Available online at https://journal.
com/index.
php/ijrcs/index International Journal of Research in Community Service e-ISSN: 2746-3281 p-ISSN: 2746-3273 Vol.
No.
3, pp.
138-146, 2025
Cognitive Dissonance Behavior in Online Consumers (Demographic and Cross Cultural Approac.
Irwan Christanto Edy1* Business Faculty.
ITBK Bukit Pengharapan.
Karanganyar.
Indonesia *Corresponding author email: irwan.
christanto@bukitpengharapan.
Abstract This study aims to understand cognitive dissonance behavior in online consumers in a demographic and cross-cultural context.
The research variables included cognitive dissonance, visual content of the website, psychological .
motions, preferences, perception.
, demographics and cross-culture.
This study uses a quantitative approach.
Novelty is research with a demographic and cross-cultural approach in Indonesia.
Indonesia has demographic and cultural diversity.
The sampling method is purposive sampling.
The data collection technique used questionnaires.
Data was obtained by 428 respondents.
Data analysis with SEM-PLS.
This research produced a model.
The model's findings show that demographics and cross-cultural have impact on cognitive dissonance behavior in online consumers.
The implications of future research are that online marketing strategies need to consider demographic and cross-cultural factors of a country so that online marketing strategies can be more effective.
The limitation of research is that research is carried out with a quantitative approach, in the future it can be carried out with a qualitative approach or a quantitative qualitative mix method.
Keywords: cognitive dissonance, web visual content, psychological, demographics, cross-cultural Introduction Online consumer behavior is an interesting topic (Cummins et al.
, 2.
Consumer behavior discusses individual groups and organizations (Kotler & Keller, 2009.
Notoatmodjo, 2007.
Sunyoto, 2.
Online marketing is a union of internet technology and marketing (Warschauer, 2.
A website is an online marketing platform that will have an impact on consumer behavior such as perception, emotions (Marfuah & Noor, 2023.
Patil & Chordiya, 2.
In online shopping, there is a unique behavioral phenomenon, namely cognitive dissonance (Anas et al.
, 2.
Cognitive dissonance is a spending decision that is contrary to the consumer's beliefs (Leon, 1.
Cognitive dissonance is strongly influenced by psychological factors (Gykcek et al.
, 2.
The topic of cognitive dissonance, especially in online consumers, is a unique and interesting study (Lazim et al.
, 2.
There is no specific data on the exact percentage of online consumers who shop due to cognitive dissonance.
However, cognitive dissonance is a psychological phenomenon that deviates from shopping habits.
Empirical data shows that Indonesia is a developing country with internet users increasing every year (APJII.
Internet users in Indonesia reach around 221 million .
and 215 million .
9% of Internet users are men 1% are women.
The largest Internet users in Indonesia are in the trade sector (APJII.
The largest consumer group in Indonesia is the millennial generation (Secapramana, 2.
Fashion products are the most popular products for online consumers (Ryding et al.
, 2.
Fashion products are the most popular products in Indonesia (Katadata.
Wedding & Mustika, 2019.
Pattipeilohy et al.
, 2019.
Suleman, 2.
The millennial generation has gained a lot of knowledge and experience in online shopping (Dewidita, 2020.
Sebastian, 2.
Every online shopping decision is the result of a process experienced by consumers (Pires et al.
, 2.
Information is the main basis for online shopping decisions (Setiadi, 2.
Novelty research with cognitive dissonance behavior in online consumers with a demographic and cross-cultural approach, where Indonesia is a country with diverse demographics and cultures.
Indonesia has 715 regional tribes.
The limitations of previous research on cognitive dissonance are also the basis for this research.
Most previous studies only explain the impact of cognitive dissonance but have not discussed in detail the relationship between demographic and cross-cultural factors (Abelson et al.
, 1968.
Sweeney et al.
, 2.
The problem in this study is to construct a theoretical model that explains the relationship between websites, psychology, demographics, crossculture and cognitive dissonance as one integrated study with each other.
Edy / International Journal of Research in Community Service.
Vol.
No.
3, pp.
138-146, 2025
Literature Review The grand theory of this research is consumer behavior where the behavior of individuals or groups or organizations about the decision-making process or activity to acquire, choose, buy and use goods or services is also influenced by the environment (Kotler & Keller, 2009.
Lamb, 2001.
Mangkunegara, 2007.
Peter & Olson, 2.
A website is a marketing stimulus.
Stimulus is a communication stimulus that can take physical form.
or verbal that affect individual behavior (Assael, 2004.
Menendez & Jonsson, 2010.
Pluzinski & Qualls, 1975.
