Indicators of Consumers’ Preferences for Bio-Based Apparel: A German Case Study with a Functional Rain Jacket Made of Bioplastic

Plastic pollution is an increasing global problem, however, replacing fossil resources with bioplastics made from renewable resources could be part of the solution. Currently, no research analyzing the influencing factors for consumers’ preferences for functional apparel made of bioplastic material could be found. To close this gap in research, we conducted this study with a sample of 1673 participants that were representative of the German population aged 16 years and above. We conducted a choice based conjoint analysis for a bio-based rain jacket and measured psychographic indicators that were used as covariates in the statistical estimation of participants’ preferences for the rain jacket. Our results show the high level of importance of prior product experience, Green Consumer Values and attitude towards bioplastic for selecting bio-based apparel, and thus give first insights related to the influence of psychographic characteristics of consumers when selecting bio-based apparel.


Introduction
Plastics have become an increasing global waste problem over the last decades [1][2][3][4]. Plastic waste and micro-plastics can be found in almost every ecosystem in the world. Awareness of plastic pollution in the air, on land and in the oceans has raised the need to protect these ecosystems and find cleaner solutions or even substitutes for conventional plastics. One possible strategy is changing the raw materials in plastic products by replacing crude oil with biomass. Therefore, bioplastics are considered to be part of the answer to this challenge and a substitute for conventional plastics.
Bioplastics are made of renewable resources and contain carbon dioxide sequestered by plants. Bioplastics have been part of industrial life for several decades. There is strong demand for renewable solutions with respect to plastic pollution and the substitution of fossil resources use, meaning that bioplastics are gaining importance. However, the definition of bioplastics is not easy to understand for consumers. Bioplastics can be manufactured partly or completely from biogenic raw materials from various types of plants. The raw material compounds can be, for example, bioethanol (sugar cane), oil (castor bean or canola oil), starch (potatoes) or cellulose (leaf of plants). The different types of bioplastic are not necessarily biodegradable. Additionally, current "drop-in" bioplastics are chemically identical to their petrochemical equivalents, with the only difference being that the source of the carbon is not fossil fuel. There are also "novel" bioplastics (like e.g., polylactic acid, polyhydroxyalkanoates or polyethylene furanoate), which differ in chemical structure and properties from conventional plastic [5][6][7][8].
The intensively debated food versus fuel dilemma has been examined in various studies [32][33][34]. Consumers with higher GCV possibly perceive the use of agricultural resources for bioplastic as conflicting more with food security, whereas products made from forest wood can be perceived as conflicting less with food resources. Scherer et al. [13,15] found that the resources used to create the bio-based material used for a product is relevant for consumers' preferences for products such as this. We assume that the food versus fuel dilemma influences the choice of product when the resources the product is made of are taken into account. Thus, we expect that the crop used as a source of material is a factor that influences preference for bio-based rain jackets in consumers with high GCV.
Hypothesis 2c (H2c). Higher GCV increase consumers' preferences for resources with a perceived low food versus fuel dilemma in the case of bio-based apparel.
Ellis et al. [35], Chen and Chai [36] and Umberson [37] found that environmental attitudes are positively related to a high purchase intention towards green sportswear. Furthermore, Scherer et al. [13], Kainz [18] and Kurka [19] showed that respondents with high GCV are less sensitive to increased prices for bio-based products. Thus, we assume that higher GCV would decrease the importance of low prices with respect to bio-based apparel.
Hypothesis 2d (H2d). Higher GCV decrease the relevance of low prices for bio-based apparel.
Altruism: Altruism produces prosocial behavior. As defined by Batson and Powell [38], altruism is the motivation to increase another person's welfare and has been found to be a predictor for the preference for green products and green consumption by Stern et al. [39], Pfattheicher et al. [40], Teng et al. [41] or Hefner [42]. Based on these findings, we assume that a person's level of altruism is an indicator for the preference for green apparel and is expressed by reduced relevance of paying lower prices for bio-based products.
Hypothesis 3a (H3a). Higher altruism decreases the relevance of low prices for bio-based apparel.
