2.3 Consumer decision-making
- Syllabus
- 9990–2028–2029
- Topic
- 2.3
- Level
- A2
Learning objectives
2.3.1Consumer decision-making• 2.3.1 Consumer decision-making- models of consumer decision-making focusing on explanations and examples of utility theory, satisficing, prospect theory.- strategies of consumer decision-making focusing on explanations and examples of compensatory, non-compensatory and partially compensatory strategies including explanations and examples of each.- decision-making strategies applied to internet shopping focusing on website design, including a study, e.g. Jedetski et al (2002).- Relevant issues and debates and methodology for this topic include: application to everyday life, cultural differences, reductionism versus holism, determinism versus free-will, idiographic versus nomothetic.2.3.2Choice heuristics• 2.3.2 Choice heuristics- heuristics focusing on availability, representativeness, recognition, take-the-best, and anchoring, including explanations and examples of each.- point of purchase decisions including multiple unit pricing and suggestive selling, including a study, e.g. Wansink et al. (1998).- applying heuristics to decision-making styles, including a study, e.g. del Campo et al. (2016).- Relevant issues and debates and methodology for this topic include: application to everyday life, individual and situational explanations, experiments, quantitative and qualitative data, objective and subjective data.2.3.3Mistakes in decision-making• 2.3.3 Mistakes in decision-making- thinking fast and thinking slow (system 1 and system 2) including explanations and examples of each, e.g. Shleifer (2012).- choice blindness, preferences and defending a choice (exemplified by the following key study).- Key study for choice blindness when tasting and smelling food items: Hall et al. (2010).- consumer memory for advertising including how retroactive and proactive interference affect memory, including a study, e.g. Burke and Srull (1988).- Relevant issues and debates and methodology for this topic include: individual and situational explanations, determinism versus free-will, experiments, interviews, reliability.
Decision models set the goal; strategies set the attribute rule
| Model | Decision rule | Example | Boundary |
|---|---|---|---|
| Utility theory | Rationally compare expected benefits/costs and choose the option with greatest utility | Score laptops by expected performance, reliability and price | Assumes stable preferences, usable probabilities and enough time/information |
| Satisficing | Set minimum requirements, search until an acceptable option appears, then stop | Buy the first coat meeting warmth, size and budget thresholds | ‘Good enough’ can be adaptive under limits; it is not automatically careless |
| Prospect theory | Judge outcomes relative to a reference point; losses generally loom larger than equal gains and sensitivity diminishes farther from the reference | A £50 surcharge can deter more than a £50 saving attracts; framing a warranty as avoiding loss changes response | Describes patterned risk/framing effects, not one fixed choice for every person |
| Strategy family | Rule | Named forms |
|---|---|---|
| Compensatory | A strong attribute can offset a weak one | Equal-weight totals; weighted-additive totals give important attributes more weight |
| Non-compensatory | A failure cannot be rescued by other strengths | Satisficing; elimination-by-aspects applies sequential cut-offs; lexicographic choice takes the best option on the most important attribute |
| Partially compensatory | Some trade-off occurs within a restricted comparison | Majority of conforming dimensions compares two options across attributes then keeps a winner; frequency of good/bad features counts attributes above/below cut-offs |
| What the shopper does | Best diagnosis | Why |
|---|---|---|
| Adds weighted battery, price and camera scores | Compensatory | Weak price may be offset by strong camera/battery |
| Rejects every phone above £600 before comparing cameras | Elimination-by-aspects | Price cut-off cannot be compensated |
| Chooses first chair that is sturdy, wooden and comfortable enough | Satisficing | Search ends at aspiration threshold |
| Compares pairs and repeatedly retains the option winning most dimensions | Majority of conforming dimensions | Limited within-pair trade-off |
| Jedetski et al. (2002) | Evidence |
|---|---|
| Question/design | Experiment (Nov 2000–Feb 2001): whether attribute comparison and number of alternatives alter strategies |
| Tasks | Participants chose answering machines, baby monitors, golf clubs and toasters |
| Interface contrast | CompareNet allowed sorting by attributes and side-by-side alternative comparison; comparison-poor design did not |
| Finding | CompareNet produced more compensatory processing and greater satisfaction, but did not improve confidence or task time |
| Application | Comparison tables enable trade-offs; without them shoppers shift toward non-compensatory rules such as satisficing, lexicographic choice or elimination |
| Debate | Evaluation |
|---|---|
| Application | Retailers can match comparison/filter tools to strategy, but a design that helps complex purchases may overload routine choices |
| Reductionism–holism | Categories make cognition testable yet omit emotion, identity, social influence, budget and store atmosphere |
| Determinism–free will | Framing/interface constrain processing, while shoppers can change goals, seek reviews and choose effort |
| Idiographic–nomothetic/culture | General rules predict patterns, but product importance, joint choice, affluence and choice availability shift strategy within and between people |
Do not infer a stable ‘type of shopper’ from one purchase. The same person may use weighted comparison for an expensive computer and satisficing for routine groceries; observed purchase alone may not reveal the private rule.