Setiadi, 2.
The main components that can be used for a website include logos, color palettes, typography, messaging, user experience.
the psychological concept, consumers are known as affective, cognitive and behavioral (Gagne, 1.
The affective realm is often associated with aspects of emotions, feelings, interests, attitudes, and moral compliance.
Cognitive is often linked to systems of meaning and empirical understanding through various interactions and experiences with the environment (Piaget, 1.
Emotions can be desires, hatreds, sadness, miracles, love and joy (Gunarsa, 2.
There are three types of emotions, namely: fear, anger, love (Yusuf, 2.
Cognition is associated with the ability to think, learn and process information (Suharman, 2005.
Walgito, 2.
Cognitive dissonance is a cognitive process.
Cognitive dissonance occurs when a person makes decisions that are not in accordance with personal beliefs or values (Leon.
Cognitive dissonance behavior occurs when individuals make decisions by changing beliefs.
From an empirical perspective, cognitive dissonance can occur in the purchase of fashion products, namely when consumers feel uncomfortable or anxious before and after buying fashion products.
Cognitive dissonance has a positive impact on producers but negatively impacts consumers.
Researchers observed that in Indonesia, demographic and crosscultural developments in society have an impact on cognitive dissonance behavior (Bae, 2.
The number of indigenous peoples in Indonesia will reach 4.
57 million people in 2023 (Aman, 2.
Based on previous theories and research studies, the propositions proposed in this study are:
Figure 1.
Conceptual Model of Cognitive Dissonance Forming Factors Source: (Assael, 2004.
Kotler & Keller, 2007.
Leon, 1968.
Pluzinski & Qualls, 1.
(Goleman et al.
, 2002.
Piaget, 1999.
Solomon, 2.
Methods The research design is hypothesis testing (Prajitno, 2.
The research location was conducted in Indonesia.
The characteristics of the objects are: online consumers on Shopee, including the millennial generation (Simbolon, 2.
similar products.
The bound variable is cognitive dissonance, the independent variable is the visual content of the website, the mediating variable includes emotions, preferences, perceptions, moderation variables include demographics, cross-culture.
The sampling method is a purposive sampling technique (Hair et al.
, 2.
Data were collected with a questionnaire instrument with a Likert scale.
Testing of the instrument was carried out by validity and reliability tests.
The number of samples was 428 respondents.
Data analysis using SEM-PLS (Ghozali, 2.
The reason for using SEM-PLS is because of some of the advantages it offers, especially in situations where the data do not meet the assumptions of normality, the research sample is small, or when the purpose of the research is to predict or develop new theories.
PLS-SEM is also more flexible in handling complex and multi-variable models.
Edy / International Journal of Research in Community Service.
Vol.
No.
3, pp.
138-146, 2025
Results and Discussion Male Female Total Table 1.
Characteristics of Respondents by Gender Frequency Percent Cumulative Percent Source: Data 2024 Table 1 gives an idea that most of the respondents have a female gender.
This data is only a reference in the findings of this study.
Measurement Model Analysis (Outer Mode.
, the criteria for using the data analysis technique with SmartPLS 4 in assessing the outer model are by looking at convergent validity, discriminant validity.
Composite Reliability and Cronbach's Alpha.
Convergent validity is the loading factor value of the latent variable and its indicators.
Table 2.
Outer Loading Variable Website Visual Content Emotion Preference Perception Demography CrossCultural Cognitive Dissonance Indicator X1.
X1.
X1.
X1.
X1.
X1.
X1.
X2.
X2.
X2.
X2.
X3.
X3.
X3.
X3.
X3.
X4.
X4.
X4.
X5.
X5.
X5.
X5.
X6.
X6.
X6.
X6.
Source: Data processing 2025 Website Visual Content Emotion Preference Preception Demography CrossCognitive Cultural Dissonance Edy / International Journal of Research in Community Service.
Vol.
No.
3, pp.
138-146, 2025
Based on Table 2, it is known that all indicators in this study have an outer loading value of > 0.
Indicators that have an outer loading value of > 0.
70 are interpreted as having the indicators in the variable declared to meet the Convergent Validity requirements in the adequate and good category so that this study can be continued for the next stage of validity testing.
Table 3.
Fornel-Lacker Criterion Value Table Variabel Cognitive Dissonance (Y) Cross-Cultural (X.
Emotion (X.
Demography (X.
Perception (X.
Preference (X.
Website Visual Content (X.
Cognitive Dissonance (Y) CrossCultural (X.
Emotion Demography (X.
(X.
Preception (X.
Preference (X.
Website Visual Content (X.
