate, only relevant if solver lacks relevant knowledge for an immediate answer. - Well defined problems -\> no mystery in terms of when the goal is met and how to...

Week 10: problem solving - Features -\> initial state, goal state, obstacles, Problem solving is; goal directed, deliberate, only relevant if solver lacks relevant knowledge for an immediate answer. - Well defined problems -\> no mystery in terms of when the goal is met and how to meet that goal (actual execution of those steps may not be easy) - Ill-defined problems -\> underspecified, many variations of actions/strategies could be attempted to resolve the problem -- how do I know that they will get me towards the goal state? - Knowledge rich -\> to solve problem, solver requires relevant specific knowledge - Knowledge-lean -\> info contained in initial problem statement Strategies - Algorithms -\> Guidelines for task completion with step-by-step operations across the entire problem space are described. ensure a solution, ideal for well-defined problems. However, algorithms can be effort intensive, slow, and prefer heuristics due to careful planning required. - Heuristic -\> rules of thumb/short cutes for our thinking (based on previously successful solutions, produce answers that sometimes incorrect, simple and fast) Means-end analysis -\> heuristic based on creating subgoals to reduce difference between current state and goal state - Hill climbing -\> step-by step approach, less sophisticated than means end, useful when problem not clear - Progress monitoring -\> monitor rate of progress towards goal, if too slow we employ dif strategy - Availability heuristic -\> first solution that comes to mind, probability based on ease (relevant instances and operations (retrieval, construction, association)) can be performed. - Analogies -\> detecting similarities between current problem and problems solved in past. No guarantee that analogous will solve problem, needs to b noticed, ppl need to know it's relevant. 3 similarities between problem (1) superficial similarity (commonalities in solution irrelevant details), (2) structural similarity (casual relations between some main components) (3) procedural similarity (common actions for turning principle into concrete operation). Superficial similarities tend to be easy to perceive but misleading. If participants are asked to generate analogies rather than perceive = more sensitive to structural and procedural similarities. Isn't always 2 dif things or settings that are mapped in analogous problem solving. We can map from a simple version of the same problem and apply to the complex version - Insight (solution pops into mind, require construct different versions of the problem space, bc 1^st^ representation is wrong -- we have to restructure) vs non-insight (deliberate WM processing), hints (Thomas and lleras) and sleep (wallas) facilitate insight, sleep bc solver may have initially been using wrong strat, taking break allows us to come back fresh and forget wrong strat. - Representational change theory -\> insight =conceptualisation/conceptualization, blocks occur when this is done wrong, solved by; (1) constraint relaxation, (2) re-encoding, (3) elaboration Expertise - Exposure to variety of dif problem types and solutions mean experts already have solution in memory, allows ppl to plan further ahead and get to important info faster - **Chess (plan ahead) -\>** info stored in templates, experts and non experts don't vary in general memory but increased memory for specific atea bc templates have larger spans, perform well under time pressure, remember locations of pieces if locations map onto templates - Medical (faster) -\> use implicit reasoning (faster, automatic), more efficient eye movements shift from attentional search to perceptual processes reflect explicit -\> implicit - Deliberate practice -\> elements; appropriate level of difficulty, feedback on performance, chances to repeat task, chance to correct errors. If used enough there is a shift from WM to LTM. Factors like IQ task type and genetics factor into efficiency of practice. Issues w problem solving - Functional fixedness -\> fixation of an object's usual function at the expense of realizing it could be used in another way to solve the problem. Can be reduced if objects is less associated with its typical function or if the key objects needed are highlighted. - Mental set -\> applying a strategy that worked in the past even though it is no longer a good strat - Counteracting these issues -\> If we apply the generic-parts-technique which involves: Producing function free descriptions or Look at whether the descriptions afford new uses of the part, being aware that we are susceptible to these things can help counteract them Hypothesis testing - Test by verification (test, amend and retest if incorrect, popper argued this isn't enough and susceptible to confirmation bias) or falsification (showing it to be false, what should happen).