Human Cognitive -Read Entirely FIRST!!

What Is a Problem?

· Components of a Problem:

· Initial State

◘ The starting point

· Goal State

◘ The desired end

· Rules

◘ Constraints that must be met

Well-Defined and Ill-Defined Problems

· Well-Defined Problems

· Clear and structured

· Initial states, goal states, constraints understood

· Solution is clearly right/wrong

· Ill-Defined Problems

· Unclear and vague

· Initial states, goal states, constraints imprecisely specified

· Solution accuracy not immediately assessable

❖

Routine and Non-Routine Problems

· Routine Problems

· Consists of procedures that have been performed many times

· Non-Routine Problems

· Consists of procedures that are new or unfamiliar

· Routine problems tend to be well-defined due to the previous experience

Problem Solving Research: Some Methodological

Challenges

· Problem solving can be difficult to study, given its complexity and span of duration

· Speed and accuracy are common DVs in cognitive psychology

· Often uninformative in problem solving research

» Verbal protocols

· Verbal reports given as problem solvers “think out loud” during the problem solving interval ● Frequently used

· Limitations:

− Verbal ability necessary − Accuracy of report?

− Introspection may alter processing

◘ Most research indicates minimal interference

❖

❖

» The Varied Nature of Problems

· Transformation Problems

– Moves that transfer one from initial state to a goal state

· Arrangement Problems

– Figuring out how to arrange problem elements

· Induction Problems

– Given specifics, figure out general rule

· Deduction Problems

– Given general principles, determine specific conclusion

· Divergent Problems

– Generate as many solutions as possible

Approaches to the Study of Problem Solving

❖Behaviorism: Problem Solving as Associative Learning

· Early study of “problem solving”

· E.L. Thorndike – Cats as solvers

− Problem: Trapped in “puzzle box” − Goal: Freedom and food

Problem solution was not immediate

· Learning occurred gradually, through trial and error

· Thorndike proposed the law of effect to account for the trial-and-error learning pattern

◘ Response that lead to satisfying outcomes— strengthened

◘ Nonsatisfying outcome—response weakened

· Gestalt Psychology: Problem Solving as Insight

· Mind has an inherent tendency to organize incoming information

· Problem solving involves a restructuring of problem elements

· Result is a sudden realization of the solution

– Called “insight”

» Contrasting the Behaviorist and Gestalt Views

· Behaviorist View

– Strengths

◘ Simplicity and precision

− Weakness

◘ Fails to explain novel and creative behaviors

· Gestalt View

· Strength

◘ Accounts for novel and creative behaviors in terms of mental representations

· Weakness

◘ Imprecise and ill-specified

· Cognitive Psychology: Problem Solving as Information

Processing

» General Problem Solver

· Proposed by Newell and Simon

· A computer model of human problem solving

· Subgoal Analysis

− Minimize “distance” between initial and goal states by breaking problem down into subgoals

· Problem Space

– Solver’s representation of the initial and goal states, all intermediate states, and operators

◘ Problem solving as “excursion through problem space”

Problem Representation

· Problem solving involves specifying the problem space

· Successful solution depends largely on appropriate representation

~ Example

Visualization makes it apparent that there is a single spot that will be passed at precisely the same moment

❖Rigidity in Problem Representation

~ Example: What’s the rule that generates the sequence?

8, 5, 4, 1, 7, 6, 10, 0

» Mental Set

· Tendency to rely on habits and procedures that have worked in the past

» Functional Fixedness

· A type of mental set

· Tendency to view items in terms of their most typical function(s)

· Not always a bad thing

− Most life situations require conventional thinking

Duncker (1945)

· The candle problem

· Task: attach the candle to the wall so that it burns properly using the materials on the table

· Conditions

· Functional fixedness condition—critical items (candles, tacks and matches) in boxes

· Control condition 1—boxes were empty and critical items were on the table

· Control condition 2—boxes filled with non-critical items and critical items

· Results

· In control condition 1, all participants solved the problem successfully

· In the functional fixedness condition and control condition 2, only about 1/3 solved the problem

