Abstract
This article explores consumers' preferences for different chestnut (Castanea spp.) attributes and studies differences across potential market segments. The study was conducted between 2003 and 2007 during the Missouri Chestnut Roast festival. The festival, held annually in October during the chestnut harvest season, is one of mid-Missouri's premier family-oriented events. A longitudinal study completed among festival visitors in 2003, 2004, and 2006 to identify chestnut characteristics that influence purchasing decisions was complemented with a conjoint analysis in 2007. The conjoint analysis used a conditional logit model to investigate responses from pairwise product profile comparisons. The attributes investigated include chestnut size (small, medium, and large), price ($3, $5, and $7 per pound), production process (organic and conventional), and origin (Missouri, United States, and imported). Results suggest a strong preference for locally and U.S.-grown compared with imported chestnuts. Local growers that provide the market with medium-size chestnuts that carry organic certification could command a market premium compared with imported/nonorganic certified chestnuts.
Edible chestnuts were well known in the United States before their near extinction at the beginning of the 20th century. The accidental introduction of the chestnut blight (Cryphonectria parasitica) from Asia in the early 1900s almost eradicated american chestnut trees (Castanea dentata) (Anagnostakis, 1987). Research efforts are currently being conducted in Missouri, Michigan, Tennessee, and other states to identify improved cultivars and to develop management practices suitable for commercial chestnut production. These initiatives aim to revitalize the chestnut industry and to provide an alternative crop to growers and a “new” product to consumers.
Chestnuts have favorable nutritional characteristics. High moisture and low oil content make chestnuts virtually fat free. Chestnuts have a high concentration of complex carbohydrates, a low glycemic index, are cholesterol free, and contain only one-third the calorie content of peanuts (Arachis hypogaea) and cashews (Anacardium occidentale) and as much ascorbic acid as their equivalent weight of lemons (Citrus limonium). Chestnut flour is sweet, nutty, and gluten-free, the latter making it appropriate for consumption by people with celiac disease (Chestnuts Australia Inc., 2008; Erturk et al., 2006; University of Missouri Center for Agroforestry, 2006).
The market for edible chestnuts has considerable potential for increase in production and demand given growing consumer interest in alternative and healthy foods (Gold et al., 2004a). Consumers across Europe, Australia, New Zealand, and the United States have an increasing interest in chestnuts (Kelley and Behe, 2002). Current demand for chestnuts in the United States exceeds national production, which is offset by imports. According to the U.S. Department of Agriculture (USDA), chestnut imports have grown steadily in value since 2003, reaching $11.6 million in 2006 (USDA, 2007). Chestnuts are mostly imported from Europe (82% of total imports in 2006). According to the USDA (2007), the main exporters to the United States are Italy (55%), China (23%), and Korea (15%).
In the United States, chestnut cultivation can be an attractive enterprise due to high product demand, favorable prices, and relatively low initial investment requirements (Gold et al., 2006). However, aside from a small number of successful enterprises, the U.S. chestnut industry is in its initial stages of development, with the majority of producers in business for less than 10 years and just beginning to produce at a commercial scale (Gold et al., 2006).
At the other end of the market value chain, U.S. consumers are unfamiliar with chestnuts, are not fully aware of their healthful properties, in what form and where to buy them, or how to prepare them (Gold et al., 2005). In regions surrounding successful growers or active research centers, consumers' familiarity with chestnuts, and knowledge and frequency of consumption is continually increasing (Carlman, 2007; Cernusca et al., 2008). This situation represents a market opportunity for expansion of chestnut production in the United States that demands a better understanding of consumer preferences.
Market knowledge is fundamental for the successful establishment of agroforestry enterprises (Gold et al., 2004a). Gold et al. (2006) consider that one of the biggest barriers to success in the chestnut industry is the lack of information for producers, retailers, and consumers. Over the last two decades there have been a limited number of studies focused on improving understanding of chestnut markets (Gold et al., 2004b, 2006). A study by Wahl (2002) assessed the interest of upscale restaurant chefs in value-added chestnut products and found that product freshness and quality were the main attributes that influenced consumers' and chefs' interest in purchasing food products and ingredients. Wahl (2002) recommended that growers should promote chestnuts on the basis of quality, uniqueness, and local production. Another study developed by the Midwest Nut Producers Council and Michigan State University identified marketing opportunities for chestnuts in upscale restaurants in Michigan (Kelley and Behe, 2002; Smith et al., 2002). Chefs that participated in the study preferred peeled chestnuts and incorporated them in a variety of dishes. Gold et al. (2004b, 2005) also reported that U.S. consumers prefer to buy chestnuts from grocery stores or farmers markets and that organic, brand name, and cultivar chestnuts can help capture price premiums.
