Unraveling the Fallacy of Expertise: Exploring the Influence of Product-Related Experience on Consumer Perception of Product Knowledge

Abstract

This paper investigates the impact of product-related experience on consumers’ perception of their knowledge about extended warranties for automobiles in Germany and its implications for sustainability. Utilizing a structural equation model and analyzing data from 467 participants, the study explores the influence of general self-confidence and product-related experience on both subjective and objective knowledge. The empirical findings highlight that while high general self-confidence and product-related experience significantly shape participants’ subjective knowledge, they do not have a similar impact on objective knowledge. Notably, previous possession, representing the highest level of product-related experience, emerges as the primary influencing factor on participants’ subjective knowledge. Surprisingly, subjective and objective knowledge does not significantly influence participants’ willingness to pay. The analysis presented in this paper underscores the importance of understanding the distinction between self-perceived knowledge and objective knowledge, particularly concerning extended warranties for automobiles and the need for accurate knowledge dissemination to inform sustainable consumer choices.

1. Introduction

Numerous studies have investigated various aspects of consumer decision-making, with some attempting to predict a consumer’s choice [1,2]. Most studies assume that consumers use available information, whether external, such as research on the Internet, or internal, such as previous experience with the product, to make a purchase decision. To influence a consumer’s decision, companies invest heavily in advertising to provide information about their product, hoping that it will help consumers choose their product. This strategy aligns with the concept of Homo economicus, which states that humans will ultimately make the economically best decision, creating the highest expected value [3].

Recent research suggests that decision-making competence is dependent on general intelligence and numeracy [4]. This leads to the general assumption that decision-making is mainly a matter of an individual’s ability to process information from internal (i.e., memory) and external sources (i.e., advertisements) [5]. Personality traits have also been taken into account, with research examining the influence of self-control [6], happiness [7], social uncertainty on decision-making [8], the importance of environmental sustainability of products [9,10,11], price thresholds [12], and inter-temporal decision-making [13]. Overall, economic research on decision-making currently focuses on the capacity of individuals to process a vast amount of information, which is influenced by personal traits, cultural background [14,15], and situational effects. To conclude and finalize these studies, models are created to measure and forecast these effects.

Previous research has primarily focused on the effect of ownership on decision-making using laboratory experiments [16]. It is generally believed that ownership influences the evaluation of the owned good [17], which is known as the endowment effect [18]. However, a crucial question remains open: What type of memories are created by owning an object, and how are these memories used during a later decision-making process? Recent research indicates that consumers feel more affiliated with self-produced products [19]. However, can this effect also be found for purchased goods, and if so, does it increase consumers’ willingness to pay? This study aims to bridge different types of current research on the effect of ownership by focusing on the perceived and actual knowledge created by ownership and research its effect on the price consumers are willing to pay for a product.

This paper aims to contribute to existing knowledge on decision-making and related background theory by exploring the concept of knowledge as a double-edged sword. On one hand, there is real knowledge, and on the other, there is assumed knowledge about a product. This distinction is particularly relevant for non-physical products such as insurance, where the line between fact-based knowledge and perceived knowledge can become blurred. Hadar et al. [20] have shown that financial decision-making is more reliant on assumed knowledge than on actual knowledge. This research draws on their findings and differentiation of subjective and objective knowledge to evaluate the effect of this phenomenon on extended warranties for automobiles in Germany. Extended warranties can be perceived as a means to increase the lifespan of a product, therefore persuading consumers to use it for a more extended period. As demonstrated by studies on electronic devices, consumers with extended warranties tend to quote significantly higher buy-back prices compared to those without, indicating a greater willingness to prolong the usage of their devices [21]. It is hypothesized that extended warranties offer a sense of security, encouraging consumers to use products for an extended period, thus promoting sustainability. This phenomenon is particularly relevant to the German automobile market, which is known for having a high percentage of consumers inclined toward making eco-friendly purchasing decisions. The primary challenge lies in influencing consumers to opt for extended warranties and ensuring they are adequately informed about the associated benefits.

In summary, this paper investigates how product-related experiences, specifically previous possession of an extended warranty and general self-confidence, impact consumers’ knowledge—both perceived and real—along with its influence on their willingness to pay. The central research question addresses whether these experiences lead to the development of factual, evidence-based knowledge or merely a perception of knowledge regarding extended warranties offered by a German automobile manufacturer in its domestic market. The study’s critical implications for sustainability highlight a significant gap between subjective and objective knowledge, particularly influenced by prior product possession. This underscores a potential challenge in promoting sustainable choices, as reliance on perceived knowledge, shaped by ownership experience, may lead to overlooking essential sustainability factors. The findings underscore the need to reassess decision-making processes, advocating for a more informed approach to knowledge dissemination and emphasizing the importance of accurate information for fostering sustainable consumer decisions.

