Birey Ve Grup Kararlarının Bilişsel Yanılma Eğilimleri Açısından Karşılaştırmalı Analizi

Abstract

Klasik karar modellerinde, insan faydasını en üst düzeye çıkartmaya çabalayan, rasyonel bir varlık olarak kabul edilmiştir. Ancak daha sonra yapılan disiplinler arası çalışmalarla birlikte rasyonel insan modelinin gerçek hayattaki karar süreçlerini açıklamadığı görülmüştür. Özellikle risk ve belirsizlik altındaki karar süreçlerinde rasyonel insan modeli verilen kararları açıklamakta yetersiz kalmaktadır. Karar süreçlerinde klasik karar modelleri varsayımlarının ihlal edildiği farklı çalışmalarla ortaya konmuştur. Sınırlı rasyonellik kavramı ile insanın bilişsel ve bilgi kapasitesinden dolayı sınırlı rasyonel olduğu kabul edilmiştir. Daha sonra yapılan çalışmalarla davranışsal karar verme yaklaşımı öne sürülmüştür. Bu yaklaşıma göre, insanların rasyonelliği sınırsız değildir. İnsanlar karar verme sürecinde bazı sezgisel kurallar kullanabilir, duygularının ve ruh hallerinin etkisinde kalabilir ve yanılgıya düşebilir. Dolayısıyla insanlar karar verme sürecinde bazı sistematik hatalar yapabilir ve rasyonel davranış kalıplarından sapmalar gösterebilir. Geleneksel karar sürecinde insanın beklenen faydayı maksimize edecek yönde karar alması beklenir, davranışşal karar verme ise insanların davranış biçimlerini ve karar süreçlerini gözlemler ve daha sonra bunları açıklayabilecek karar modelini geliştirir. Her insanın kendine özgü bir karar modeli vardır. Davranışsal karar verme alanında yapılan çalışmalarda insanların karar verirken rasyonellikten uzaklaştıran bazı bilişsel yanılma eğilimlerinin olduğu tespit edilmiştir. Bu çalışmanın amacı birey kararlarında yanılgılara sebep olan bilişsel yanılma eğilimlerinin grup kararlarına olan etkisinin incelenmesi ve birey ve grup kararları arasındaki farklılığın araştırılmasıdır. Aileden, şirket yönetimine kadar toplumun her biriminde bir çok önemli karar gruplar tarafından alınmaktadır. Bilgi birikimlerin paylaşılması ve grup içi tartışma ve etkileşimin olması grup kararlarda farklılık oluşturabilir.Literatürde birey ve grup kararları arasında farklılığı içeren çalışmalar araştırılmıştır. Yapılan deneysel çalışmalar sonucunda bazıları anlamlı bir farklılık olduğuna dair kanıt bulurken, bazı çalışmalarda birey grup kararları arasında fark olmadığı sonucuna varılmıştır. Literatürde yer alan ve bireylerin yanılma eğilimlerini ortaya koyan sorulara benzer yapıda sorular oluşturulmuştur. Çerçeve ve demirleme etkisi, temsil etme ve ulaşılabilirlik buluşsallarının birey ve grup kararları üzerinde etkisini incelenmiştir. Literatürde yer alan çalışmalar uygulamada genellikle tek eğilime odaklanırken, bu çalışma birden fazla eğilimi incelemektedir. Literatürde yer alan deneysel yaklaşımlar incelenmiş ve uygun bir yaklaşım önerilmiştir. Uygulama İstanbul Teknik Üniveristesi Endüstri Mühendisliği bölümü lisans, yüksek lisans ve doktora öğrencileri arasında yapılmıştır. Demografik özelliklerin verilen kararları etkilemediği kabul edilmiştir. Çalışma ilk önce bireylere, bir hafta sonra üçer kişilik gruplara uygulanmıştır. Birey çalışmasına 80 kişi grup çalışmasına 28 grup katılmıştır. Verilerin düzenlenmesi için tanımlayıcı istatistik yöntemleri kullanılmıştır. Veri toplama açaşamasından sonra her bir problem için uygun istatistiksel analiz yöntemi seçilmiştir. Demirleme etkisi varyans analizi sonucunda birey ve grup kararlarının ve demirleme etkisinin olup olmamasının anlamlı bir farklılık oluşturduğu görülmüştür. Ayrıca karar türü (birey/grup) ve demirlemenin etkileşimli etkisininde anlamlı olduğu diğer bir sonuçtur. Çerçeve etkisi, ulaşılabilirlik ve temsil etme buluşsalları için birey ve grup kararları arasında anlamlı bir farklılık olmadığı sonucu çıkmıştır. Bu deneyin sonuçları deneye özgü olup farklı koşullarda ve farklı kişilere sorularak farklı sonuçlar elde edilebilir. Çalışmada incelenen dört farklı bilişsel yanılma eğiliminden birey ve grup kararları arasında anlamlı farklılık oluşturan sadece demirleme eğilimi olduğu görülmüştür. Çerçeve etkisi, ulaşılabilirlik ve temsil etme buluşsallarının birey ve gruplarda benzer yanılgılara sebep olduğu sonucu varılmıştır. Gelecek araştırmalarında her bir bireyin birey kararı ve daha sonra yer aldığı grubun kararı karşılaştırılabilir. Birey kararlarının, grup içi kararlarda hangi yönde değiştiği incelebilir. Karar farklılıklarının sebepleri bilgi paylaşımı, grup içi etkileşim, grup içi kutuplaşma ve grup içi baskınlık gibi özellikler doğrultusunda araştırılabilir.