This study was designed to further current knowledge of consumer preferences for chestnuts. There was a need to identify salient product characteristics that influence consumption and that could help in fostering product demand and developing new markets. Specifically, this research aimed to 1) identify how much attributes such as size, price, quality, locally grown, and nutrition-diet-health, influence consumers' decisions to purchase chestnuts, 2) generate information on the relative importance that consumers place on particular chestnut characteristics (price, size, production process, and label of origin), 3) determine salient product characteristics that current and future chestnut producers should evaluate and adopt for effective marketing purposes, and 4) identify differences in product preferences by demographic characteristics such as urban and rural consumers and gender.
Materials and methods
A two-step research approach was used to assess consumer preferences for chestnut attributes. The first step addresses objective 1 and it involved the study of perceptions among a sample of consumers over 3 years. The second step corresponds to the application of a conjoint analysis (CA) for the exploratory study of chestnut attributes. All information reported in this article was gathered during the annual Missouri Chestnut Roast festival in New Franklin, MO. The festival held annually in October during the chestnut harvest season is one of mid-Missouri's premier family-oriented events. The one-day event is not dedicated solely to chestnuts. It is an outstanding opportunity to introduce participants to the broad range of possibilities and benefits that can result from implementing agroforestry practices or consuming specialty products. Festival participants include consumers and nonconsumers of chestnuts. As a case in point, 67% of participants in 2003, 46% in 2004, 45% in 2006, and 34% in 2007 had never consumed a chestnut before attending the festival. In addition, in 2007, about 75% of participants attended the festival for the first time.
Study of characteristics that influence purchasing decisions.
A question about how much each of the following chestnut attributes (i.e., price, quality, locally grown, and nutrition-diet-health) influence people's decision to purchase chestnuts was included in a survey that was administered over 3 years (2003, 2004, and 2006). Because very similar results were obtained in 2003 and 2004, 2005 was skipped, but the survey was retaken in 2006. The selected attributes (price, quality, locally grown, and nutrition-diet-health) were identified after a review of the literature (Jekanowski et al., 2000; Frank et al., 2001). The influence on consumer preference for each attribute was evaluated on a five-point Likert-type scale ranging from “not at all” (1) to “very strongly” (5). Each participant took between 5 and 10 min to complete the questionnaire. No incentive in the form of monetary or in-kind reward was offered to study participants.
Conjoint analysis.
CA was used to design and analyze surveys distributed during the Missouri Chestnut Roast festival in 2007. The CA helped validate and quantify the results obtained in the previous surveys and provided responses to examine the remaining research objectives. CA is based on the premise that consumers can judge the value of several hypothetical products that are described by different attributes that make up product profiles and choose the one that gives them the most utility (Green and Srinivasan, 1978; Carson et al., 1994).
Under a random utility framework, a consumer (i) has a utility function defined over an array of J possible choices of chestnut products. It is further assumed that a consumer utility function contains a deterministic component and a set of unobservable variables that introduce a random error element (Hanemann, 1984). For the ith consumer faced with a total of J choices, the assumed utility of choice j can be represented by:
where the Vs represent utility values that summarize the preference of the ith consumer for the jth product alternative. The xs specify values of chestnut attributes, and the βs denote importance weight parameters (partworths) that were used to determine the specific importance placed by consumers on each attribute. Vector ***ij, j = 1,…J captures unobservable variables affecting consumer preferences, and reflects the idiosyncrasy of the individual in taste for the different product alternatives. Under the assumption that random errors are independent and identically distributed, this model specification corresponds to McFadden's conditional logit model (McFadden, 1974, 1986; Punj and Staelin, 1978).