2. Theoretical Framework

Consumer knowledge has been extensively studied and shown to impact information search, processing, and decision-making [22,23]. Earlier studies such as Newman and Staelin [24] attempted to measure consumer knowledge, but the perspective shifted in the mid-1980s towards operationalizing the rational and emotional aspects of consumer knowledge more clearly. Brucks [23] differentiated between actual knowledge and self-assessed knowledge (SAK) or subjective knowledge (SK), while Alba and Hutchinson [25] differentiated between familiarity and expertise. Although both concepts aim to operationalize the emotional or subjective element of information and memory, the concept of interconnection between the emotional and rational part of knowledge differ greatly: Brucks [23] does not assume a direct relationship between subjective knowledge (SK) and objective knowledge (OK), while Alba and Hutchinson [25] generally expect a positive effect of increased product familiarity on expertise.

Recent research suggests that expertise in a hedonistic domain, such as movies, wine, or beer, can decrease the intensity of emotions caused by consumption [26]. Additionally, having a high degree of consumer knowledge or expertise can also change the way individuals interact with service employees [27] and experience consumption. However, it is still unclear whether this expertise is based on real knowledge or just a feeling of familiarity interpreted as knowledge. The familiarity effect describes the tendency of individuals to invest in things that seem familiar to them as a helpful shortcut to avoid investing time and resources into gathering and processing information when choosing between different products [28,29]. However, solely relying on familiarity when making decisions can increase the possibility of not choosing the best alternative if it is not based on actual knowledge about the product. Therefore, it is crucial to investigate the difference between subjective and objective knowledge and recognize that they may not be interconnected.

In this paper, we examine the discrepancy between objectivity and subjectivity in purchase decision-making situations based on Brucks’ concept of consumer knowledge. Furthermore, we adopt the emotional and rational aspects of consumer knowledge by employing Brucks’ terminology of objective knowledge (OK) as rational and subjective knowledge (SK) as emotional, respectively.

2.1. Objective Knowledge (OK)

Objective knowledge refers to accurate information about a product that is stored in a person’s long-term memory [22]. This knowledge can be acquired through extended use or repeated purchase of goods or through extensive research. Objective knowledge is typically evaluated through a set of questions about a product [23,30].

The current models of economic theory and social policies assume that increased objective knowledge improves decision-making. A “homo economicus” with unlimited access to all necessary facts will always make the decision with the highest expected value. This concept has been applied to financial education programs, which aim to increase the population’s objective knowledge of financial instruments to help them make good financial decisions [20]. The results of Fernandes et al. [31] suggest a small but significant correlation between consumer financial literacy and several financial behaviors. However, when manipulating the amount of objective knowledge, no significant effect can be found [20].

2.2. Subjective Knowledge (SK)

Subjective knowledge (SK) refers to what a consumer believes they know about a product in comparison to its alternatives and is based on various factors such as expertise and experience [25,32]. Previous research on green consumption has categorized knowledge as either concrete or abstract knowledge, where the former is considered objective, and the latter is self-rated or self-evaluated knowledge [33]. However, it is widely acknowledged that there is a difference between a consumer’s real knowledge and their perceived knowledge [33].

Recent studies suggest that subjective knowledge has a greater impact on decision-making than objective knowledge [34]. However, the factors that contribute to and strengthen subjective knowledge (SK) are not yet well understood. This study aims to investigate the factors that shape a consumer’s self-evaluation of knowledge using real-world purchase decision data.

Park and Lessing [35] argue that subjective knowledge (SK) is more useful than objective knowledge (OK) in understanding decision-makers’ biases and heuristics. They suggest that measures of subjective knowledge should include self-confidence and general knowledge levels, both of which can influence decision-making. For instance, a low level of self-confidence may prompt a consumer to seek more information regardless of their actual knowledge level. Park et al. [36] proposed a model that explains the degree of subjective knowledge (SAK) as a function of stored product-class information (SPCI), product-related experience (PRE), and general self-confidence (GENSC), as illustrated in Figure 1.

2.3. Stored Product-Class Information (SPCI)

Stored product-class information refers to the amount of information a consumer has stored about a specific class of products. It is not independent of product-related experience (PRE), but it does not directly influence it. For example, a consumer can have much information stored about seafood in general without ever having tasted or even heard of a specific dish. In such cases, a high degree of stored product-class information (SPCI) can help a waiter to recommend a specific unknown dish, increasing the chance of a satisfying decision-making outcome for the consumer [36].

The influence of product-related experience (PRE) on stored product-class information (SPCI) is more direct. Increasing experience with a specific seafood dish increases the stored information about seafood in general. When consumers are faced with a purchase decision, having this information available can help them turn it into objective knowledge (OK) or at least increase their feeling of being knowledgeable, which is called subjective knowledge (SK). Therefore, stored product-class information (SPCI) is seen as directly influencing both types of knowledge [36].

2.4. General Self-Confidence (GENSC)

General self-confidence (GENSC) is a type of self-confidence that influences decision-making along with specific self-confidence. However, specific self-confidence requires prior experience with a particular decision-making process, such as purchasing an extended warranty for a new automobile. As a result, for this research, general self-confidence was chosen as the focus, as it is better suited to describing the decision-making of first-time purchasers [37].