In classical decision models, a person is accepted as a rational creature who tries to maximize the benefit. But, after the interdisciplinary studies, it was understood that this kind of rational person model does not match the real world situations. Especially, the rational person model is incapable. It is asserted with different studies that in the decision processes, the assumptions of the classical decision models are violated. With bounded rationality concept, it was accepted that a person is bounded rational because of his/her cognitive information capacity. In the further studies, behavioral decision making approach was brought forward. According to this approach, the rationality of human being is not unlimited. Human beings use some kind of heuristics and biases; and they may be under the influence of his/her emotions and different moods. Consequently, they could make some kind of systematical mistakes and show some deviations from the rational behavioral pattern. In classical decision making process, it is expected that the human being will choose the decision that causes the maximum benefit. On the other hand, behavioral decision making first observes the behavior pattern, and then develops a model that explains these patterns. Each human being has his/her own decision model. So, behavioral decision making model is interested in how human beings move. .In studies about behavioral decision making, it is confirmed that there are some tendencies which keeps the person from being rational. To put forward these tendencies, some kind of multiple-choice questions were asked to the people. It is confirmed that, people has a reference point when they move, how the event is presented effects the decisions of people, and they move according to what they had in their mind before. In the studies, it is put forward that people show different kind of tendency patterns such as framing effect, anchoring effect, availability heuristic and representativeness heuristic. The main purpose of this study is to investigate how the mentioned effects and heuristics that cause delusion in personal decisions, affects the group decision; and whether if there is a difference between personal decisions and group decision. From the family issues to the business decisions, a lot of important decision is taken by group. Does being interactive and discussable together with having lots of knowledge and experience prevent the group from delusion? In literature, the differences between the personal and group decisions investigated. As a consequent of these experimental studies, some of them found evidences that there is significant differences. On the other hand, in some studies the point is “there is no significant difference between these two.” In our experiment, similar questions that has similar structure with literature were constructed. Besides, framing and anchoring effect, and availability and representativeness heuristic on personal and group decisions were investigated. While the studies in literature generally give information only on one bias , our study includes more than one bias/tendency. The questions which include some kind of bias keeping people away from the rationality were applied, firstly personally then on groups. Because there is no certain answers of some biases like yes/no, and because these kind of questions tested people’s preferences and tendencies, and people can be tested according to the consistency of their preferences , bilateral questions were constructed. For framing effect, there were two questions; one of them was presented in a positive manner and the other one was in negative manner. On the other hand, for anchoring effect, again bilateral two questions were prepared that one of them includes a ratio and the other one does not. More over bilateral two question were prepared for subjects to test their representativeness heuristic. The questions are the same but the answers are different. To test availability and representativeness heuristic there was one question for each. In the study, different kinds of application models were investigated. After that, three different models were found in literature. Asking the prepared questions firstly to individual participant and secondly to groups consisting three members was selected as a model. H0 and H1 hypothesis was constructed for each bias. Whereas H0 is there is no difference between individual and group decision, H1 is there is difference between individual and group decision. The prepared test was applied on undergraduate, master and doctorate degree student of Istanbul Technical University. It is assumed that demographical properties don’t effect decisions. Prepared questions were asked firstly to individuals. A week later they were asked to groups consisting three members. Eighty people attended to individual application and twenty-eight groups to group application. In study, gathering required information took three weeks. Because, there were bilateral questions in both applications, the study consists of two phases. In first phase; there are some questions related with demographical properties and the three questions of the first phase. The