According to Arnold et al. (1981) the conditional logit model has several advantages, one of which is good model approximation even when working with small samples (50 observations or more). Traditional conditional logit includes product-specific attribute information in the xs, (e.g., product price, size, etc.), but cases that include respondent-specific variables (e.g., region, gender, age, etc.) can also be incorporated. The use of product- and respondent-specific variables simultaneously results in what is commonly known as a mixed-effects conditional model (Hoffman and Duncan, 1988; Long and Freese, 2006). Under this model, Eq. [1] becomes
where Si captures respondent-specific demographic information.
In this study, various models were tested. First, a fixed model with only attribute-specific characteristics was estimated. Differences in partworth parameters were then explored between place of residence (urban and rural), gender, age, and income for market segmentation purposes. Brown (2003) found that rural residents in Missouri are more willing to seek out locally produced food than are urban residents, but household location did not significantly influence the price consumers are willing to pay for locally grown food. Aguilar and Vlosky (2007) identified gender and income differences when eliciting willingness to pay premiums for environmentally certified wood products.
CA helps determine a set of partial utilities (partworths) for the individual attributes that are consistent with the respondent's overall preferences (Green and Wind, 1975). The partworths identify the relative importance that consumers place on each particular chestnut attribute. CA has been applied to a number of horticultural products, including bell peppers (Capsicum annuum) (Frank et al., 2001), peanuts (Nelson et al., 2005), tabletop Christmas trees (Behe et al., 2005), mandarins (Citrus reticulata) (Campbell et al., 2006), and asparagus spears (Asparagus officinalis) (Behe, 2006).
The first step in a CA study is the identification of key attributes and their appropriate levels. To quantify the attributes considered in the 2003–06 studies, the following product attributes and levels were selected: price ($3, $5, and $7 per pound), chestnut size [small (<1 inch in diameter), medium (>1 inch but <1-1/4 inch diameter), and large (>1-1/4 inch diameter)], label of origin (produced in Missouri, produced in the United States, and imported), and production process (conventional and organic; Table 1). The second column in Table 1 provides a brief description of the attribute level and the last column indicates expected signs for the utility partworths. The reader should note that no attribute regarding nutrition-diet-health was included. This is the case because all chestnuts products are assumed to provide the same level of nutrition. No particular country or region from where the chestnuts might be imported from was identified. Also, no specific definition to the concept of organic production was included in the survey, but if asked by participants, it was indicated that chestnuts were produced, processed, and certified to be consistent with USDA national organic standards that avoid most synthetic chemical inputs. Participants were presented with a list of product profiles as depicted in Fig. 1. No photographs or other visual aids were used in the conjoint analysis because all visitors were offered samples of chestnuts as they entered the festival.
Chestnut attributes, corresponding attribute levels, and expected partworth signs used in the conjoint analysis of consumer preferences for chestnuts in a survey distributed in 2007 during the Missouri Chestnut Roast festival.
Example of the conjoint analysis pairwise instrument used to gather consumer preferences for chestnuts' attributes (i.e., size, price, production process, and label of origin) during the Missouri Chestnut Roast festival in 2007 (small = <1 inch diameter, medium = >1 inch but <1-1/4 inch diameter, large = >1-1/4 inch diameter; 1 inch = 2.54 cm, $1.00/lb = $2.2046/kg).
Citation: HortTechnology 19, 1; 10.21273/HORTTECH.19.1.216
Differences in preferences among groups are explored in the mixed-effects conditional model by allowing for interactions with the product attributes and the socioeconomic variables presented in Table 2.
Socioeconomic variables, corresponding levels, and expected partworth signs used in the conjoint analysis of consumer preferences for chestnuts in a survey distributed in 2007 during the Missouri Chestnut Roast festival.
Preference elicitation in CA was done using the full profile pairwise comparison method as described by Hair et al. (1998). The latter was chosen primarily because it adds more realism to the product selection process. The profiles of hypothetical products were generated using the Bretton-Clark designer program following a fractional design. This program produces a subset of hypothetical profiles based on the attribute levels provided by the researcher. The program minimizes the confounding of attribute main effects by selecting a subsample of orthogonal product combinations (Harrison and Sambidi, 2004). An example of two of the nine pairwise profiles generated by the Bretton-Clark designer program is presented in Fig. 1. Eighteen profiles for a total of nine comparisons were produced. An additional 10th pair was generated by the authors to compare attribute levels that were omitted in the other nine comparisons and deemed relevant.