Individuals with high levels of general self-confidence tend to have enhanced self-esteem, self-image, and self-worth. They are more comfortable with making new decisions, taking risks, and seizing new opportunities. Furthermore, they tend to feel confident in their choices and can handle negative outcomes better than those with low general self-confidence [38].

In this paper, we assume that general self-confidence has a direct influence on subjective knowledge (SK), which refers to a consumer’s feeling of being knowledgeable about extended warranties. It is worth noting, however, that this effect is purely subjective and abstract and has no connection to objective knowledge (OK).

2.5. Willingness to Pay (WTP)

Willingness to pay has been added to the model of Park et al. [36] following the research of Hadar et al. [20], stating that subjective knowledge (SK) and objective knowledge (OK) have a direct influence on the willingness to purchase. Previous research findings indicate that consumers are willing to pay higher prices for goods that are already familiar than for goods that are unknown to them. Figure 2 shows Willingness to pay (WTP) added to the original model.

2.6. Product-Related Experience (PRE)

Product-related experience refers to the extent of familiarity a consumer has with a product or service. A product with a familiarity score of 0 means that it has never been encountered or heard of by the consumer. In the model proposed by Park et al. [36], product-related experience (PRE) is measured based on the time spent on information search, product usage, and ownership. These measures follow the theoretical model of a high-priced purchase decision, where a consumer increases their familiarity with the product gradually through the steps of information search, test usage, purchase, and post-purchase usage. As familiarity with the product increases, it influences other personal traits as well. The model suggests that product-related experience (PRE) has a significant impact on the amount of stored product-class information (SPCI), self-assessed knowledge (SAK), and objective knowledge (OK).

However, research on the endowment effect shows a strong link between product possession and a consumer’s evaluation of it [39]. Existing studies suggest that possession alone can increase the self-assessed value of a product in markets [40]. Thus, we use a two-step approach based on the Park et al. [36] model. First, the general effect of product-related experience (PRE) on subjective knowledge (SK) and objective knowledge (OK) was evaluated. Second, the influence of possession, as a single factor within the construct of product-related experience (PRE), on subjective knowledge (SK) and objective knowledge (OK) was examined.

Based on the work of Park et al. [36], H1 and H2 were formulated as follows:

H1: 

The influence of product-related experience (PRE) on subjective knowledge (SK) is stronger than the influence of general self-confidence (GENSC).

H2: 

The influence of product-related experience (PRE) on subjective knowledge (SK) is stronger than the influence of stored product-class information (SPCI).

Based on the work of Hadar et al. [20], H3 was formulated as follows:

H3: 

The influence of subjective knowledge (SK) on willingness to pay (WTP) is higher than the influence of objective knowledge (OK).

Based on the work of Shtudiner [39], H4 and H5 were formulated as follows:

H4: 

Previous ownership of an extended warranty increases the subjective knowledge (SK) of participants.

H5: 

Previous ownership of an extended warranty does not increase the objective knowledge (OK) of participants.

3. Methodology

3.1. Product Class

For this study, we focus on extended warranties as the product class. In Germany, where the research was conducted, extended warranties are offered for various types of electronic products, such as coffee machines, computers, automobiles, televisions, and others. Therefore, it was assumed that a significant number of participants would already have experience with the product class of extended warranties. Pretests conducted at the end of 2020 indicated that, for this specific product, the influence of previous ownership significantly affects future decision-making. Additionally, since extended warranties are an insurance product, they are linked to the theory of financial decision-making, which is relevant to this study’s underlying theory based on the research of Hadar et al. [20].

With respect to sustainability, extended warranties have been found to be influencing consumers’ future purchasing behavior by extending the duration of use of a good. This effect has been researched on electronic devices, where it influenced the willingness to trade in old devices for discounts on new ones. Consumers with extended warranties significantly charged higher discount rates on the new electronic devices than consumers without extended warranties. This can be seen as a higher willingness to use the product longer instead of an early exchange for a new one. In terms of sustainability, extended warranties have been found to influence consumers’ future purchasing behavior by extending the duration of use of a product. This effect has been studied in the context of electronic devices, where it impacted the willingness to trade in old devices for a discount on new ones. Consumers with extended warranties applied significantly higher discount rates to new electronic devices compared to consumers without extended warranties. This can be interpreted as a greater willingness to use the product for a longer duration instead of opting for an early exchange for a new one [21].

3.2. Sample and Procedure

The project underwent review by the Human Research Ethics Committee regarding the data collection method and procedures and was granted expedited ethics approval based on the National Statement on Ethical Conduct in Human Research of 2007. The period of ethics approval was from 30 July 2021 to 30 July 2022, and the project was assigned ethics approval number S211607.

The survey instrument underwent pretesting to ensure that the questions, possible answers, and general format were clear and valid for participants. The pretest was conducted with 50 valid participants and reviewed by experts from the service provider Qualtrics. No changes were made to the online questionnaire after the pretest, and filter variables were set for pretest answers that appeared illogical.