♦ Boxes were viewed only as containers for what they were holding

German and Barrett (2005)

· Functional fixedness in “technologically sparse” culture

· Objects are not as specialized for a given function, so may not show functional fixedness

· Participants were members of Shuar tribe in Ecuador

· Task

· Construct a “bridge” as part of a story in which character needs to cross a river ✓Materials presented:

♦ Spoon, lollipop stick, plastic cup, eraser, clear ball, cup of rice

▪Spoon was the target object as it was the only object long enough to be “span the river”

· Two presentation conditions:

· Objects presented separately

· Spoon presented sticking in the cup of rice

· Dependent variable: RT to pick the spoon and time to solve

· Results

· Functional fixedness evident, even in non-technological culture

❖Individual Differences in Problem Representation

Stereotype threat

– Negatively stereotyped group member feels the stereotype will be used to judge their behavior

◘ Pressure/anxiety surrounding the propagation of the stereotype undermines performance’

Quinn and Spencer (2001)

· Presented females and males with GRE math problems word problems or algebraic equivalents

· Mathematical knowledge needed to answer both was the same

· Word problems require transformation into proper mathematical representation

Results

· Woman and men possessed the mathematical knowledge to solve the problems

✓Equal performance in algebraic condition

· Stereotype threat presumably affected women during the problem representation stage

· Men outperformed women in the word problem condition

2nd experiment tested if poor performance was due to stereotype threat

· Added a low stereotype threat condition

♦ Told there were no sex-related differences found on the word problems

· Results

· When stereotype threat was eliminated, no sex differences were found

Sex difference due to difficulty in problem representation?

· Used verbal protocol technique when solving word problems

· Dependent variable

♦ Failure rate: inability to determine a strategy (i.e., proper problem representation)

· Results

♦ Failure rates were equivalent in low stereotype threat condition

· Difficulty occurred during problem representation

· Locus in executive control: activation of stereotype is resource demanding

➢

➢

➢

» Stereotype Threat Meets Mental Set

Mere effort account

− Stereotype threat can be negative or positive depending on problem-solving strategy

◘ Stereotype threat activates prepotent response

Prepotent response = mental set

· If prepotent response incorrect or inappropriate—performance suffers

· If prepotent response correct or appropriate—performance enhanced

Jamieson and Harkins (2009)

· Compared GRE math problems

✓Prepotent response to math problems is to use a formula

♦ “Solve” problems

· Prepotent response is appropriate

· Performance should be enhanced under stereotype threat

♦ “Compare” problems

· Prepotent response is inappropriate

· Performance should be hurt under stereotype threat

· Results

· Solve problems (prepotent response appropriate)

♦ Stereotype threat enhances performance

· Compare problems (prepotent response inappropriate)

♦ Stereotype threat deters performance

Problem Solution

Finding a solution = “traveling through” problem space

− Algorithms

− Heuristics

· Algorithms

· Rules that can be applied systematically to solve a problem

· Solution is guaranteed if algorithm correctly applied

− Limitations

◘ Often not feasible given limits in human information processing

◘ Algorithms don’t exist for many problems

· Heuristics

· Shortcuts that improve efficiency, but don’t guarantee success

· Types

− Means-end analysis

− Analogies

» Means-End Analysis

· Breaking a problem into smaller subgoals

· Each subgoal moves the solver closer to solution

» Analogies

· Problems that have already been solved as aids for representing and solving a current problem— In general, people do not readily pick up on analogies

Gick and Holyoak (1980)

· Developed the “radiation problem”

Solution focus many lower-level rays from numerous different directions

♦ Convergence solution

· Materials

· Story which involved solving a problem analogous to the radiation problem (commander story)

♦ Referred to as the source problem

· Radiation problem

♦ Referred to as the target problem

· Conditions

· Given source problem to memorize, followed by being asked to solve the target problem

· Target problem only

· Dependent variable: % solving radiation (target) problem

· Results

· Source + target: 10% ✓Target only: 30%

· So 20% spontaneously used analogy

Gick and Holyoak (1983)

· When does analogical transfer occur?