The generated profiles were included in a survey questionnaire together with questions about frequency of consumption, familiarity with chestnuts, and demographic information (i.e., age, gender, marital status, level of education, total household income, and ethnic group). For the CA, respondents were asked to review the 10 pairs of hypothetical chestnut products and to select one product (A or B) that they would be most likely to purchase (Fig. 1). A note also mentioned that all products meet a standard of quality that guarantees freshness, good flavor, and nuts free of fungal- or bacteria-caused damage. The survey was pretested on 20 subjects and received required human subject-approval before being distributed to participants.
Data collection and analysis.
Respondents for each of the surveys were selected following the tailored design procedure for in-person delivered questionnaires recommended by Dillman (2000). Interviewers randomly selected festival attendees, explained the purpose of the study, and asked for their participation. Those who agreed to participate completed the survey on site. The sample generated by this method is a convenience sample and implies caution in generalizing the results to the whole population of consumers. For the surveys distributed in 2003–06 and the first part of the survey distributed in 2007, frequencies, cross tabulation, and correlations were calculated to analyze data.
The information presented in the product profiles corresponded to the chestnut attributes (x). Data were entered as a series of dummy variables representing the different categories composing a chestnut profile. The only variable not categorical is price, which is continuous, and the actual price presented on the survey was entered in the data matrix. For each of the three categorical variables (chestnut size, production process, and label of origin), the base level indicated in Table 1 was left out as the point of comparison in the analysis, and a “1” or “0” was entered if the profile contained or lacked information for a corresponding attribute level. As such, after leaving out the base level in our data, we obtained two categories for chestnut size (medium and large), one category for production process (certified organic), and two categories for origin (produced in Missouri and produced in the United States). For the case of the mixed-effects conditional model, a set of binary variables were created to capture the respondent characteristics as described in the socioeconomic variables presented in Table 2, and each one was multiplied by the product attribute levels to generate interaction terms that allow for the study of differences among different group segments.
Results and discussion
Characteristics that influence purchasing decisions (2003–06 survey).
We collected 232 questionnaires in 2003, 217 in 2004, and 487 in 2006. In 2003, quality (69% of respondents were strongly and very strongly influenced by quality in their decision to buy chestnuts) and nutrition-diet-health (55%) were the most important attributes mentioned. Price was listed as the least important attribute (26%) (Gold et al., 2004b). Similar results were obtained from the 2004 data (Gold et al., 2005). In 2006, an increase was noted in the preference for locally grown chestnuts. As shown in Fig. 2, quality remained the top attribute (72% of respondents were strongly and very strongly influenced by quality in their decision to purchase chestnuts), followed by locally grown (56%), nutrition-diet-health (54%), and price (23%). Some of these results are in accordance with those reported by Brown (2003). Brown found that quality and freshness were the most important concerns when shopping for fresh fruits and/or vegetable for 82% of consumers in southeast Missouri, while price was important for only 8%.
Attributes that influence consumer decisions to purchase chestnuts as resulted from surveys distributed between 2003 and 2006 to Missouri Chestnut Roast participants.
Citation: HortTechnology 19, 1; 10.21273/HORTTECH.19.1.216
A crosstab analysis of the 2006 data revealed relationships between price and age of participants [χ2 (8) = 15.66, P < 0.05]; locally grown and age [χ2 (8) = 24.01, P < 0.01], locally grown and income [χ2 (12) = 23.36, P < 0.05], nutrition-diet-health and age [χ2 (8) = 28.19, P < 0.001], and nutrition-diet-health and income [χ2 (12) = 22.06, P < 0.05]. Participants over 55 years of age (21%) tended to be less influenced by price than people under 35 years of age (14%) or people between the ages of 36 and 55 years (10%). Participants under 35 years of age (14%) were less influenced by locally grown than participants from 36 to 55 years old (28%) or older than 55 years of age (23%). Participants between 35 and 55 years old (29%) were more influenced by nutrition-diet-health when making purchase decisions than people under 35 years of age (10%). Respondents with a total household income between $25,000 and $50,000 valued locally grown products more than respondents in other income ranges. A correlation analysis revealed a small, negative significant correlation between price and age [r(425) = −0.141, P < 0.01], the older the respondents, the less influenced they are by price when making decisions to purchase chestnuts.