The final online survey was conducted from 1 September to 25 September 2021. The decision to conduct the survey online was based on several factors. First, it allowed the research to reach more participants at a lower cost and in less time than if participants were invited to a specific location to complete the survey. Second, online research is less prone to errors and is more convenient to manage and store data accurately. Finally, online surveys allow participants to participate from a quiet and comfortable location with a high degree of anonymity, which can improve the quality and validity of the results [41,42]. Moreover, with the increased online activity in educational fields due to the ongoing global pandemic [43], it is natural to conduct academic research surveys online. A total of 504 valid responses from participants were provided by Qualtrics at the end of the period. Based on filter criteria, we deleted:

-

19 valid responses from participants owning a company car, as the research only accounts for privately used cars.

-

8 incomplete responses where at least 1 answer was skipped.

-

7 valid responses of participants stated that they do not know if they own an extended warranty for their current automobile. This information as “yes” or “no” is needed for the group comparison.

-

3 valid responses of participants that stated their willingness to pay for an extended warranty to exceed EUR 10,000. Based on all the other answers given, amounts above this value seem to be typed in randomly.

Consequently, a total number of 467 valid responses was used for the analysis.

3.3. Measures

Survey participants were required to have purchased a new Audi automobile within the six months preceding their participation date. To hone in on private consumption decision-making, participants were specifically asked to indicate whether they utilized the vehicle for personal use or as a company car.

To collect data from participants, a self-administered questionnaire was formulated, and participation was entirely voluntary, without any remuneration. Subjective knowledge (SK) measures were elicited prior to the collection of product-class information, while objective knowledge (OK) questions were posed subsequently. This sequencing was deliberately chosen to prevent potential bias, ensuring that responding to objective questions first did not impact the subsequent responses to questions about perceived knowledge.

Product-related experience (PRE) is defined by Alba and Hutchinson [25] as the sum of a consumer’s contact with a product, including exposure to advertising, information search, interactions with salespersons, and decision-making processes such as purchase and usage. To measure PRE, participants were asked to evaluate three levels of product-related experience, following the approach of Park et al. [36]. (PRE_1: very few = 1, few = 2, “ ” = 3, much = 4, very much = 5; PRE_2: yes/no/I do not know; PRE_3: never = 1, almost never = 2, “ ” = 3, sometimes = 4, often = 5).

Subjective knowledge (SK) was measured by asking participants to self-assess their knowledge of extended warranties for automobiles using three questions based on Park et al. [36] and Hadar et al. [20]. Self-assessed knowledge (SAK) was used as the measure for SK in this research. A 5-point Likert scale was provided for answering the questions. (PRE_1: very few = 1, few = 2, “ ” = 3, much = 4, very much = 5; PRE_2 and PRE_3: much less = 1, less = 2, same = 3, more = 4, much more = 5).

Objective knowledge (OK) was measured using a set of 13 multiple-choice questions developed to test participants’ knowledge of the extended warranties for automobiles offered by Audi within the German market. The questions covered duration, definitions of parts included, maintenance requirements, and available services. The percentage of correctly answered questions was used as the measure for OK. Correct statements are marked in italics in

Table 1. Constructs, indicators, and questions asked.

Construct	Indicator	Question	Source
PRE	PRE_1	How much time did you spend so far to search for information about the extended warranty offered by Audi?	Alba, Hutchinson [25]
	PRE_2	Do you own an extended warranty for your current automobile?
	PRE_3	How many times have you made use of the extended warranty so far?
GENSC	GENSC_1	On the whole, I am satisfied with myself.	Filej et al. [44]
	GENSC_2	At times, I think I am no good at all.
	GENSC_3	I feel that I have several good qualities.
	GENSC_4	I am able to do things as well as most other people.
	GENSC_5	I feel I do not have much to be proud of
	GENSC_6	I certainly feel useless at times.
	GENSC_7	I feel that I am a person of worth, at least on an equal plane with others.
	GENSC_8	I wish I could have more respect for myself.
	GENSC_8	All in all, I am inclined to feel that I am a failure.
	GENSC_9	I take a positive attitude toward myself.
SK	SK_1	How much do you feel you know about extended warranties in general?	Hadar et al. [20]
	SK_2	How much do you feel you know about extended warranties compared to your friends?
	SK_3	How much do you feel you know about extended warranties compared to a warranty expert?
SPCI	SPCI_1	Please name as many car manufacturers as possible that offer an extended warranty for their automobiles.	Park et al. [36]
	SPCI_2	Please name as many attributes of an extended warranty as possible.
OK	OK_1	The extended warranty starts immediately after the manufacturer’s warranty ends.	Developed by the authors, drawing on Hadar et al.’s work [20].
	OK_2	The extended warranty automatically ends after the contracted period.
	OK_3	The extended warranty is not limited by the mileage of the car.
	OK_4	To keep the validity of the extended warranty, servicing must be carried out in accordance with vehicle manufacturer recommendations.
	OK_5	There are nine different variants of the extended warranty available for purchase.
	OK_6	It is possible to switch between these variants at any time during the contracted period.
	OK_7	The extended warranty meets the same conditions as the manufacturer’s warranty.
	OK_8	Bodywork components, like glass and paintwork, are covered by the extended warranty.
	OK_9	Parts that are subject to natural wear and tear are not covered by the extended warranty.
	OK_10	During the repair of your vehicle, the extended warranty covers the costs for a replacement vehicle.
	OK_11	You can only claim the extended warranty at a dealership or service partner approved by Audi in Germany.
	OK_12	You cannot claim the extended warranty for automobiles that have been imported to Germany from neighboring countries.
	OK_13	The warranty for parts replaced within the terms of the extended warranty ends with the end of the extended warranty.
WTP	WTP	What amount of money would you be willing to pay for an extended warranty (2 years or 100,000 km extension) for your next Audi brand automobile?	Developed by the authors

.