· Tested several “source problem + hint” conditions

· Source problem + diagram

· Source problem + general principle

· Source problem + another analogous problem

♦ Find relationship between the two problems

· Results

Successful use of analogy requires schema induction

· Schema = mental representation of the underlying representation shared by two problems

· Steps needed for schema induction

◘ Noticing: must notice that a relationship exists between the two problems

◘ Mapping: must be able to map the key elements of the two problems

◘ Development : must develop the general schema that can be used to solve the target (i.e., current) problem

◘ Most common failure is in the initial (noticing) stage

Memory problem: current problem fails to trigger the memory of the earlier problem

Surface vs. Structural Features

· Related to the spontaneous recognition and retrieval of an analogous problem

− Surface features = specific elements of the problem

− Structural features = underlying relationships among surface features of problems

· Analogies likely to prove helpful when surface features match

· Differences in surface features hinder the effectiveness of analogies

− People tend not to notice structural similarity

Lane and Schooler (2004)

· Investigated the effects of verbalization on effectiveness of analogies

· Conditions

· Participants read 16 initial problem-solving scenarios (initial scenarios)

· Followed by 8 test problem-solving scenarios (test scenarios)

♦ ½ similar in surface features

♦ ½ similar in structural features

· Participants verbalized or were silent

· Dependent Variable: for each test scenario indicate the initial scenario to which it was most similar

· Results

Data represents the mean number of test scenarios in which the correct analogous initial scenario was selected (range 0-8)

· Verbalization:

◘ enhanced the ability to notice surface similarity

◘ impaired the ability to notice structural similarity

· Talking leads one to focus on surface similarities as they are easier to talk about

In real-world situations, people seem better at picking up on analogies than data would indicate

− Potential artifact of the laboratory

Blanchette and Dunbar (2000)

· Participants were asked to generate their own analogies to target problems

· Analogies shared structural similarity not surface similarity

· Suggests people may be more sensitive to structural features than traditional lab-based studies indicate

Markman, Gentner, and Taylor (2007)

· Previous studies generally present problems in written form

· Few everyday problems are presented in written form, so examined another modality—listening

✓Anaphoric reference (Chapter 10) is more likely to be apprehended with spoken presentation

· Focused on memory as that is problem in the use of analogies

· Procedure

✓Presented proverbs (e.g., the swiftest steed can stumble)

· Conditions

· Later cued recall with proverb cues that shared:

♦ Surface features (e.g., a rough steed needs a rough bridle)

♦ Structural features (e.g., the greatest master is wrong from time to time)

· Participants either listened to, or read, the initial and cue proverbs

· Dependent variable: memory accuracy

Predictions

· In listening condition

♦ Structural Surface>

· In reading condition

♦ Structural Surface=

· Results: as predicted

· Implication: studying problems in written form may underestimate the use of analogies

Catrombone, Craig, and Nersessian (2006)

· Encoded the Commander problem

· Three Retrieval phases

· Phase 1: recall it in one of three modes:

♦ Verbal—recount the story

♦ Visual—recount the story while sketching it

♦ Enactment—recount the story and use blocks to describe what happened

· Phase 2: given the radiation problem and asked to come up with as many solutions to it in 8 minutes

· Phase 3: start fresh and attempt to find a solution based on the initial story (commander problem)

Results

Phase 1

· Enactment during phase 1 was associated with higher rates of solution in phase 2

· When told to use initial story (phase 3), memory did not differ

♦ As found in other studies, when told to do so, people are able to use analogies

» Problem Solution: Dual Processes Revisited

Both algorithms and heuristics can involve system 1 and system 2

Pretz (2008)

· Compared system 1 (intuition) and system 2 (analytic) processes in the context of problem solving

· Hypothesis:

✓Most effective problem-solving mode depends on the experience level of the problem solver