CA statistics (2007).
A total of 104 surveys were collected in 2007, of which 54% were completed by women and 46% by men. A lower number of surveys reflects inclement weather throughout the 2007 festival. Because of the limited sample size available for analysis, this research is deemed exploratory. Regarding age, respondents were classified in six categories. Seven percent were under 25 years old, 12% were between 26 and 35 years of age, 13% were in the 36- to 45-year range, 29% were in the 46- to 55-year range, 24% were in the 56- to 65-year range, and 15% were over 65 years old. Regarding place of residence, 60% of the study participants lived in towns or cities with a population of at least 10,000. Forty percent came from rural areas and towns with less than 10,000 people. Twenty-eight percent of respondents had high school or technical school education, 39% were college educated, and 30% had a graduate degree. Thirteen percent of survey participants had a total household income of less than $25,000, 21% between $25,000 and $50,000, 32% between $50,000 and $100,000, and 20% over $100,000 per year. In terms of ethnicity, 95% were Caucasian, 4% were Asian, and 1% were Hispanic. Compared with the demographic characteristics of Missouri, the following population segments were overrepresented: people between 46 and 55 years old (53% for our sample vs. 26% for the Missouri population), people with more than a high school education (69% vs. 31%), the Caucasian population (95% vs. 83% in Missouri), and people representing households with more than $100,000 annual income (20% vs. 13%), while people with an annual income under $50,000 were underrepresented (34% in our sample compared with 57% in Missouri).
Thirty-four percent of respondents have never eaten a chestnut, 43% consumed them once per year or less, and 23% consumed chestnuts two to three times per year or more. Fifty-two percent of respondents were not at all familiar with roasting chestnuts and 51% had never bought chestnuts before the event. Compared with data collected during previous annual Missouri Chestnut Roasts festivals in 2003, 2004, and 2006, we noted a continuous increase in the frequency of consumption from year to year (Cernusca et al., 2008).
The estimated attribute partworth coefficients of the conditional logit models, along with P values and odds ratios, are presented in Table 3. Odds ratios were obtained by exponentiating the β coefficient and can be interpreted as changes in the odds of a consumer choosing a product with a given category over the base level. An odds ratio of 1, then, indicates that there is no difference in consumer preference between that attribute and the base level, while an odds ratio of 5 suggests that consumers like a given attribute five times better than the base level. Table 3 also presents the results for the models that allow for market segmentation allowing preferences to differ between men and women, and urban and rural settings. Additional models were tested, but no statistical significant differences between previous chestnut consumers, age, income, and education groups were detected. These results of the latter are omitted, as they provided little additional information.
Conditional logit partworth estimates of chestnut product attributes for fixed and mixed-effects conditional models derived from a conjoint analysis of consumer preferences for chestnuts performed in 2007 during the Missouri Chestnut Roast festival.
Results of the fixed-effects model show that all product attributes were statistically significant at the α = 0.10 level and all but the large-size characteristic were significant at α = 0.05. Odds ratios indicate that consumers had a preference for medium-sized chestnuts that is 1.89 times higher than the small-size base level. Large-size chestnuts were preferred over small ones and the effect was about the same compared with medium-size chestnuts (1.85). This finding translates into recommendations for chestnut producers to plant proven chestnut cultivars instead of seedlings to obtain medium to large chestnuts, consistent quality, and more consistent yields (Hunt et al., 2006). Consumers also were 1.31 times more likely to select chestnuts that are organically certified compared with conventionally produced chestnuts. The most salient product attribute was the label of origin. Information collected indicates that consumers were 10 times more likely to choose chestnuts grown in Missouri, and five times more likely to select chestnuts grown in the United States compared with imported nuts. These results reflect the support for locally grown products reported by Darby et al. (2006) in Ohio, Gallons et al. (1997) in Delaware, Jekanowski et al. (2000) in Indiana, and Patterson et al. (1999) in Arizona. In a study conducted in New England, Giraud et al. (2005) found that favorable attitudes toward local goods are positively correlated with the probability of purchasing the local good.