Source: Data processed 2025 Discriminant Validity is used to ensure that each concept in a latent variable is different from the others.
The way to test the validity of discrimination is with Fornell Lacker.
In (Wong, 2.
explained by Fornell and Larkell if the value of the square root.
The Avarange Variance Extracted (AVE) of each variable is greater than the correlation value between the variables and other variables in the model, so the model is said to have a good discriminant validity value.
In measuring the internal consistency of the measuring instrument on PLS, it is carried out using a reliability test.
Reliability is measured by 3 criteria, namely Cronback's Alpha (CA).
Composite Reliability (CR), and Average Variance Extracted (AVE).
In reliability testing, it shows the accuracy, consistency of a measuring instrument in making measurements, where reliability will refer to an instrument that is considered reliable to be used as a data collection tool because the instrument is good.
The following is presented the composite reliability value for each variable:
Table 4.
CronbachAos Alpha.
Composite Reliability.
Average Variance Exctracted (AVE) Variable Cognitive Dissonance (Y) Cross-Cultural (X.
Emotion (X.
Demography (X.
Perception (X.
Preference (X.
Website Visual Content (X.
CronbachAos Alpha Composite Reliability Average variance Exctracted (AVE) Source: Data Processed 2025 A variable is declared tested or reliable if it has a Cronbach'c alpha value greater than 0.
Based on the data in table 6, it can be seen that all constructs have met the reliability criteria, this is evidenced by a composite reliability value greater than 0.
All indicators have consistency in measuring five variables.
A variable is recognized as tested or reliable if it has a composite reliability value above 0.
Structural Model Analysis (Inner Mode.
, the outer model is a measurement model to assess the validity and reliability of the model.
Inner model is a structural model to predict causal relationships between latent variables.
The results of the inner model path diagram in this study are described as follows:
Edy / International Journal of Research in Community Service.
Vol.
No.
3, pp.
138-146, 2025
Figure 2.
Inner Model Source: Data processing with Smart PLS 3, 2025 Based on Figure 2, it can be obtained that: Website Visual Content has a positive impact on the emotions of online consumers.
Website Visual Content has a positive impact on cognitive dissonance.
Website Visual Content has a positive impact on online consumer preferences.
Website Visual Content has a positive impact on the perception of online consumers, demography has a positive impact on cognitive dissonance, but cross cultural has a negative impact on cognitive dissonance.
Positive impact means that if the causative variable increases, it will have an impact on the increase in the causal variable, and vice versa for negative influences.
Table 5.
R-Square Test Results Cognitive Dissonance (Y) R Square Adjusted R Square 0,667 Source: Data processing 2025 The value on the R-Square to see the magnitude of the impact of the free latent variable on the dependent variable.
Based on table 5, it is concluded that the R-Square value for the latent variable Cognitive Dissonance is 0.
715 which shows that the variability of website visual content, emotion, preference, perception, demography, cross-cultural has an impact of 0.
715 or equivalent to 71.
5% on cognitive dissonance while the rest is influenced by other variables that are not studied.
Effect Size (F-Squar.
, the value of F-Square (F.
is used to determine the influence of the predictor variable (X) on the dependent variable (Y).
The value of F-Square ranges from 0.
02 OF2<0.
15 indicating that the variable is having a weak effect, 0.
15OF2<0.
35 indicates that the variable is having a moderate effect, and F2Ou0.
35 indicates that the variable is having a high effect.
Based on Table 8, it can be seen the weak, medium and strong influences on each variable relationship.
Hypothesis testing in SEM (Structural Equation Modelin.
PLS (Partial Least Square.
using the bootstrap method was carried out to test the significance of the influence between variables in the model.
The bootstrap method involves resampling the original data to create multiple simulated samples.
Then, the t-statistical values obtained from these samples are compared with the critical values .
-tabl.
to determine the significance of the influence between variables.
Model Evaluation, with Statistics calculation by Bootstrapping on SmartPls 4.
0 As follows:
Edy / International Journal of Research in Community Service.
Vol.
No.
3, pp.
138-146, 2025
Figure 3.
Path Coefficient and p Values Bootstrapping on SmartPls 3.
Source : Data processing with Smart PLS 4, 2025 Table 6.