♦ Little knowledge, experience: system 1 better

♦ Experience, knowledge available: system 2 better

Participants: 1st year and 3rd year college students

· Dependent variable: ratings of problems on College Student Tacit Knowledge Inventory

➢

➢

Conditions: varied instructions

· System 2 (analytic)

♦ Engage in 4-step process for analyzing the problem

· Define the problem

· Identify relevant information

· Decide how to use resources to solve the problem

· Evaluate possible solutions and consequences

· System 2 (intuition)

♦ Vividly imagine the situation, think about it holistically, trust your gut, incubate (take a break and come back)

· Control

♦ Solve the problem any way in which you feel comfortable

Results

Problem solving distance = ratings distance from consensus solution rating

· First-year students performed better under system 2 instructions

♦ Not have enough experience to use system 1 instructions

· Third-year students performed better under system 1 instructions

♦ Had enough experience to use instructions effectively

Experts: Masters of

Problem Representation and Solution

Expertise = Exceptional knowledge and/or performance in some problem domain

– Early view: innate capacity or talent

− Recent view: an information-processing account

◘ Expertise is an extremely well-learned set of cognitive abilities and skills

◘ 10 years of extensive practice

➢

➢

❖Expert Advantages

· Core of expertise is memory

· Experts = skilled memorizers

· Skilled memory theory

· Experts advantages (relative to novices)

◘ More extensive semantic knowledge networks

◘ Quick and efficient coding in LTM

◘ Quicker and more direct access to LTM

deGroot (1948/1978)

· Compared chess players with various levels of experience

· Briefly presented meaningful chess board configurations

· Task: reconstruct configurations

· Dependent variable: accuracy

· Results

· Chess experts could recall board configurations near perfectly

Chase and Simon (1973)

· Is the memory ability of experts a general one or specific to their area of expertise?

· Participants: expert and novice chess players

· Conditions: varied arrangement of chess pieces

· Random configurations

· Game configurations

· Results

· Superior memory in experts, but only for game configurations

· Equivalent recall of random configurations

Explanation for superior memory performance

− Superior pattern recognition and chunking ability in experts

− Not simply due to immediate memory (IM) ability

◘ Not subject to limitations found in IM

Charness (1976): expert memory for chess pieces doesn’t diminish with delay (even with interference) Gobe and Simon (1996): more “chunks” than possible in IM

Long-term working memory (Ericsson and Kintsch, 1995)

· Experts bypass limits of immediate memory by using information in immediate memory to access LTM directly

− Unitary view of memory: domain specific activation of LTM that is currently in consciousness

Strategy differences between experts and novices

− Experts tend to work through problem space in a forward fashion, from initial state to the goal

◘ Novices start with the goal and work backwards

· Experts are better at picking up on structural features of problems

◘ Novices are more likely to focus on surface features

− When faced with a problem in area of expertise, more likely than novices to notice analogous problem/situation







Adaptive Strategy Model

– Expert-novice differences exist at 4 levels

◘ Strategy existence: experts have more available strategies than novices

◘ Strategy base rate: know which strategies work, in general, and are biased toward using them

◘ Strategy choice: expert advantage at discerning which strategy would be best for a specific problem

◘ Strategy execution: expert advantage (speed and accuracy) in carrying out the chosen strategy

» Expertise Advantages vs. Age-Related Deficits

Nunes and Kramer (2009)

· Participants

· Older and younger air-traffic controllers (ATC)

· Older and younger non-ATC controls

· Tasks

· Non-ATC tasks: IM capacity, inductive reasoning, processing speed

· ATC tasks (benefit from experience as ACT): task switching and inhibitory control

· Computer simulation tasks that replicated specific ACT functions

♦ Conflict detection: judging if two aircraft would collide

♦ Conflict resolution: resolving a conflict with an appropriate response

♦ Vectoring: sequencing aircraft within corridors around an airport

♦ Airspace management: managing flow of air traffic within airspace growing increasingly crowded

· Results

· Non-ATC tasks showed age-related deficits

· ACT tasks showed experience-related sparing

· Computer simulations

♦ Conflict detection and resolution

· Experience-related sparing was not found

♦ Vectoring and airspace management

· Showed experience-related sparing

» Expert Disadvantages: Costs of Expertise?