Stored product-class information (SPCI) was measured by asking participants to name attributes of an extended warranty for automobiles and car manufacturers within the German automobile market that offer such a product. The number of named attributes and manufacturers was counted and taken into the model as a single variable.

Rosenberg’s 10-point self-esteem scale was used to measure general self-confidence as it is a valid measurement in various experiments [44,45], and the construct was referred to as general self-confidence (GENSC) in this research. A 5-point Likert scale (with strongly agree = 1, agree = 2, undecided = 3, disagree = 4, strongly disagree = 5) was used for the responses.

Willingness to pay (WTP) was measured by asking participants which amount of Euros they were willing to pay for an extended warranty. Participants could respond to this question with any amount of Euros they deemed reasonable.

Table 1 shows the questions asked to measure each variable.

4. Results and Analysis

4.1. Overview

A two-step approach was undertaken to test the hypotheses. First, the data were analyzed descriptively. Second, PLS-SEM [46] was used to test the proposed model. The partial least squares approach (PLS-SEM) aims to maximize the covariance between the predictor latent variable and the dependent latent variable [46]. PLS-SEM is currently preferred in business and marketing research disciplines [47]. Despite criticism [48] stating that PLS-SEM is a less rigorous and unsuitable approach for examining relationships among latent variables, it is accepted as a more robust approach for estimating the structural model [47]. Furthermore, the demand for a normal distribution of observed variables in the covariance approach is viewed as unrealistic, especially for social science disciplines. The covariance-based approach (LISREL) uses the maximum likelihood (ML) function to minimize the differences between the sample covariance and those predicted by the theoretical model. When applying the maximum likelihood function, the observed variables should follow the normal distribution, and observations should be fully independent from one another [49]. Afterwards, the hypotheses were tested. The raw data and the results for every statistical indicator used to evaluate the measurement model can be requested from the authors.

4.2. Descriptive Statistics

The descriptive statistics presented in

Table 2. Descriptive statistics and correlations among the questions used in the research.

	Variable	Mean	SD	1	2	3	4	5	6	7	8	9	10	11	12	13	14	15	16	17	18
1	PRE_1	3.61	0.84	-
2	PRE_2 (yes/no)	1.75	0.43	0.25 ***	-
3	PRE_3	3.65	0.83	0.22 ***	0.16 *	-
4	GENSC_1	1.62	0.80	−0.10	0.00	−0.03	-
5	GENSC_2	3.14	1.22	0.01	0.20 **	−0.02	0.18 **	-
6	GENSC_3	1.77	0.85	−0.05	0.06	−0.06	0.44 ***	0.17 *	-
7	GENSC_4	1.95	0.89	−0.07	0.04	0.03	0.43 ***	0.07	0.31 ***	-
8	GENSC_5	2.89	1.25	0.01	0.21 ***	0.06	0.17 *	0.51 ***	0.10	0.08	-
9	GENSC_6	2.39	1.24	−0.02	0.08	0.09	0.26 ***	0.52 ***	0.26 ***	0.26 ***	0.52 ***	-
10	GENSC_7	1.89	0.84	−0.06	0.04	−0.09	0.42 ***	0.15	0.52 ***	0.36 ***	0.10	0.20 **	-
11	GENSC_8	4.11	0.81	0.04	0.06	0.04	−0.36 ***	−0.09	−0.31 ***	−0.30 ***	−0.08	−0.19 **	−0.36 ***	-
12	GENSC_9	2.48	1.24	−0.01	0.09	0.14	0.26 ***	0.44 ***	0.15	0.12	0.44 ***	0.62 ***	0.10	−0.21 ***	-
13	GENSC_10	1.77	0.83	0.02	0.18	0.08	0.45 ***	0.08	0.43 **	0.35 ***	0.11	0.22 ***	0.39 ***	−0.27 ***	0.19 **	-
14	SK_1	3.61	0.81	0.48 ***	0.34 ***	0.32 ***	0.00	0.15	−0.02	−0.02	0.12	0.08	−0.01	0.01	0.14	0.07	-
15	SK_2	3.52	0.90	0.34 ***	0.29 ***	0.30 ***	−0.11	0.15	−0.07	−0.06	0.10	0.05	−0.10	0.10	0.13	−0.07	0.51	-
16	SK_3	3.41	1.16	0.32 ***	0.25 ***	0.36 ***	0.11	0.13	0.10	0.13	0.17 *	0.17 *	0.09	−0.07	0.24 ***	0.19 **	0.55 ***	0.48 ***	-
17	SPCI_GES	2.48	1.63	0.06	0.28 ***	−0.02	−0.23 ***	−0.10	−0.17 *	−0.04	−0.08	−0.21 ***	−0.08	0.11	−0.31 ***	−0.16	0.03	0.08	0.01	-
18	OK	53.63	12.22	−0.06	−0.02	−0.19 **	−0.11	−0.11	−0.05	−0.03	−0.11	−0.10	−0.06	0.06	−0.10	−0.10	−0.19 **	−0.05	−0.17 *	0.00	-
19	WTP	800	1880	0.01	−0.06	−0.07	0.36 ***	−0.11	0.28 ***	0.20 **	−0.01	0.09	0.24 ***	−0.07	0.05	0.19 **	−0.02	0.002	0.14	−0.13	0.03