As would be expected, the model suggests that as price increases, the likelihood of purchasing chestnuts decreases. However, the relative influence of price on consumer choice is much smaller than that of the other attributes (0.83).
Statistically significant differences were obtained when allowing for variability in respondent-specific preferences for two sociodemographic characteristics: gender and place of residence. When consumers were classified according to their gender, there was evidence that women are more likely to select chestnuts certified to organic standards. These results are in accordance with findings of other studies (Harris et al., 2000). Harris explained these findings by the fact that women tend to be the primary food shoppers of a household and may be more aware of food issues. The results of the mixed-effects conditional model controlled for place of residence (urban and rural) diverge from those reported by Brown (2003) for the state of Missouri who found that household location influenced preference for local products and did not influence the price a buyer was willing to pay. In our case, the price attribute was the only statistically significant variable in the model, suggesting that urban residents are more price sensitive than rural residents.
Building on a market-based approach, efforts to encourage the expansion of chestnuts production in the United States should be focused on the use of chestnut cultivars that produce a medium- to large-size chestnut. Organic certification would help to increase the likelihood of purchase. However, “locally grown” is the most powerful attribute. Results on price effects suggest that prices of $5 or $7 per pound are attainable in the market for products that meet the above characteristics. In fact, retail chestnut prices above $5 per pound are very common (Gold et al., 2006). Above all, marketing should emphasize “locally grown,” as this attribute was shown to be the most important to survey respondents.
Conclusion
Results of this and other studies (Gold et al., 2005; Cernusca et al., 2008) indicate that frequency of chestnut consumption is increasing. Quality, locally grown, and nutrition-diet-health were consistently perceived as the most important attributes influencing chestnut purchase decisions. Three-year findings (2003, 2004, 2006) confirmed initial results indicating that consumers who participate in the Missouri Chestnut Roast festival value product quality, local production, and nutritional value, constantly ranking price as a relatively lower priority attribute. These results support promotion focused on local production and quality (quality chestnuts are firm, heavy for their size, and free of external defects) to receive a premium price.
The exploratory CA suggests that medium and large chestnuts were preferred over small ones by consumers in Missouri. There were small differences between medium and large sizes and a slightly higher preference for medium- compared with large-size chestnuts. This is a finding that actually matches consumer preferences with agronomic reality. Chinese chestnut (Castanea mollissima) cultivars that reach a medium size are already proven producers and are recommended to be grown in Missouri and the midwestern United States (Hunt et al., 2006).
Chestnuts that originate in Missouri and the United States would be preferred over imported ones and there was a preference for organic compared with conventional production. Higher prices had a relatively small negative influence on consumer preference for chestnuts. Looking at all attributes, the highest importance in the purchase decision is the label of origin (produced in Missouri followed by produced in the United States). The next most important attributes are chestnut size (medium followed by large) and organic production. The least important attribute is price. These findings suggest that chestnuts that are produced locally and carry organic certification could command a market premium compared with imported/nonorganic certified chestnuts. In particular, there will be a considerable competitive advantage for local growers, as consumers give preference to domestically and locally grown chestnuts.
Literature cited
Aguilar, F.X. & Vlosky, R.P. 2007 Consumer willingness to pay price premiums for environmentally certified wood products in the U.S For. Policy Econ. 9 1100 1112
Anagnostakis, S.L. 1987 Chestnut blight: The classical problem of an introduced pathogen Mycologia 79 1 23 37
Arnold, S.J., Roth, V. & Tigert, D.J. 1981 Conditional logit versus MDA in the prediction of store choice Adv. Consum. Res. 8 1 665 670
Behe, B.K. 2006 Conjoint analysis reveals consumers prefer long, thin asparagus spears HortScience 41 5 1259 1262
Behe, B.K., Walden, R.M., Duck, M.W., Cregg, B.M., Kelley, K.M. & Lineberger, R.D. 2005 Consumer preferences for tabletop Christmas trees HortScience 40 2 409 412
Brown, C. 2003 Consumers' preferences for locally produced food: A study in southeast Missouri Amer. J. Altern. Agr. 18 4 213 224
Campbell, B.L., Nelson, R.G., Ebel, R.C. & Dozier, W.A. 2006 Mandarin attributes preferred by consumers in grocery stores HortScience 41 3 664 670
Carlman, S.F. 2007 Cherished chestnuts. No longer endangered 7 Oct. 2008 <http://www.pantagraph.com/articles/2007/12/10/news/doc475c7f54c1ca5818953424.txt>.