Lusk et al. [43] found that individuals who are more altruistic are willing to pay more for environmentally certified pork. Grolleau et al. [44] also showed that altruism influences the preference for eco-labelled environmentally friendly products. Thus, we conclude that consumers with a higher level of altruism are more likely to prefer certified bio-based apparel.
Hypothesis 3b (H3b). Higher altruism increases consumers' preferences for the certification of bio-based apparel. Attitude towards bioplastics: Hartmann and Apaolaza-Ibáñez [45] and Teng et al. [41] revealed that consumers' environmental attitudes have a positive influence on the preference for green products. Recent studies by Rumm [17], Kurka [19] and Scherer et al. [13] found a connection between environmental attitudes and preferences for bioplastic products. As bioplastics are expected to help solve environmental issues caused by conventional plastics, they can be considered environmentally advantageous. Thus, environmental attitudes and the attitude towards bioplastics seem to be strongly related. Hence, we expect that attitude towards bioplastics has a similar influence on preferences as environmental attitudes have on green products. Furthermore, environmental attitude is an indicator for the preference for bio-based apparel and a higher percentage of bioplastic in the fabric.
Hypothesis 4b (H4b). Higher attitude towards bioplastics increases consumers' preferences for a high percentage of bioplastic in the fabric of bio-based apparel.
As consumers can hardly distinguish bio-based plastic products from conventional plastic products, the recognition of bio-based products needs to be supported through visual aids, such as labelling or certification. Nam et al. [29] revealed that consumers' attitude have a positive influence on the intention to purchase green sportswear. D'Souza et al. [46] found that there is a relationship between consumers perceiving labels as relevant for the willingness to pay higher prices for green-labelled alternatives and argued that environmental attitude can be a relevant factor for perceiving eco-labels. We propose that attitude towards bioplastics as a type of environmental attitude influences preference for the labelling of bio-based apparel.
Hypothesis 4c (H4c). Higher attitude towards bioplastics increases consumers' preferences for a certification of bio-based apparel.
Trust: Trust as defined by Beierlein et al. [47] is an indicator of someone's ability to trust other persons' actions. According to Atkinson and Rosenthal [48], trust in green products is influenced by the source of an eco-label and the argument specificity of the product. Several previous studies measured green trust scales as predictors of green consumption [48][49][50][51][52][53][54]. Atkinson and Rosenthal [48] and Young et al. [49] argued a possible general influence of trust towards green products. Thus, we assume that trust is a possible factor influencing preferences for bio-based apparel generally as trust in claims can be understood as trust in agents' products. Especially eco-labels must be seen as an effort to gain consumers' trust [55,56]. Therefore, we claim that trust can be connected towards the trust into the certification of product labels of bio-based apparel.
Hypothesis 5a (H5a). Higher trust increases consumers' preferences for a certification of bio-based apparel.
For consumers to purchase products, they need to trust in the product and this is connected to familiarity with products and their origin, as found by Fandos Herrera and Flavián Blanco [57] and Jiménez and San Martín [58]. Consequently, we assume that lower levels of trust increase the preference for bio-based apparel with a biomass source that is produced nationally.
Hypothesis 5b (H5b). Lower trust increases consumers' preferences for bio-based apparel made with biomass from sources that are produced nationally.
Innovation-friendly: Consumers' innovativeness is also an important driver for the adoption of environmentally friendly product innovations, as found by Tellis et al. [59], Rahman et al. [60] and Jansson [61]. Recently, Osburg et al. [62] showed that choices for wood-plastic composites over conventional plastics correlated with the respondents' innovativeness. Furthermore, Scherer et al. [13] Sustainability 2020, 12, 675 5 of 20 revealed that ecologically sensitive respondents are more innovative. Thus, we assume that respondents that are more innovative (short scale of Tellis et al. [59] innovativeness-scale) prefer bio-based apparel with a higher percentage of bioplastic in the fabric.
Hypothesis 6 (H6). Higher innovation-friendly increases consumers' preferences for a high percentage of bioplastic in the fabric of bio-based apparel.

Data and Methods
Here, we describe how the data was collected and the theoretical background on the choice based conjoint analysis (CBC) and the covariates used in our model.

