이중 프로세스 이론과 멘탈 에너지

이중 프로세스 이론과 멘탈 에너지; Feeling and Thinking in Memory-Based versus Stimulus-Based Choices

 

#Dual process theory, 이중 프로세스 이론, Kahneman, Frederick, automatic, controlled, 시스템1, 세스템2, 기억, 지식, 에너지 고갈, 소비자행동, 마케팅, 마케팅 논문, 경영학 논문

 

제 전공인 Consumer Behavior에서 박사자격 시험때 공부했던 논문 소개 들어갑니다= Memory, Knowledge, Beliefs (기억, 지식, 신념, 믿음) 범주에 속한 논문입니다.

 

Rottenstreich, Yuval, Sanjay Sood, and Lyle Brenner (2007), “Feeling and Thinking in Memory-Based versus Stimulus-Based Choices,” Journal of Consumer Research, 33(4), 461-469.

 

이 논문을 이해하기 위해서는 먼저 Dual process theory(이중 프로세스 이론)를 알아야 합니다. Kahneman 과 Frederick는 우리가 어떤 결정을 내릴 때 두 가지 체계가 있다고 얘기합니다.

1. 하나는 자동적이고 반사적이고 감성적인 사고로 이를 System 1이라고 합니다 (automatic, reflex-like, affective processing).

2. 반면에 잘 통제된 이성적인 사고를 우리는 System 2 사고라고 합니다 (controlled, deliberate processing).

 

우리는 대게의 경우 System 1으로 먼저 사고를 합니다. 그리고 System 2로 이성적인 생각을 하죠. 예를 들어 맛있는 디저트가 있으면 먹고 싶다! 라는 자동적인 생각이 System 1입니다. 반면에 "에이 살찔 거야"라고 절제하는 것이 System 2적인 사고이지요.

 

이 논문에서는. System 2적인 사고를 하려면 우리의 멘탈 에너지(plentiful mental resources)가 있어야 한다는 것을 밝힌 논문입니다.

 

논문에서 디저트를 고르게 하는 실험이 있었습니다. 한 집단의 사람들에게는 4개의 디저트 이름을 외우고 있으라 한 뒤 실험이 마칠 때쯤 그중 하나를 고르게 했습니다. 반면 다른 집단의 사람들에게는 실험 마지막에 디저트 4개를 한 번에 보여주고 고르라고 했지요. 그러자 외우라고 한 집단은 "외우는데 에너지를 이미 소비하였기에" 충동적으로 건강하지 못한 디저트를 골랐습니다. 반면 마지막에 한 번에 보여준 그룹은 에너지가 남아있기에 이성적으로 사고하여 건강한 디저트를 상대적으로 더 많이 골랐어요. 이 뿐만 아니라 비싼 물품 구매하는 상황과 과제를 해야 하는 상황 등 다양한 경우를 테스트했는데 한번 아래의 내용들을 확인해보세요.

 

1. memory-based → effortful → system1 (automatic)
2. stimulus-based → pentiful mental resources → system2 (controlled)

 

 

초록

 

• Memory-based vs stimulus-based choices
  • We contrast memory-based and stimulus-based choices, using dual-process theories such as Kahneman and Frederick’s system 1/system 2 dichotomy.
    system1, 2 분류를 이용해서
    • Systems 1 and 2 are conceptualized as distinct modes of thought,
      1. the former automatic and affective,
      2. the latter controlled and deliberate.
  • Cognitive load impedes system 2, yielding greater reliance on system 1.
• Memory-based choice ∝ congnitive load
  • In memory-based choice,
    • consumers must maintain relevant options in working memory.
    • Thus, memory-based choices are associated with greater cognitive load than stimulus-based choices.
  • Indeed,
    1. we find that memory-based choices
       • favor immediately compelling, affect-rich system 1 options,
    2. whereas stimulus-based choices
       • favor affect-poor options whose attractiveness emerges from deliberative system 2 thought.

 

 

내용 정리

 