* p < 0.10, ** p < 0.05, *** p < 0.01.

show that for PRE_1, which measures the time participants have spent gathering information about the extended warranty offered by Audi, the mean value was 3.61 with a standard deviation of 0.837. The median value for this question was 4, while the modal value was also 4, indicating that, on average, participants feel they have spent a significant amount of time gathering information for this product. PRE_2 is a yes/no answer regarding current possession of an extended warranty, and 116 participants stated that they currently own an extended warranty for their automobile, while 351 participants did not. The last item for the PRE construct is the number of times the extended warranty has been used by participants. This question had a mean value of 3.647 with a standard deviation of 0.834, and the median and the modal values were 4. This indicates that participants feel they have made significant use of the extended warranty, even if they do not currently own one. General self-confidence (GENSC) was measured using a set of 10 questions, of which 5 were asked as positive, the other 5 as negative statements. For assessment, the results of Questions 2, 5, 6, 8, and 9 were inverted.

SK_1, which measures participants’ feeling of knowledgeability about extended warranties in general, has a mean value of 3.61 with a standard deviation of 0.811. This suggests that, on average, participants feel knowledgeable about extended warranties. The second item of the SK construct, SK_2, measures participants’ feelings of knowledgeability compared to their friends. The results show a mean value of 3.411, a standard deviation of 1.157, a median value of 4, and a modal value of 4. These values indicate that participants feel more knowledgeable about extended warranties than their friends. However, compared to SK_1, the level of estimated knowledgeability decreases when participants are asked to compare themselves to other people.

The last item of the SK construct, SK_3, asked participants to evaluate their knowledge about extended warranties compared to experts in the field. The mean value for SK_3 is 3.411, with a standard deviation of 1.157, a median value of 4, and a modal value of 4. These results suggest that participants feel knowledgeable about extended warranties even compared to experts. However, the values for SK_3 are slightly lower than those for SK_2. The step-by-step decrease in mean values from SK_1 to SK_2 and SK_3 seems logical, as the knowledge level of the compared groups increases with each question.

Willingness to pay (WTP) was evaluated by asking participants to state the amount of Euros they are willing to pay for an extended warranty. The results show an average of EUR 800 with a standard deviation of EUR 1880.

The last item, which measures objective knowledge (OK), consisted of a quiz with 13 true-or-false questions. On average, participants answered 53.632% of the questions correctly (mean value), with a standard deviation of 12.22 percentage points. The lowest score achieved was 23.077%, while the highest was 92.308%.

4.3. Measurement Model

The reliability of the indicators was used to identify variables that were not suitable for the model, with a threshold of 0.7. Due to low results, General self-confidence (GENSC) was reduced to variables GENSC_2, GENSC_5, GENSC_6, GENSC_9, and GENSC_10. The variables for product-related experience (PRE) were kept in the model, even though variables PRE_2 were found to be slightly below the threshold. All the variables for subjective knowledge (SK) passed the 0.7 threshold.

Next, internal consistency reliability was tested using Cronbach’s Alpha [50] and Composite reliability, both with a threshold of 0.7. Except for product-related experience (PRE), all the constructs passed the threshold for Cronbach’s Alpha. However, as product-related experience was found to be reliable using composite reliability, it was kept in the model.

Convergent validity and discriminant validity were tested next. Average variance extracting (AVE) was used for convergent validity testing. Product-related experience (PRE) was found to be slightly below the threshold of 0.5 for this test (AVE = 0.474). However, it was still accepted as valid. The Fornell-Larcker Criterion was passed by all constructs using more than a single variable, indicating that the model passed the test for discriminant validity.

To avoid collinearity of the variables, the variance inflation factor (VIF) was used to ensure the descriptive power of the model. All results were found to be below the threshold of 3.3. The resulting model is shown in Figure 3.

4.4. Testing Hypotheses H1 and H2

For testing hypotheses H1 and H2, the thresholds shown in

Table 3. Thresholds for statistical indicators.

Value	Threshold	Source
Path coefficient β	β > 0.10	[51]
t-value (two-tailed)	t > 2.58	[52]
p-value	p < 0.05	[53]

were set for the path analysis.