Data Collection
We collected data through an online access panel in June 2016. This was done through a subcontracted market research company, which recruited the respondents. The participants in the study received pre-determined incentives for completing the full survey. Web surveys are less costly and allow a fast data collection [63]. We aimed for a representative group of the German population aged 16 or above, with the sociodemographic characteristics being based on the German sample census from 2015 (="Mikrozensus") [64]. Quotas were set on gender, age, education and the number of inhabitants in the town of residence ( Table 1). The questionnaire of the survey included-amongst other topics-a CBC experiment, items operationalizing the chosen psychographic variables, and sociodemographic information related to the respondents. Subsequent to the data collection, we cleaned the data to remove answers that were not feasible (straight-liners, Christmas-tree-behavior, conflicting answers, incorrectly answered test-questions, etc.), which left 1,673 responses for the statistical analysis. Data management and analysis was performed using Sawtooth 8 [65] and SPSS 23 [66]. The identical data set was used in the study by Klein et al. [67] that was targeted to other research questions.

Methods
We conducted a CBC and used Hierarchical Bayes (HB) estimation to analyze the data. We included several covariates in the HB estimation, which we derived through factor analysis. In this section, we describe the approach and methods used.

Choice Based Conjoint Analysis
Choice-based conjoint analysis is a well-known and valid method to research preferences for innovations or existing products [68,69]. Even though the measured choices do not necessarily lead to purchase behavior, they provide insights into motivations for buying certain products, for example, the attributes that consumer consider important or willingness-to-pay. This enables research on the current state of consumers' demand or for marketing strategies to be developed. Choice measurements with a high number of respondents can help to reveal the predictors of decision making processes. A respondent's preference results from the chosen product concepts in the CBC. Therefore, a set of attributes with realistic and existing occurrences (levels) need to be set ( Table 2). The product concepts are randomly constructed (random tasks) or explicitly selected (fixed tasks). The fixed tasks are part of the validation of the method. By choosing the illustrated product concepts, respondents value each attribute-level. Through the utility function, it is possible to translate the choices made into the consumers' perceived preferences, which can potentially be used to predict purchase decisions [69].

Attributes, Levels and Product-Concepts
When conducting CBC-experiments, a set of attributes with different levels is needed for the product concepts. In our study, we considered attributes that either already exist or potentially play a role in the German market. We defined the attributes and levels based on expert interviews, a market analysis and a literature review ( Table 2). We found the lowest attribute-levels for "percentage of bioplastic in fabric" (20%) in the market for rain jackets in 2016. The higher levels (50%, 100%) were set in order to simulate the effects of more ambitious targets. The biogenic resources shown in the CBC contain plants/byproducts that can be used to manufacture bioplastics for use in textiles. Sanad [70] found that the origin of the apparel is relevant for consumers. Furthermore, Scherer et al. [13,15] showed that the origin of the plant material is important for consumers' preferences for bioplastic products. Thus, the attribute "origin of resource" was included in the choice set, with its levels representing the globalized economy and considering important regions in which raw material plants for apparel could be produced. According to Chekima et al. [71] and Borin et al. [72] eco-labels or certificates are of value for consumers to build trustworthy products, to communicate certain properties of a product and to promote the purchase of green products. Hence, we included the attribute "product certificate" in our CBC design and defined three certificate types frequently found on the German Sustainability 2020, 12, 675 7 of 20 market. The prices we chose for the CBC cover a wide range (69 € to 349 €), but this depicts a realistic view of the German market for functional rain jackets. Additionally, we included a "None" option that is interpreted as a choice threshold, where respondents prefer none of the three listed product concepts. Three randomly created product-concepts were part of every choice set plus the "None option". Therefore, respondents could choose one option out of four (three alternative product concepts or the "None option", Tables A1 and A2). Thereby the "None" option represented the influence on the utility of not choosing a shown product.

Factor Analysis
All factors were examined for their construct validity and internal reliability. Factor analysis is often used to reduce the dimensions, measured through certain defined items. Through factor analysis, item-groups of the measured scales that highly correlate with each other are found and can be aggregated in one influencing factor. We conducted the principal components analysis as extraction method used within the factor analysis. We applied the rotation method Varimax with Kaiser-normalization and the rotation was converged in 5 iterations. The variable "prior behavior" was not included in the factor analysis as it was measured by one item only. Table 3 shows the derived factors and the statements used for covariate-estimation in this study. The results of the original factor analysis are displayed in the Appendix A (Table A3). No evidence of multicollinearity was found. All factor loadings are reflected in the factors given (Table 3).