Choice heuristics trade information for speed using different cues
| Heuristic | Cue/rule | Consumer example | Common confusion |
|---|---|---|---|
| Availability | Judge likelihood/importance by how easily examples come to mind | Vivid battery-fire news makes one phone seem unusually risky | Ease of recall is not objective frequency |
| Representativeness | Judge by similarity to a prototype/category | Leader-like packaging is assumed high quality | Resemblance can ignore base rates and real attributes |
| Recognition | If one option is recognised and another is not, favour the recognised one | Buy the familiar shoe brand worn for years | Familiarity is the cue; it is not mere recent recall |
| Take-the-best | Inspect cues by validity and stop at the first discriminating cue | Choose the egg carton with the clearest decisive welfare/price cue | It does not total every attribute |
| Anchoring/adjustment | Start from an initial number/information and adjust insufficiently | ‘Limit 12’ or ‘buy 18’ raises quantity considered normal | Anchor need not be informative or a genuine discount |
| Wansink, Kent & Hoch (1998) | Exact teaching claim |
|---|---|
| Model/design | Anchoring-and-adjustment account tested in two field experiments and two laboratory studies |
| Point-of-purchase anchors | Multiple-unit price (‘5 for £2’), purchase limit (‘limit 5’) and suggestive quantity (‘buy 18 Snickers for your freezer’) |
| Finding | Numerical/quantity anchors increased purchase quantity; even general and explicit Snickers suggestions raised sales |
| Boundary | Consumers who retrieved self-generated anchors such as budget, list or normal usage resisted more effectively; more units can reflect stockpiling/switching rather than greater consumption or retailer profit |
| Technique | Cognition encouraged | Risk/control |
|---|---|---|
| Multiple-unit pricing | Makes a larger bundle feel attractive/normal | Show unit price; compare against a one-unit control |
| Quantity limit | Implies scarcity/high demand and supplies a numerical anchor | Can feel manipulative; observe whether customers actually need it |
| Suggestive selling | Adds a use or complementary item and a quantity idea | Standardise wording and distinguish cross-sell from pressure |
| Consumer defence | Recall usual quantity, budget/list and compute unit value before using the display anchor | Test whether internal anchor removes the effect |
| del Campo et al. (2016) | Evidence |
|---|---|
| Aim/design | Between-subjects online experiment in Vienna and Madrid; time pressure present/absent |
| Task | Choose one of five egg cartons designed for recognition, take-the-best, emotional, cognitively complex or filler processing; order randomised |
| Pressure | Random assignment; 40-second limit versus no limit; then open reason, buying questions and 25-item decision-style scale |
| Findings | Pressure reduced cognitively complex choice in both samples but increased take-the-best clearly only in Vienna; Madrid shifted toward recognition; style relationships differed by country |
| Conclusion/boundary | Situation and individual style can interact, but results were inconsistent across culture; chosen product only indirectly identifies the heuristic and eggs are one low-involvement product |
| Evidence lens | Strength | Limit |
|---|---|---|
| Experiment/objective choice | Standardised products, random assignment and recorded selection permit causal comparison | Artificial cartons, inferred cognition and time awareness reduce validity |
| Quantitative scales | Compare style dimensions and conditions reliably | Scores are subjective and translation/factor structure may vary culturally |
| Open reasons | Reveal why a cue mattered | Post-hoc rationalisation, social desirability and coding subjectivity |
| Determinism | Heuristics predict systematic responses useful for interventions | More than one shortcut may operate; lists, knowledge and deliberation preserve agency |
Identify the heuristic from the operative cue, not speed alone. Numerical promotions can anchor quantity without creating a real saving, and del Campo did not find one universal time-pressure or personality effect across Vienna and Madrid.