papers of the participants who end up the application were taken and it was passed to the second phase. In both phases, names of the participants were taken. To prepare and arrange the data, descriptive statistical methods and graphics were used. For each bias, percentage of the answers of both individual and groups were demonstrated on graphics. Besides, the questions placed in study were compared to the questions placed in literature. By that way, the power of reflection of constructed questions was showed. With the framing effect; it was demonstrated that preferences of participants may change when the questions asked in different manners like positive-negative, win-lose, life-death. While individual who faced with –gain- exhibit risk averse attitude, the individual who faced with -loose- behaved as risk-taker.In anchoring effect, it was monitored that the median values of two question that includes percentage and does not include percentage, were different from each other. Among bilateral questions, only mental accounting question did not reflect the bias. It was expected that decision maker makes his/her decision based on a reference point while s/he gives a decision that include different options and also s/he chooses different preferences for same amount of loses. Decisions of individuals didn’t work in that way. For the availability heuristic, participants showed tendency to select the answer that comes to the mind first and known as deluded. Representativeness heuristic was inspected under two subtitle. After data collection phase, appropriate statistical analysis method was selected. As a result of anchoring effect and analysis of variance, it was seen that personal or group decisions, and having anchoring effect or not caused a significant difference on given decisions. Besides, it is another result that the decision type (personal/ group) and the interactive influence of anchoring is meaningful. According to the results, framing effect doesn’t have a meaningful difference on personal or group decisions. Therefore, H0 hypothesis couldn’t be refused. Rather than personal or group decisions, positive or negative presentation of the questions influences the decisions. It was indicated that groups have similar tendencies with the person. Mental accounting question didn’t have a meaningful difference on personal or group decisions. This question couldn’t explain the tendency as seen on the dual answers of personal participants. Availability bias question didn’t have a meaningful difference on personal or group decisions, either. H0 hypothesis couldn’t be rejected. On two different questions that evaluate representativeness bias, H0 hypothesis couldn’t be rejected. It results in that there is no significant difference between personal or group decisions. The consequences of the experiment is specifical. Different results can be handled with different circumstances and different people. Among the four affect and heuristics inspected in this study, the anchoring tendency has significant difference between personal and group decisions. It is demonstrated that framing tendency, availability tendency and representative tendency may cause similar delusions. By the analysis of descriptive statistics methods, it was identified that there is little difference in group decisions compared to personal decisions. However, these differences weren’t meaningful as statistically. One reason for this may be that sample size is limited. On the application of this study predecided person insists on the decision that s/he has taken beforehand as s/he participated on the group discussion with his/her own truth. On the experiential work, as the method was applied simultaneously on unpredecided and predecided groups, the given decisions may be different. Not having a week between the group and personal experimental studies, making group study just after personal application may cause different results in personal or group decisions. To summarize, the change of the participant profile in application like academician or worker, the different selection of the experimental method and the duration between the applications may result in the supporting evidence for the difference of personal or group decisions. According to the results of the study, group decisions doesn’t guarantee that group decisions isn’t deluded compared to the personal decisions. To pass on the group decision duration from the personal decision duration cannot be taken alone as a precaution. It the further studies, the decisions of each of person and the group that the person is in may be compared, how personal decisions change in a group may also be investigated. The reasons of the changes in decisions may be investigated in accordance with information sharing, synergy, polarization in group and dominance in group.