Mathematical Approach for the Estimation of the Covariates
We estimated the part-worths of the CBC data using the Hierarchical Bayes statistic (HB), which requires repeated choices per respondent within the CBC as it is necessary to measure variation within and between the respondents' choices. The HB models respondents' preferences as a function of an upper-level model and a lower level individual model [75]. The upper-level model describes the part-worths β i distributed with means α and covariance matrix D, subscript i is the respondent, as shown in Equation (1).
The lower-level model is described by Equation (2) and estimates the utility u k that respondent i gains through the choice for alternative k.
Furthermore, the probability p of the choice for alternative k made by respondent i is estimated by Equation (3).
The integration of the model for the psychographic features into the HB-logit-choice-model takes place by the assumption through the regression Equation (4). Where Θ is a matrix (q by b) of regression parameters, z i is a q vector of covariates, and ε i is a b vector of random error terms and D is a covariance matrix, the output β i displays the part-worths of an analyzed attribute-level for respondent i [75].
We conducted the CBC with 8 random tasks and 2 fixed tasks using the default balanced overlap estimation. The HB had 40,000 iterations using 30,000 for the statistical analysis. The estimation was computed using only the random tasks. Six covariates were included into the statistical estimation. Factor 3 Attitude towards bioplastics Below you can see statements from various organizations on bioplastics. Please state per statement whether you would convince them or prevent them from supporting increased use of bioplastics. If the statement does not matter for your rating, select "I do not care." 1. The long-term goal is to produce bioplastics from non-edible plant residues. 2. Bioplastics can withstand just as much as conventional plastics. 3. The carbon footprint of bioplastics is lower than that of conventional plastics. 4. The plants for bioplastics come from non-regional cultivation. 1 = I do not agree at all; 2 = I rather do not agree; 3 = I do not care; 4 = I rather agree; 5 = I totally agree 0.782 Reformulated according to [17]

Results and Discussion
Below, we show the relative importance and the utilities of the tested attribute levels as computed through the HB estimation and compare our results with the existing literature. Additionally, we show the estimation of the covariates in Section 4.3 and discuss their influence on consumers' preferences for selected attributes of bio-based apparel. Table 4 shows the average relative importance values of the attributes of the functional jacket based on the HB estimation. Price was the attribute with the highest average relative importance (45.3%). The relevance of price for green apparel is well known and supported by previous studies described in the literature [27,70,71]. Product certificate showed the second highest average importance, which indicates the importance of labeling and information for many consumers. This is also in line with other studies [71,76]. The average relative importance of the attributes "percentage of bioplastic in fabric" and "origin of resource" had very similar values. The least important attribute was the "biogenic resource" itself. It seems that abstract information provided through certificates was more valuable to consumers than detailed information like the exact source of the biomass. Several previous studies have confirmed the importance of eco-labels for the awareness and the consumption of green products [20,21,53,71,[77][78][79].  Table 5 displays the part-worth utilities of the levels of the analyzed attributes. Our results show that 100% bioplastic in the fabric had the highest marginal utility within this attribute. The highest utility within the biogenic resource attribute was for the level chipped wood. Forest wood as a raw material is not directly related with the production of food crops on arable land and therefore does not contribute to the food-versus-fuel dilemma. This could be the reason for respondents' preference for biomass from this source in comparison to maize or potatoes. The European Union as location for the raw material plants had the highest marginal utility for respondents and was preferred over all overseas locations. This finding is also supported by Scherer et al. [13,14], who found a preference for national or regional production of raw material plants used for bioplastic products. The marginal utilities of the product certificates "free of pollutants", "climate protection" and "fair production" ranged from 17 to 20, whereas the level "no certificate" had a high negative utility value. This supports the value of certificates for informing consumers about the specific characteristics of green products, which is in line with previous studies [29,46,76,80]. The highest marginal utility within the price attribute was found for the lowest price of 69 € followed by 109 €. Prices are often important for consumer choices and the majority of respondents also preferred low prices also for bio-based products [13,14]. The "None" option, meaning that no product was chosen in the CBC, had a mean utility value of 37.4. This means that the sum of the utilities of a product concept with the displayed attributes and levels must show a higher value than the utility of the "None" option so that consumers prefer this concept. Otherwise, they would decide not to choose any of product alternatives offered in the CBC.