The results for the paths of product-related experience (PRE), general self-confidence (GENSC), and stored product-class information (SPCI) to subjective knowledge (SK) are shown in

Table 4. Results of statistical analysis of path links.

Underlying Path Links for H1, H2 and H3	Path Coefficient (β), t-Value, p-Value
Product-related experience (PRE) to subjective knowledge (SK)	β = 0.546
	t = 13.338
	p = 0.000
General self-confidence (GENSC) to subjective knowledge (SK)	β = 0.237
	t = 4.977
	p = 0.000
Stored product-class information (SPCI) to subjective knowledge (SK)	β = 0.074
	t = 2.044
	p = 0.000
Subjective knowledge (SK) to Willingness to pay (WTP)	β = 0.055
	t = 1.400
	p = 0.162
Objective knowledge (OK) to Willingness to pay (WTP)	β = 0.037
	t = 0.888
	p = 0.375

.

Based on the results, it can be concluded that the influence of stored product-class information (SPCI) on subjective knowledge (SK) is not statistically significant, as the p-value exceeds the threshold. The influence of subjective knowledge (SK) and objective knowledge (OK) on willingness to pay (WTP) is not statistically significant, as all indicators exceed the threshold. However, the influence of product-related experience (PRE) and general self-confidence (GENSC) on subjective knowledge (SK) passed all the tests for significance. Among these connections, the strongest influence is given to product-related experience (PRE) with β = 0.573. Therefore, it can be inferred that H1 and H2 are not rejected based on the data.

4.5. Testing Hypotheses H4 and H5

The responses provided by 467 participants were divided into two groups, namely possessors (n = 351) and non-possessors (n = 116) of an extended warranty. Visual data analysis was conducted using notched box plots, which are a variant of the original box-and-whisker plots. The box plots offer the advantage of showing all five descriptive statistics simultaneously, making it easier to compare groups.

In a notched box plot, the lower end of the line represents the minimum of the data set, while the upper end represents the maximum. The horizontal line in the middle of the box is the median, and the spaces above and below the median show the first and third quartiles of the data. A dot outside of the lines represents an outlier. If the notches for the medians of both groups do not overlap, then the medians are significantly different at a 95% confidence interval, indicating a significant difference between the two groups.

The box plots for SK_1, SK_2, SK_3, and OK are shown in Figure 4. The first question answered by participants to indicate their subjective knowledge was SK1, which asked for their perceived knowledge about extended warranties in general. As the notched box plots indicate, possessors of an extended warranty generally feel more knowledgeable about this product than non-possessors on average. The median for possessors is given at 4 (“I know much about extended warranties”), while the median for non-possessors is at 3 (“I know neither much nor little about extended warranties”). The notches do not overlap and, therefore, indicate a significance of data at a 95% confidence interval.

The second question regarding subjective knowledge (SK) asked participants to indicate their feeling of knowledgeability compared to their friends. This question is useful for verifying the answers given in the first question, as it adds a neutral comparison group to the evaluation of knowledgeability. Theoretically, a participant’s friends could consist entirely of experts in extended warranties, which could significantly alter the results compared to the first question. However, it is more likely that participants have one or two friends who are more knowledgeable about extended warranties, leading to only small deviations compared to the first question. The results for SK2 support the assumption that the feeling of knowledgeability remains relatively stable for both groups when asked to compare their knowledge to that of their friends. Once again, the notches do not overlap, indicating a significant difference in data on a 95% confidence interval.

The second question regarding subjective knowledge (SK) asked participants to compare their feelings of knowledgeability to their friends. This question is useful to verify the answers given to the first question, as it adds a neutral comparison group to the evaluation of knowledgeability. The results for SK2 underline the assumption that the feeling of knowledgeability stays rather stable for both groups when asked to compare their knowledge to that of their friends. Again, the notches do not overlap, indicating a significant difference in data at a 95% confidence interval.

The last question regarding subjective knowledge (SK) required participants to indicate their feeling of knowledgeability compared to a group of experts in that field. This question adds a comparison group that is knowledgeable about the product, helping to explore how stable the feeling of knowledgeability for both groups is. The results given in Figure 4 show no significant difference for possessors between SK1, SK2, and SK3. This can be interpreted as possessors being very confident in terms of their level of knowledgeability even when confronted with experts in that field. However, non-possessors’ self-evaluation falters when confronted with experts in that field, with the first quartile moving down to answer number 2 (“I know less about extended warranties compared to experts”). This result indicates that possession has a supporting influence on self-evaluated knowledge. The notches do not overlap and, therefore, indicate a significant difference in data at a 95% confidence interval.

All variables were assessed for statistical significance using multivariate analysis of variance (MANOVA). As indicated by the notched box plots, SK1, SK2, and SK3 passed the test for statistical significance with p-values significant at the 0.00% confidence interval. In contrast, OK did not pass the MANOVA test for statistical significance, suggesting that there is no significant difference between the amount of correctly answered questions of possessors and non-possessors. The analysis indicates that possessors of an extended warranty tend to rate their subjective knowledge (SK) higher than non-possessors, while their measured objective knowledge (OK) does not differ significantly from the comparison group. Therefore, H4 and H5 are not rejected.