• Literature Review
  • Stimulus based vs Memory based — contrasted decisions
    • stimulus based
      눈 앞에 물건 사
      • in which relevant choice options were specified or listed for consumers
        • (as when the consumer peruses options stocked at the grocery store)
    • memory based
      떠올려
      • from decisions in which consumers on their own recalled the identity of relevant options
        • (as when the consumer draws up a shopping list while at home).
  • System 1 vs System 2
    • For instance, Kahneman and Frederick (2002) contrast what they call system 1 and system 2 mental processing. They describe system 1 as automatic, rapid, associative, and affective and system 2 as controlled, slow, deliberative, and deductive.
• KEY
  • memory-based → effortful → system1 (automatic) Vs. stimulus-based → pentiful mental resources → system2 (controlled)
    • We argue that
      memory-based - system1(automatic), stimulus-based - system2(controlled)
      • memory-based choices tend to reflect relatively more system 1 processing,
      • whereas stimulus-based choices reflect relatively more system 2 processing.
        • Thus, memory-based choices tend to favor
          • immediately compelling, affect-rich options,
            이미 회상하면서 에너지를 쓰기 때문에 바로 유혹되는 매력적인 제품을 선택해
        • whereas stimulus-based choices tend to favor
          고민해야 매력적인 제품을 선호해...
          • affect-poor options whose attractiveness emerges only given more deliberative thought.
      • A key premise underlying our claim is that memory-based choices are made in a context of relatively depleted processing capacity.
        • Recalling a consideration set of relevant options is effortful.
        • More critical, in memory-based choice,
          • a consumer engages in the task of identifying a favorite from his or her consideration set
            • while also devoting resources to maintaining this set in working memory.
              목록들을 계속 떠올리는데 working memory 사용
      • In contrast, under a stimulus-based procedure,
        오히려 pentiful mental resources가 남아있어!
        • there is no need to recall the relevant options
          • and no need to maintain them in working memory;
        • thus, the task of identifying a favorite occurs in a context of relatively plentiful mental resources.
  • system 1 may tend to guide memory-based choices, whereas system 2 may tend to guide stimulus-based choices.
• Studies
  • Study1&2
    • Predictions
      • Our first two studies examine decisions that
        1. pit immediately compelling, higher-affect, hedonic, fun options (e.g., chocolate cake)
        2. against lower-affect but more sensible or appropriate options (e.g., fruit salad).
      • As predicted, we find that
        1. memory-based choices favor the higher-affect, hedonic options,
        2. whereas stimulus-based choices favor the more sensible or appropriate options.
    • Study 1
      • Hypotheses
        • H1: Stimulus-based choices will tend to favor healthy desserts, whereas memory-based choices will tend to favor tasty desserts.
          Stimulus Based -> Healty desserts
        • H1a: Recalling a list of options, writing down the list, and selecting a favorite from the list will yield choices that are equivalent to standard stimulus- based choices.
          Recall but writing down = Stimulus-based choices
        • H1b: Making a stimulus-based choice while performing the digit-maintenance task will yield choices that are equivalent to memory-based choices.
          stimulus-based-with-memorization participants = memory-based participants
      • IV
        • four between-subject conditions
          1. Memory-based participants were asked,
             • “Of the four desserts you saw listed on the blue page, which would you most prefer to have?”
          2. Stimulus-based participants were asked,
             옵션 4가지가 마지막에 주어짐
             • “Which of the following desserts would you most prefer to have?” and this question was followed by a listing of the four desserts.
          3. Stimulus-based-with-maintenance participants were asked
             7-digit memorize
             • the same question and saw the same list but with cognitive load.
               • Specifically, these respondents were asked to memorize a seven-digit sequence of numbers
                 • while indicating their choice among the desserts on the list.
          4. Memory-based-without-maintenance
             옵션 4가지 다시 떠올려보고 이제 골라라!
             • participants encountered instructions that asked them to list the four desserts they had seen listed on the blue page,
             • and then, on the subsequent page in the packet, they encountered instructions stating:
             • “Please review the desserts that you listed. Then, please circle the listed dessert that you would most prefer to have.”
             • cf) In this memory-based-without-maintenance condition,
               • participants engaged in recall,
               • but, because they created a list for themselves and subsequently chose from this list,
               • they did not need to maintain options in working memory while selecting a favorite.
               • In other words, these participants actually made their choice under stimulus-based conditions.
        • maintenance?
          • One of our aims was to disentangle the impact of recalling options from the impact of maintaining options in working memory.
      • Result
        • Fruit salad를 봐봐
          •  

            