Influences of the Covariates
Covariates interpretation: As displayed in Table 3, the psychographic scales of the survey were factorized and the scores for each factor were implemented as covariates in the HB estimation. The results of this analysis are shown in Table 6 and illustrated in Figure 1. The intercept shows a respondent's individual utility considering the zero centered means of all factors. Additionally, the intercept shows the characteristic of having prior product experience, whereas the covariate product experience shows the theta-values of having no prior experience. We considered the significant theta-values of the included covariates for the interpretation. Significant positive theta-values lead to an increase of the individual utility of a respondent for a specific attribute-level if the measured value of the covariate is above the mean given all other variables are held constant. For example, a person's utility for the attribute-level "100% bioplastic" was estimated as shown in Equation (5). The figure of the intercept (1.13) added by the product of the theta-weight (0.36) and the measured value of "attitude towards bioplastics" of an individual results in the utility of the attribute level for the individual.
β 100% bioplastic = 1.13 + 0.36 × Attitude towards bioplastics (5)  including the "None" option. The expected RLH value for a chance model would be 0.25. Our computed value of 0.700 for 40,000 iterations is interpreted to be 2.8 times better than the chance level. The percent certainty typically varies between 0 and 1, with a value of 0 meaning that the model fits the data at only chance level, and a value of 1 meaning perfect fit. Our percent certainty value of 0.743 indicates that the log likelihood is 74.3% of the way between the value that would be expected by chance and the value for a perfect fit [81]. Product experience: As stated in Section 2, prior product experience with green products can be regarded as an influencing factor for purchasing bioplastic products. The theta value of the "None option" indicates a highly significant positive value for respondents without product experience and illustrates that these respondents tend to reject bio-based rain jackets. This finding is consistent with previous studies considering actual green behavior based on prior behavior [27][28][29]82]. Therefore, hypothesis H1 "No prior product experience with bio-based apparel decreases the probability of consumers' preferences for it" is supported.
Green Consumer Values: The utility for 100% bioplastics in the fabric increases with higher GCV as shown in Table 6. Additionally, a significant decrease in the utility for 20% bioplastics in the fabric is seen as GCV increases. The influence of GCV are in line with previous studies considering the consumption of green products and the influence of GCV product choice [13,15,31]. Hence, we assume that higher GCV lead to a higher preference for green functional clothes. Thus, hypotheses H2a "Higher GCV increase consumers' preferences for a high percentage of bioplastic in the fabric of bio-based apparel" and H2b "Lower GCV decrease consumers' preferences for a high percentage of bioplastic in the fabric of bio-based apparel" are supported. We also expected consumers with a higher GCV to prefer chipped wood over other resources. We postulated that consumers with green attitudes are more sensitive to the food-versus-fuel-dilemma identified in previous studies [32][33][34]. However, our results show no evidence for hypothesis H2c "Higher GCV increase consumers' preferences for resources with a perceived low food versus fuel dilemma in the case of bio-based apparel", which we therefore reject. Additionally, with increasing GCV, the utilities for the pricelevels 349 € and 239 € increase, whereas the utility for the price level of 69€ decreases. The influence of GCV on respondents accepting higher prices was also found in previous studies [13,18,19]. Hence, Fit-statistics: The root likelihood (RLH) is the geometric mean of the predicted probabilities for the concepts the respondent actually chose. This data set has four alternatives per choice task, including the "None" option. The expected RLH value for a chance model would be 0.25. Our computed value of 0.700 for 40,000 iterations is interpreted to be 2.8 times better than the chance level. The percent certainty typically varies between 0 and 1, with a value of 0 meaning that the model fits the data at only chance level, and a value of 1 meaning perfect fit. Our percent certainty value of 0.743 indicates that the log likelihood is 74.3% of the way between the value that would be expected by chance and the value for a perfect fit [81].
Product experience: As stated in Section 2, prior product experience with green products can be regarded as an influencing factor for purchasing bioplastic products. The theta value of the "None option" indicates a highly significant positive value for respondents without product experience and illustrates that these respondents tend to reject bio-based rain jackets. This finding is consistent with previous studies considering actual green behavior based on prior behavior [27][28][29]82]. Therefore, hypothesis H1 "No prior product experience with bio-based apparel decreases the probability of consumers' preferences for it" is supported.