Result For Inner Wights Hypotesis Research Direct Influence Path Coefficient P Values Result Website Visual Content y Emotion Website Visual Content y Preference Website Visual Content y Perception Emotion y Cognitive Dissonance Preference y Cognitive Dissonance Perception y Cognitive Dissonance Website Visual Content y Cognitive Dissonance Demography y Cognitive Dissonance Cross-Cultural y Cognitive Dissonance Demography Moderate Website Visual Content y Cognitive Dissonance Sign Sign Sign Sign Demography Moderate Emotion y Cognitive Dissonance Demography Moderate Preference y Cognitive Dissonance Demography Moderate Perception y Cognitive Dissonance Cross-Cultural Moderate Website Visual Content y Cognitive Dissonance Cross-Cultural Moderate Emotion y Cognitive Dissonance Cross-Cultural Moderate Prefernce y Cognitive Dissonance Cross-Cultural Moderate Perception y Cognitive Dissonance Source: Data processing 2025 Based on table 6, the data that has been processed using Bootstrapping on Smart PLS4, each relationship is In this study, a bootstrap method was used on samples.
Table 6 shows the results of the test using the Smart PLS bootstrapping method and found the conclusion that there are significant direct relationships but there are not .
ignificant level 0.
A weak but insignificant influence means that there is a relationship between two variables, but the relationship is not strong enough to be considered important or meaningful in the context of the study.
This can happen for a variety of reasons, including large data variations, small sample sizes, or the influence of other variables that aren't measured.
In research, significant refers to whether an outcome is strong enough to prove that the observed relationship did not occur by chance.
If the results show that the effect is weak and insignificant, it means that even if there is little association between the variables studied, the relationship is not strong enough to be convincing that the relationship is real, or that the result cannot be explained by mere chance.
Table 6 illustrates that the influence of demography and cross-cultural mediators is not significant.
This means that even if the effect is not strong, the relationship is consistent enough that it can be concluded that the effect is real, even if it is not the result of random fluctuations in the data.
Edy / International Journal of Research in Community Service.
Vol.
No.
3, pp.
138-146, 2025
Discussion A search of previous research shows that there are limitations to the findings.
Demographic factors in the context of cognitive dissonance behavior have not been studied in detail.
Previous research noted that research on cognitive dissonance in the context of online marketing is still very limited and researchers suggest for research on cognitive dissonance in the context of online marketing (Yap & Gaur, 2.
The results of this study, which show that website visual content has an impact on cognitive dissonance, reinforce the findings of research conducted by Anas et al.
which reveal that website experience has a significant effect on online purchases (Anas et al.
, 2.
The results of this study also reinforce the findings of research conducted that suggest that cognitive processes have a significant effect on cognitive dissonance behavior (Chatterjee et al.
, 2.
The latest findings from this research are that demography has a significant effect on cognitive dissonance both as an antecedent factor and a moderation factor in the relationship of website visual content to cognitive dissonance.
The findings have not been previously researched on the relationship between demography and cognitive dissonance, but this research can strengthen and develop research conducted by Hoda et al.
in Saudi Arabia .
which states that demographics, namely gender, age, marital status, occupation, and income are significantly correlated with online shopping (Hoda et al.
, 2.
Demography has a noticeable effect on the relationship between website visual content and cognitive dissonance The results of previous research shows that research on the relationship between culture and cognitive dissonance is still very limited, especially for online consumers.
The researchers noted that the findings of this study reinforce a study from Tao and Jin .
which stated that culture influences cognitive dissonance behavior in the United States and China (Tao & Jin, 2.
This research also strengthens the research of Altanlar et al.
who stated that culture influences cognitive dissonance behavior, with case studies of western and eastern cultures (Altanlar et al.
, 2.
Although limited by previous research, the findings of this study which states that cross-cultural has a significant impact on cognitive dissonance behavior in online consumers provide reinforcement from several previous research findings in understanding the relationship between wbsite, cognitive dissonance, and culture.
Cross-Cultural has a real effect on the relationship between website visual content and cognitive dissonance.
Conclussion The results of this study have provided evidence of findings that demographic and cross-cultural have a real impact on the relationship of website visual content to cognitive dissonance behavior.
Demographic factors such as age, race, ethnicity, gender, education, religion and socioeconomic conditions have a real impact on cognitive dissonance behavior in online consumers, especially the millennial generation.
Similarly, cross-culturalism that includes aspects of values, norms, and traditions has a real impact on cognitive dissonance behavior in online consumers, especially the millennial These two factors, namely demographics and cross-culturalism, cannot be ignored as factors that shape cognitive dissonance behavior in online consumers.
Recommendations The limitation of this research is the research method with a quantitative approach.
Future research can be carried out by qualitative method or mixed method.
The future research agenda is needed to examine more deeply online consumer behavior from other approaches such as geographical approaches, future research agendas are needed to research marketing strategies on websites, research other generational consumer groups such as generation Z which is a generation that has information technology capabilities.
References