Fast processing can hide a switched choice and similar adverts can overwrite memory
| System | Processing | Consumer use | Error/boundary |
|---|---|---|---|
| System 1 / thinking fast | Automatic, rapid, low-effort, associative and often unconscious | Routine groceries, familiar brands, immediate affect and heuristics | Efficient but cue/framing sensitive; not always wrong |
| System 2 / thinking slow | Effortful, deliberate, sequential and consciously monitored | Expensive computer comparison, checking unit price and reviewing evidence | Can correct an intuition but needs attention/time and can rationalise rather than guarantee truth |
Shleifer (2012) applies the dual-process distinction to judgement: System 1 supplies quick intuitions; System 2 may monitor or revise them. Behavioural speed alone cannot reveal the private process.
Choice blindness occurs when a person fails to notice that the outcome presented as their choice differs from the one selected, yet may confidently defend the manipulated outcome. Preference reports can therefore be partly constructed from the current situation rather than retrieved as fixed facts.
| Hall et al. (2010) | Exact evidence |
|---|---|
| Sample/setting | 180 supermarket customers in Lund, Sweden (118 women, 62 men; age 16–80; three recordings later excluded) |
| Cover/deception | Presented as independent consultants running quality control; double-ended divided jars enabled covert switch while an experimenter distracted the participant |
| Task | Taste paired jams and smell paired teas, choose/rate on 1–10; on manipulated trials the chosen item was switched before re-sampling and reasons were elicited |
| Motivation manipulation | Half expected their selected tea/jam as a thank-you gift |
| Results | Only 33% of manipulated jam and 32% of tea trials were detected; dissimilar jam switches were detected more, but similarity did not show the same pattern for tea; promised jam reduced detection, promised tea did not |
| Conclusion | Considerable choice blindness survived meaningful sensory differences and participants could justify switched outcomes |
| Interference | Timeline | Advertising consequence |
|---|---|---|
| Proactive | OLD memory → disrupts NEW learning/recall | Earlier competitor-phone advert makes new phone details hard to learn or confuses them |
| Retroactive | NEW material → disrupts OLD memory | Later competitor advert alters/blocks memory for the original brand |
Similarity and close timing increase competition. Use the arrow test: proactive acts forward from old; retroactive reaches back from new.
| Burke & Srull (1988) | Result |
|---|---|
| Experiment 1 | After 12 magazine adverts and purchase-likelihood judgments, later same-line/competitor advertising inhibited memory for earlier brand information: retroactive interference |
| Experiment 2 | Prior competitive advertising impaired learning/recall of distinctive information in later adverts: proactive interference; a predicted interaction was absent, possibly due to short exposure–test interval |
| Experiment 3 | 22 full-colour magazine adverts and value ratings; competitive advertising significantly inhibited cued recall, and repetition helped recall mainly when similar-product advertising was low/absent |
| Boundary | Controlled laboratory evidence supports causality/reliability, but advert exposure, motivation and recall tests are less natural than real repeated campaigns and purchase |
| Lens | Hall | Burke–Srull |
|---|---|---|
| Experiment/validity | Field setting, real sensory materials and low demand; less control over supermarket context | Lab control, standardisation and replication; demand/artificial exposure |
| Data/interview | Detection rates plus reasons reveal blindness and post-choice defence | Recall measures quantify memory inhibition but reveal less lived meaning |
| Ethics | Covert switch/false cover mean deception and incomplete consent, though debrief/low stakes can limit harm | Participants know study context more, but aim may still be concealed |
| Determinism–agency | Unnoticed manipulation constrains report; detection shows not everyone is blind | Order/similarity constrain recall; attention, motivation and strategy still matter |
| Reliability/generalisation | Standardised jars/procedure, but Swedish supermarket and food senses | Repeatable tasks, but magazine ads and student/experimental context may not generalise |
System 1 is adaptive as well as error-prone, and System 2 is not an accuracy guarantee. Remember the interference arrows: old blocks new = proactive; new blocks old = retroactive. Hall measured failure to detect a covert switch, not absence of all genuine preference.