Green Consumer Values: The utility for 100% bioplastics in the fabric increases with higher GCV as shown in Table 6. Additionally, a significant decrease in the utility for 20% bioplastics in the fabric is seen as GCV increases. The influence of GCV are in line with previous studies considering the consumption of green products and the influence of GCV product choice [13,15,31]. Hence, we assume that higher GCV lead to a higher preference for green functional clothes. Thus, hypotheses H2a "Higher GCV increase consumers' preferences for a high percentage of bioplastic in the fabric of bio-based apparel" and H2b "Lower GCV decrease consumers' preferences for a high percentage of bioplastic in the fabric of bio-based apparel" are supported. We also expected consumers with a higher GCV to prefer chipped wood over other resources. We postulated that consumers with green attitudes are more sensitive to the food-versus-fuel-dilemma identified in previous studies [32][33][34]. However, our results show no evidence for hypothesis H2c "Higher GCV increase consumers' preferences for resources with a perceived low food versus fuel dilemma in the case of bio-based apparel", which we therefore reject. Additionally, with increasing GCV, the utilities for the price-levels 349 € and 239 € increase, whereas the utility for the price level of 69€ decreases. The influence of GCV on respondents accepting higher prices was also found in previous studies [13,18,19]. Hence, hypothesis H2d "Higher GCV decrease the relevance of low prices for bio-based apparel" is supported.
Altruism: With increasing levels of altruism, the utilities for the higher price-levels increase, whereas the utilities for the lower price-levels decrease (Table 6). Thus, the more altruistic a person is, the less price sensitive she or he is with respect to bio-based products. This result is also in line with previous studies of Scherer et al. [15], Stern et al. [39], Lusk et al. [43] and Thøgersen and Ölander [83] who found that altruistic values are predictors for a higher willingness to pay for more environmentally-friendly products. This supports hypothesis H3a "Higher altruism decreases the relevance of low prices for bio-based apparel". Additionally, with increasing levels of altruism, the utility for the certificate "free of pollutants" increases. In contrast, the higher the level of altruism of a respondent, the lower the utility for a rain jacket without a certificate is. Thus, hypothesis H3b "Higher altruism increases consumers' preferences for a certification of bio-based apparel" is supported.
Attitude towards bioplastic: A higher positive attitude towards bioplastics increases the utility for 100% bioplastics in the fabric, whereas the utility for 20% bioplastic decreases ( Table 6). This result shows that a strong attitude towards bioplastics is positively related to the preference for functional clothes made of bioplastic. In addition, the utility of the "None option" of the covariate attitude towards bioplastic shows a negative utility value. Therefore, hypotheses H4a "Higher attitude towards bioplastics increases consumers' preferences for bio-based apparel" and H4b "Higher attitude towards bioplastics increases consumers' preferences for a high percentage of bioplastic in the fabric of bio-based apparel" are supported. With an increasingly strong attitude towards bioplastic, the utilities for the certificates "protects the climate" and "fair production" increase. In contrast, a significant decrease in the utility for no certificate is seen for an increasingly strong attitude towards bioplastic. Thus, hypothesis H4c "Higher attitude towards bioplastics increases consumers' preferences for a certification of bio-based apparel" is supported. Trust: The higher the level of trust, the higher the utility for a certificate "fair production" is. In contrast, the utility for a certificate stating "free of pollutants" decreases. Thus, hypothesis H5a "Higher trust increases consumers' preferences for a certification of bio-based apparel" is partially supported. Padel and Foster [56] revealed that locally grown organic food generates higher trust than products from abroad. Furthermore, previous studies have found the influence of trust on familiarity and the origin of products for the choice for them [57,58]. However, we could not find a relation between a low levels of trust and lower preferences for bio-based apparel produced abroad. Therefore, hypothesis H5b "Lower trust increases consumers' preferences for bio-based apparel using biomass sources that are produced nationally" is rejected.
Innovation-friendly: The higher the levels of innovativeness, the higher the utilities for 100% bioplastic and the lower the utility for 20% bioplastic in the rain jacket (Table 6). Thus, hypothesis H6 "Higher innovation-friendly increases consumers' preferences for a high percentage of bioplastic in the fabric of bio-based apparel" is supported. This result is in line with previous studies showing that consumer innovativeness promotes the adoption of innovative green products [13,59,62].