The last variable assessed is the amount of correctly answered questions of both groups, measuring their objective knowledge (OK). The notched box plots for both groups show a closely linked median with overlapping notches in both directions. It can already be assumed at this point that there is no statistically significant difference in objective knowledge (OK) between the group of possessors and non-possessors. This result can be interpreted as possession of an extended warranty not increasing the ability of participants to answer objective questions on the responsibilities, benefits, and contractual details of an extended warranty correctly.

5. Summary and Discussion

By assessing consumer knowledge about extended warranties in the German automobile market, this paper supports the previous findings of Park et al. [36] that consumer knowledge can be divided into perceived knowledge (subjective knowledge) and actual knowledge (objective knowledge). Furthermore, our analysis also supports the view that product-related experience and personal characteristics influence the degree to which knowledge is perceived by individuals. However, our work does not support the view that increasing consumers’ knowledge about a product affects their willingness to pay for it, which was suggested based on Shtudiner [39] and Hadar et al. [20].

The findings of this study provide valuable insights into the current research on consumer knowledge about extended warranties in the German automobile market. It shows that the main factor influencing perceived knowledge among the 457 participants was their previous ownership of an extended warranty. However, the study also indicates that this perceived superior knowledge does not translate into an increase in objective knowledge.

5.1. Discussion

Recent research on decision-making has focused on the human cognitive ability to process vast amounts of data in making decisions, as well as the effect of social uncertainty on information search behavior and the degree of self-control individuals possess. Prior studies have also explored the effect of ownership on decision-making behavior. However, the results presented in this paper suggest the presence of a bias between the two stages of the decision-making process, which had not been previously detected. The data indicates that humans accumulate a significant amount of knowledge based on product-related experience (PRE), with the highest level of PRE being associated with previous possession of a product. The findings of this study support this assumption: participants who had not possessed an extended warranty estimated their knowledge about it significantly lower than the group of previous possessors did when asked to evaluate how much they felt they knew about extended warranties in general, compared to their friends and experts. Both groups’ estimation of their knowledge was also supported by a higher general self-confidence.

The data suggests that this self-estimated knowledge, built up by personal confidence, product-related experience (PRE), and especially by previous possession, is just an illusion of real knowledge. Fact-based tests do not show any evidence that people with a high estimation of their knowledge score better results than those with a lower estimation of their knowledge. These results support the idea of a difference between subjective and objective knowledge. Additionally, the formation of strong subjective knowledge was not found to be supported by stored product-class information, as previous research has indicated. Based on the analyzed data, it appears that humans falsely believe themselves to be knowledgeable about a product by having owned it once.

The recent research on humans’ ability to process large amounts of information for decision-making raises questions about the reliability of the results when the information processed is biased by previous possession. Previous research has shown that participants tend to reduce the time spent on information searches when they believe they have sufficient information to make a decision [54]. Based on the clear results of this study, future research on decision-making should not only focus on the quantity and circumstances of information processing but also question the value and reliability of all memorized information used in the decision-making process.

This research on consumer knowledge about extended warranties in the German automobile market reveals critical implications for sustainability. The study underscores the discrepancy between subjective and objective knowledge, particularly among consumers who have previously owned extended warranties. The fallacy of personal expertise, where possession fosters an illusion of superior knowledge without a corresponding increase in objective understanding, poses challenges for promoting sustainable decision-making. As consumers may reduce information search time based on perceived knowledge, there is a risk of overlooking important sustainability factors. This study calls for a reevaluation of the reliability of decision-making processes influenced by biased information and highlights the need for sustainability initiatives to focus on accurate knowledge dissemination to foster more informed and sustainable consumer choices.

5.2. Conclusions

The analysis presented in this paper yields conclusions falling into two distinct categories: practical and theoretical. From a theoretical standpoint, it can be inferred that individuals with an extended warranty for their automobile may falsely perceive themselves as more knowledgeable about the product than they actually are. However, possession of an extended warranty does not exhibit a correlation with enhanced performance on objective knowledge questions when compared to those without such possession. The experiential aspect of owning the product fosters a fallacious sense of personal expertise. Importantly, neither the erroneously perceived expertise nor the objective expertise significantly influences the willingness to pay for an extended warranty.

Aligned with previous research, extended warranties play a role in encouraging consumers to use a specific product for an extended period, promoting more sustainable behavior. The practical objective of motivating consumers to opt for an extended warranty with their next purchase and consequently influencing their behavior toward sustainability cannot be accomplished solely by augmenting their product knowledge. This study underscores that previous ownership of an extended warranty exerts the most substantial influence on the perceived knowledge about the product, therefore increasing the likelihood of a repeat purchase. Crucially, this prior ownership does not impact the amount consumers are willing to pay for it.

These findings imply that the most pragmatic approach to bolstering the distribution of extended warranties would involve implementing a support program for new car purchases. Such a program could entail offering the product for free with every new car purchase for a limited time, therefore incentivizing adoption.