Intermediate effect

· Experts remember less detailed information than those at intermediate levels of expertise

· Encapsulation hypothesis

◘ Experts “chunk” information into higher level summarizing concepts using system I processing

◘ Consequently remember information at that level and don’t remember the details

Memory distortion

Castel, McCabe, Roediger, and Heitman (2007)

· Participants: high (expert) and low (novice) knowledge of

NFL football team names

· Used DRM paradigm (Chapter 8)

· Lists

· Animal names that were also NFL team names ✓Body parts

· Predictions

· Recall

♦ Animals: Experts novices>

♦ Body parts: Experts novices=

· False recall

♦ Animals: Experts novices>

♦ Body parts: Experts novices=

· Results

· As predicted

Expert Mental Set?

Bilalic, MacLeod, and Gobet (2008)

· Participants: skillful and super expert chess players

· Task: game problem that could be solved in two ways

· Non-optimal, but very familiar solution

· Optimal, but unusual and unfamiliar solution

· Results

· Skilled experts chose familiar, but non-optimal move

· Super experts picked the unusual and optimal move

· Existence of mental set depends on level of expertise

Insight and Creativity

Stages of creative problem solving (Walls, 1926)

· Preparation

· Incubation

· Illumination

· Verification

❖Insight

· Sudden realization of a problem’s solution

· Insight problems

· Solved with a (seemingly) sudden realization of a problem’s solution

· Non-insight problems

· Solved through conscious, step-by-step procedures

» Removal of a Mistaken Assumption?



Nine dot problem

· Use continuously drawn line to connect the dots

· Mistaken assumption: stay within the boundaries created by the nine dots

· Single hint often fails to lead to solution

· Sources of difficulty in solving insight problems

· Perceptual Factors

◘ The ways the problem is seen initially

· Process Factors

◘ Size of problem space, complexity

· Knowledge Factors

◘ Application of previous experience (mental set?)

· All of the above factors (not one) lead to the difficulty in solving insight problems

» The “Aha!” Experience

Metcalfe and Wiebe (1987)

· Procedure

· Presented insight and non-insight problems

· Given four minutes to solve each problem

· Dependent variable—every 15 seconds provide a:

· Rating of warmth

♦ How close the person believes they are to solving the problem

· Judgment of the likelihood of solving the problem

· Prediction

· Non-insight problems

♦ Ratings of warmth and likelihood of solving judgments should increase over time

· Insight problems

♦ Ratings of warmth and likelihood of solving judgments should be low until solution is suddenly realized

· Results

· As predicted

· Fundamental difference between insight and non-insight problems

− Metacognition for non-insight problems are accurate and predictive of actual performance

− Metacognition of insight problems is unrelated (or negatively related) to probability of solving the problem

· Pattern of ratings of warmth during problem solving may be used to classify a problem as insight or non-insight

· Some suggest the Aha! experience is phenomenological

− All problems are solved incrementally we are just unaware of it, so it appears sudden

» Intuition as insight

· Two-stage process

− Stage 1 (guiding stage)

◘ Mnemonic networks relevant to the problem are activated and begin to spread

◘ Working on the problem unconsciously

− Stage 2 (integrative stage)

◘ Buildup of activation reaches enough strength to break through into consciousness

Transition from stage 1 to stage 2 is insight

Bowden and Beeman (1998)

· Right hemisphere is more involved in creativity than the left hemisphere

· Authors equate insight problems with creativity

· Task: RAT (Remote Associate Test) insight problem ✓Shown three “unrelated” words and generate one word that ties the triplet together

Example: wine, reading, sun

~ Remote associate: glasses

Procedure

· Presented RAT triad for 15 seconds to the right or left of fixation (left or right hemisphere, respectively)

· After solution word was generated or time ran out, a target word was presented that was to be pronounced