Methodological Limitations, Conclusions and Implications
Methodological limitations: As with all empirical work, this study is not without some limitations. First, respondents might have been confused by the relatively high prices of the rain jacket. However, these prices represent the current market prices of functional rain jackets in Germany in a realistic way. We even found higher prices for a superior quality jackets. Second, selection bias through the online access panel cannot be excluded, although it seems unlikely. The quotas were set considering important sociodemographic characteristics, which should ensure that a representative sample is obtained from the crowd of the panel. In this sense, the results of this study are assumed to be representative for the population in Germany. Third, all responses were self-reported and the buying process was simulated by the CBC and was therefore hypothetical. In future, a study investigating a real market situation may be a promising approach to identify further important attributes, such as style of the product or other haptic aspects of bio-based apparel. Fourth, this study was conducted in an online access panel in Germany and an international comparability of the results always faces possible influences of cultural differences related to the market or society of other countries. Therefore, results of this study should be interpreted with caution when transferring them to other countries. Future studies in other countries could improve insights on consumer's preferences for bio-based apparel in an international context. Fifth, panel-members seem to be familiar with questionnaires, which possibly influences their responses [84,85]. Conspicuous responses were cleaned from the data set as mentioned in Section 3.1 in an attempt to minimize the potential effect. In conclusion, the applied methods and their statistical analysis show results that are strongly related to previous studies. In addition, 9 of 12 hypotheses are supported and one more is partially supported.
Conclusions and implications: The major contribution of this study is to reveal the factors influencing consumers' preferences for bio-based apparel. We examined the importance and the utilities of the attributes of bio-based apparel and investigated the factors significance and their influential strength. In this study, prior product experience, GCV and attitudes towards bioplastics are the strongest factors promoting the preference for bio-based apparel. Blesin et al. [9] stated that only 12% of German consumers have product experience with bioplastics. Hence, it is advisable to increase the perception of bioplastic products in the general public and address further marketing efforts to interested groups.
Since product experience has a significant influence on the preference for bio-based apparel, it is necessary to increase the awareness of bioplastics to promote bio-based apparel. This might have the effect that more consumers intend to buy bio-based apparel and gain product experience, which promotes a preference for bio-based apparel. Possible instruments to increase the awareness are the presentation at the point of sale for better product attention, a targeted communication for specific product segments or nudging. Furthermore, policy makers should elaborate a clear definition of bioplastics and improve labeling of bioplastic products to support the awareness of them.
Consumers with higher preferences for bio-based apparel show higher GCV and attitudes towards bioplastics. Moreover, a European origin of the vegetable raw materials used is preferred by green consumers as well as high levels of bioplastics in the fabric. Thus, consumers with these preferences should be strengthened in their confidence in bio-based products. A largely positive attitude towards bioplastics in the population should be maintained and-if possible-expanded. This can be supported above all by clear claims and transparent value chains. This also includes recognizable labeling, verified life cycle assessments and standards. Policy should set the rules in these areas, and state institutions should be responsible for reviewing these tools to create security, reliability and trust among consumers.
Finally, the identified influencing factors should be taken into account and used in the "emotional" design and the tonality of communication measures. It is advisable to put positively assessed product properties of bioplastic products in the foreground and consider them in communication measures. Funding: This research with the title: "New pathways, strategies, business and communication models for bioplastics as a building block of a sustainable economy" was funded by the German Federal Ministry of Education and Research (BMBF). The responsible body for administrating the project was the German Aerospace Center (DLR) (support code for the project: 01UT1430C). This work was supported by the Technical University of Munich within the TUM Open Access Publishing Fund.

Conflicts of Interest:
The authors declare no conflict of interest. The funders had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript, or in the decision to publish the results. Table A1. CBC introduction text.

Appendix A
First, please put yourself in the following situation: You want to buy a new rain jacket. You are looking for a high-quality product from a brand you know in an outdoor clothing store. After a short search, you will find what you are looking for and have the choice between different alternatives. Size, color, cut and function of the jacket meet your expectations. The seller points out that some jackets are made of bioplastics. He goes on to say that, various renewable raw materials were used instead of petroleum for its production, but that no impairment of product quality is to be expected. He advises you further that the raw materials come from different countries of origin, and there are also various certifications on the rain jackets, which say something about production or material of functional clothing.