        • memory-based choices are guided relatively more by feeling-based considerations (say, an urge for tasty food),
        • whereas stimulus-based choices are guided relatively more by cognitive or deliberation-based considerations (say, the need to obey a sensible diet)
        • and that this difference arises because,
          • as compared to stimulus-based choices, memory- based choices are made in circumstances in which cognitive resources are relatively depleted.
    • Study 2
      • To test accessibility hypothesis
      • H2
        • H2: When schools are undifferentiated in fun and academics, no memory-stimulus difference will arise; when schools are differentiated, memory- based choice will favor fun and stimulus-based choice will favor academics.
          • H2:
            • When schools are undifferentiated in fun and academics,
              • no memory-stimulus difference will arise;
            • when schools are differentiated,
              1. memory-based choice will favor fun
              2. and stimulus-based choice will favor academics.
      • IV
        • 2 (choice process: memory based vs. stimulus based) X 2 (option set: differentiated vs. undifferentiated) between-subjects design.
          • choice process
            핵심은 princeton이 지루하고 공부 빡신 학교라는 것! 상대적으로
            • manipulation of memory-based versus stimulus-based choice with a manipulation of the identity of the options under consideration
              1. Princeton이 들어간 set — higher academic quality but less fun to attend
                 Princeton = Healthy = Stimulus based
              2. 안 들어간 set
            • We, therefore, predicted that Princeton,
              • but no other option, would be significantly more popular when choice is stimulus based rather than memory based.
              • Put differently, we predicted that a memory-stimulus difference would emerge for the choice set including Princeton but not for the choice set including Northwestern.
      • Results
        • choice set including Northwestern — No difference
          • The choice set including Northwestern revealed essentially no stimulus-memory differences. Across the four options, the mean difference in market shares between memory and stimulus-based choices was only 2%, and the largest such difference was a scant 4%.
        • choice set including Princeton — stimulus based 일 때 princeton 선택 확률↑
          • A very different picture emerged for the choice set including Princeton. The market share of Princeton was only 33% under memory-based choice but rose to 52% under stimulus-based choice. Both the simple effect difference in market shares for Princeton ( ) and the interaction in- z p 4.54, p = .01 volving Princeton and Northwestern market shares (z p 3.58, p = .01) were highly statistically significant.
  • Study3
    • Predictions
      • Memory based — 가격 높은 게 좋은 퀄리티라고 생각해서 고를 거야
        • Dual process theories,
          • such as Kahneman and Frederick’s, propose a general dichotomy of
            • automatic, reflex-like, affective processing
            • versus
            • controlled, deliberate processing meant to apply to many different domains.
          • These theories are thus compatible with, and in some respects generalize, Baumeister’s model, which focuses on self-regulation.
        • Note that the conception of a system 2 that may be somewhat disabled when resources are drained is reminiscent of Baumeister and colleagues’ theory of self-control.
          • Our third experiment thus tests implications of the system 1/system 2 dichotomy outside the domain of self-regulation.
        • In that experiment, we find that price changes
          • yield a standard negative price elasticity under memory-based choice
          • but a surprising positive price elasticity under stimulus-based choice.
            • We suggest that, given taxed mental resources, memory-based participants may rely on relatively reflexive assessments.
              • Thus, they may be repelled by higher-priced items.
            • By contrast, given ample mental resources, stimulus-based participants
              • may supplement reflexive responses with deliberate reflection.
          • In particular, our data suggest that, in the context we study,
            • stimulus-based participants infer that
              • high prices signal high quality.
              • Such an inference constitutes a system 2 override of
                • system 1’s aversion to high prices
                  system1: aversion to high prices
                • and implies that stimulus-based participants may be attracted to higher prices rather than repelled by them.
                  가격 높은 것을 격퇴하지 못해...?
    • Hypothesis
      • H3: In our purchasing context, memory-based choices will show negative price elasticity, and stimulus- based choices will show positive price elasticity.
      • H3a: Stimulus-based-with-maintenance preferences should be largely equivalent to memory-based choices.
    • IV
      • 3 (choice process: memory based, stimulus based, or stimulus based with maintenance) X 2 (pricing: lower price or higher price) between-subjects design.
        • Participants were asked which of four sandwiches they preferred to purchase: turkey, ham, roast beef, or tuna. The prices of the latter three sandwiches were always $5.
        • The price of the turkey was either $3 or $7.
    • Result
      • Table 3

        • 

      • memory-based — negative price elasticity
        • Under memory-based choice,
          negative price elasticity
          • turkey received a 47% market share when it was low priced,
          • and its market share fell to 34% when it was high priced (z p 1.77, p = .05, one-tailed test).
      • In contrast, under stimulus-based choice,
        • turkey received a 35% market share when it was low-priced,
        • but its market share rose to 58% when it was high priced(z p 3.32,p= .01, one-tailed test).
        • The interaction is highly significant (z p 3.58, p =.01).
      • Furthermore, the stimulus-based-with-maintenance condition revealed negative price elasticity,
        • similar to memory-based choice:
          • turkey’s market share was 50% when low priced
          • and 25% when high priced (z p 4.81, p = .01 for the within-condition difference; z p 5.51, p = .01 for the in- teraction with the standard, stimulus-basedcondition).
        • To reiterate, these results cannot be explained by any alternative account concerning differential weighting across memory and stimulus-based choice.
        • No shift in the weight- ing of price and quality can transform unfavorable reactions to high prices into favorable reactions.

 

[글쓴이 소개]

서강대학교 수학&심리 복수전공 최우등 졸업

KAIST 마케팅 전공 석사 최우등 졸업, 박사과정 수료