♦ The solution word

♦ Non-solution word

· Dependent variable: priming

· Non-solution word pronunciation time—solution word pronunciation time

Results

· Solved problems: more priming when RAT triad presented to right hemisphere

♦ Processing by the right hemisphere speeds ability to pronounce the solution word

· Unsolved problems: priming only found when RAT triad presented to right hemisphere

♦ Right hemisphere is “working” toward solution, so pronunciation time is speeded

♦ Left hemisphere in not “working” toward solution, so no effect on pronunciation time

Follow-up study

· Same procedure

· Changed response: indicate “yes” or “no” as to whether the target word was the solution word for the triad

· Results

♦ Right hemisphere is working towards solution, so faster to identify target word as solution when unsolved

➢

➢

➢

· Neural signatures when solving insight problems

− fMRI scans reveal increased activation in the anterior superior temporal gyrus of the right hemisphere

− EEG reveal a sudden burst of neural activity immediately preceding solution

» Does incubation lead to insight?

· Incubation: taking a break in problem solving leads to a quicker solution than does continuing effort

Smith and Blackenship (1991)

· Procedure

· Solve rebus problems: words and pictures are used to depict a common phrase

♦ Example: HEAD

HEELS

· For some problems misleading cues were presented

♦ Example: ACHE and HIGH

♦ Misleading as it suggests “head ache” and “high heels”

· Retested unsolved problems

♦ Immediately

♦ Varying periods of delay (incubation)

♦ Asked to recall the initial cue words presented with each problem

· Results

✓Longer delays were associated with:

♦ Higher probabilities of solution

♦ Poorer memory for the misleading cues

▪As cue was forgotten, problem became more solvable

· Explanation

✓Encoding specificity principle in reverse

♦ Contextual change prevents reinstatement of the context that stymied solution

❖Creativity

» What is creativity?

· Creative solutions have 2 components

− Novelty

− Appropriateness

· 4-c model of creativity: based on impact of creative product

− Big-C creativity (eminent creativity): creative products have major impact

− little-c creativity (everyday creativity): creative products have minor impact

− mini-c creativity: novel and personal interpretation of experiences

− Pro-c creativity: professional creativity that does not reach Big-C creativity level

· Creativity can be informed by focus on dimensions:

− Person

− Process

− Press

− Product

Person

− Creativity, to some extent, depends on personality characteristics

◘ Broad interests

◘ Appreciation of complexity

◘ Tolerance of ambiguity

◘ Self-confidence

◘ Independence

◘ Sensible risk taking

◘ High degree of intrinsic motivation for their field ◘ Flexibility

◘ React effectively to change

− Developmental aspects

◘ Constantly developing ability, rather than static attribute present from birth

◘ More likely to develop with a diverse set of life experiences

· Enhances ability to take fresh perspectives

◘ Depends on having faced sufficiently challenging life experiences

· Develops ability to persevere

Creative problem solving requires overcoming obstacles, which requires perseverance

− All personality data is correlation ◘ Does personality leads to creativity?

OR

◘ Does creative ability leads to development of the personality characteristics?

· Process

· Creativity can be characterized as a set of processes

− Special processes

− Normal cognitive processes

− Creative cognition approach

◘ Creative thinking can be the result of either type of processing or both

· Other important cognitive processes in creativity

− Wide and diffuse attentional processing

− Good memory in terms of knowledge (semantic memory)

− Selective encoding: ability to distinguish between relevant and irrelevant information

− Selective combination: ability to relate new information to old information in novel ways

· Press

· Creativity is subject to a variety of contextual factors and external pressures

· Product of interpersonal, disciplinary, and sociocultural environments

− Being evaluated by others may decrease creativity

◘ Brainstorming is ineffective

− Disciplinary area defines what is deemed creative

− Cultural diversity increases creativity

· Product

The outcome of the creative process

− Based on case studies of highly creative individuals

◘ Limited by the methodology—difficult to generalize

− One standard is productivity

◘ May be associated to creativity, but more does not always mean better

» A taxonomy of creative processes and products