Naslov (srp)

Neuromarketinški pristup i primena Bajesovih modela u analizi efekata regulacionog stanja na odlučivanje potrošača

Autor

Čolić, Lazar, 1987-

Doprinosi

Milošević, Bojana, 1989-
Chroneos Krasavac, Biljana, 1964-
Ognjanov, Galjina, 1974-
Prica, Ivana, 1967-

Opis (srp)

Analiza psihološkog (regulacionog) stanja u kontekstu odlučivanja potrošača je jedan od izazova ove disertacije. Budući da tradicionalne metode samoocenjivanja ispitanika nisu pouzdane u merenju regulacionog stanja, korišćen je neuromarketinški instrument elektroencefalograf (EEG). Na ovaj način je omogućeno simultano praćenje neuralnih reakcija tokom tretiranja ispitanika eksperimentalnim stimulusom (muzičke numere različitog tempa i energičnosti) i elicitacije njihovih preferencija. Za ocenjivanje preferencija potrošača korišćen je hijerarhijski Bajesov model izbora uz korišćenje neuromarketinških metrika dobijenih putem elektroencefalografa. Za ocenjivanje modela primenjen je Hamiltonov Monte Karlo metod korišćenjem NUTS algoritma zbog sposobnosti da oceni aposteriorne raspodele kompleksne strukture, kao što je to slučaj kod hijerarhijskih modela sa malim uzorcima korišćenog u ovom radu. Model je programiran u statističkom programskom jeziku Stan. Za potrebe elicitacije preferencija korišćen je Bajesov D-optimalan eksperimentalni dizajn, uz korišćenje apriornih informacija iz predistraživanja. Dobijenim modelom analiziran je uticaj regulacionog stanja na cenovnu osetljivost tražnje nad studentskom populacijom Univerziteta u Beogradu. Osnovni naučni cilj empirijskog istraživanja je bio da objasni uzročno-posledičnu vezu muzike koja se čuje u pozadini („pozadinske muzike“) i regulacionog stanja potrošača, sa jedne strane, i odlučivanja potrošača o izboru proizvoda, sa druge strane.U skladu sa teorijom regulacione usklađenosti, rezultati su pokazali da promociono regulaciono stanje smanjuje cenovnu osetljivost tražnje, dok je prevenciono pojačava. Ovaj rezultat je važio kako za hronično regulaciono stanje, tako i za situaciono regulaciono stanje. Kod merenja hroničnog regulacionog stanja, tradicionalna metrika merena BIS/BAS skalom nije bila značajan prediktor cenovne osetljivosti, dok je neuromarketinška metrika frontalne asimetrije koju pruža EEG, bila. Situaciona regulaciona stanja promocije i prevencije su primovana muzičkim stimulusima različitih nivoa energičnosti, odnosno korišćena je spora i brza verzija numere „Wake me up“ originalno izvedena od strane umetnika Avicii-a. Pokazano je da energična muzika uvodi potrošače u promociono regulaciono stanje, koje se reflektuje kroz veću emocionalnu pobuđenost ispitanika, što posledično smanjuje cenovnu osetljivost tražnje. Sporija verzija muzičkog stimulusa je primovala prevenciono stanje, reflektovano kroz stanje relaksacije (niska emocionalna pobuđenost), što je pojačavalo cenovnu osetljivost tražnje. Dobijena veza regulacionog stanja i cenovne osetljivosti je važila samo u uslovima kada potrošači nisu bili involvirani u obradi informacija prilikom izbora proizvoda, čime su potvrđeni rezultati iz ranijih istraživanja.Metod razvijen u ovom radu može se, osim muzike, primeniti na različite kontekstualne faktore koji primuju regulaciona stanja potrošača. Moguće je primeniti i druge neuromarketinške instrumente, poput uređaja za praćenje pogleda ispitanika, čime bi merenje psihološkog stanja bilo sveobuhvatnije. Apriorne raspodele korišćene prilikom ocenjivanja hijerarhijskih modela u ovom radu mogu poslužiti kao dodatne informacije u istraživanjima gde se koriste hijerarhijski Bajesovi modeli izbora.

Opis (srp)

Društvene nauke / Statistika i matematika Social sciences / statistics and mathematics Datum odbrane: 25.01.2024.

Opis (eng)

One of the challenges tackled by this thesis is the analysis of the psychological (regulatory) states within the context of consumer decision making. Since the traditional methods of interviewee self-evaluation are not reliable in estimating regulatory states, a neuromarketing instrument electroencephalograph (EEG) is used in the thesis instead. The latter allows for simultaneous tracking of neural reactions while the interviewee is being treated with the experimental stimulus (music numbers with different tempo and energy) and elicitations of their preferences. In order to estimate consumer preferences hierarchical Bayes model of choice using neuromarketing metrics obtained from electroencephalograph was used. The model is estimated using Hamilton’s Monte Carlo method with NUTS algorithm. This approach is chosen because it is the preferred approach for measuring a posteriori distributions of complex structure, such as the hierarchical models on small samples that are employed in this thesis. The model was coded using statistical programming language Stan. To elicit preferences, Bayes D-optimal experimental design was employed, using a priori information provided from the pre-research. The estimated model is used to analyse the effect of regulatory states on price elasticity for the student population of the University of Belgrade.The main scientific goal of the empirical research was to explain the causal relationship between the background music and the regulatory state of the consumer, on one hand, and the consumer decision on which product to choose, on the other hand. In line with the theory of the regulatory fit the empirical research results show that promotional regulatory state lowers the price elasticity, while the preventative state increases the price elasticity. This result was obtained in case of both chronic and situational regulatory state. In measuring the chronic regulatory state, the empirical research findings are that the traditional metrics provided by BIS/BAS scale did not provide a significant predictor of the price sensitivity, unlike the neuromarketing metrics of frontal asymmetries provided by the EEG. Situational regulatory states were primed using music stimuli of varying level of energy (tempo), i.e. slower and faster versions of the song “Wake me up” originally performed by Avicii. It was shown that the higher energy (faster) music lead the consumers to the promotional regulatory state, as reflected by higher emotional excitement of the interviewees, and this lead to the lower price elasticity. Inversely, the slower musical stimulus primed the preventative state, where the interviewees were in a relaxed state (low emotional excitement), and that increased the price elasticity of demand. The resulting relationship between the regulatory state and the price elasticity was applicable only when the consumers were not involved in processing of the available information while deciding which product to choose. These findings are in line with the already existing research.The method developed in this thesis, besides music, could be applied to other contextual factors that prime the regulatory states of consumers. It is also possible to use other neuromarketing instruments, like the ones following the eye movements, that could help to achieve a more encompassing measure of the psychological state. A priori distributions used to estimate the hierarchical models in this thesis may also be used as an additional information for other research that employs the hierarchical Bayes models of choice.

Jezik

srpski

Datum

2023

Licenca

Ovo delo je licencirano pod uslovima licence
Creative Commons CC BY-NC-ND 3.0 AT - Creative Commons Autorstvo - Nekomercijalno - Bez prerada 3.0 Austria License.

http://creativecommons.org/licenses/by-nc-nd/3.0/at/legalcode

Predmet

OSNO - Opšta sistematizacija naučnih oblasti, Ekonomske nauke

regulatory focus theory, Bayesian statistics, neuromarketing, electroencephalography, price sensitivity, choice modeling, conjoint analysis

OSNO - Opšta sistematizacija naučnih oblasti, Ekonomske nauke

teorija regulacionog fokusa, Bajesova statistika, neuromarketing, elektroencefalografija, cenovna osetljivost, modeliranje izbora, analiza združenih efekata