Carson, R.T., Louviere, J., Anderson, D., Arabie, P., Bunch, D., Hensher, D., Johnson, R., Kuhfeld, W., Steinberg, D., Swait, J., Timmermans, H. & Wiley, J. 1994 Experimental analysis of choice Mktg. Lett. 5 4 351 367
Cernusca, M.M., Gold, M.A. & Godsey, L.D. 2008 Influencing consumer awareness through the Missouri Chestnut Roast J. Ext. (Forthcoming: December issue).
Chestnuts Australia Inc 2008 Nutritious, delicious Australian chestnuts 20 Feb. 2008 <http://www.chestnutsaustralia.com.au/images/pdfdocs/nutritional%20qualities%20of%20chestnuts-updated%2016-2.pdf>.
Darby, K., Batte, M.T., Ernst, S. & Roe, B. 2006 Willingness to pay for locally produced foods: A customer intercept study of direct market and grocery store shoppers 20 Feb. 2008 <http://aede.osu.edu/programs/VanBuren/pdf/AAEA%20Selected%20Paper-156532.pdf>.
Dillman, D.A. 2000 Mail and internet surveys. The tailored survey method Wiley New York
Erturk, U., Mert, C. & Soylu, A. 2006 Chemical composition of fruits of some important chestnut cultivars Brazilian Arch. Biol. Technol. 49 2 183 188
Frank, C.A., Nelson, R.G., Simonne, E.H., Behe, B.K. & Simonne, A.H. 2001 Consumer preferences for color, price, and vitamin C content of bell peppers HortScience 36 4 785 800
Gallons, J., Toensmeyer, U.C., Bacon, J.R. & German, C.L. 1997 An analysis of consumer characteristics concerning direct marketing of fresh produce in Delaware: A case study J. Food Distribution Res. 28 98 106
Giraud, K.L., Bond, C.A. & Bond, J.J. 2005 Consumer preferences for locally made specialty food products across northern New England Agr. Resource Econ. Rev. 34 2 204 216
Gold, M.A., Godsey, L.D. & Josiah, S.J. 2004a Markets and marketing strategy for agroforestry specialty products in North America Agrofor. Syst. 61 371 382
Gold, M.A., Cernusca, M.M. & Godsey, L.D. 2004b Consumer preferences for chestnuts, eastern black walnuts, and pecans HortTechnology 14 4 583 589
Gold, M.A., Cernusca, M.M. & Godsey, L.D. 2005 Update on consumers' preferences for chestnuts HortTechnology 15 4 904 906
Gold, M.A., Cernusca, M.M. & Godsey, L.D. 2006 Competitive market analysis: Chestnut producers HortTechnology 16 2 360 369
Green, P.E. & Srinivasan, V. 1978 Conjoint analysis in consumer research: Issues and outlook J. Consum. Res. 5 103 113
Green, P.E. & Wind, Y. 1975 New way to measure consumer' judgments Harv. Bus. Rev. 53 4 107 117
Hair, J.F., Anderson, R.E., Tathan, R.L. & Black, W.C. 1998 Multivariate data analysis 3rd ed Macmillan New York
Hanemann, W.M. 1984 Discrete/continuous models of consumer demand Econometrica 52 3 541 561
Harris, B., Burress, D. & Eicher, S. 2000 Demands for local and organic produce: A brief review of the literature 17 Feb. 2008 <http://www.ipsr.ku.edu/resrep/pdf/m254A.pdf>.
Harrison, R.W. & Sambidi, P. 2004 A conjoint analysis of the U.S. broiler complex location decision J. Agr. Appl. Econ. 36 3 639 655
Hoffman, S.D. & Duncan, G.J. 1988 Multinomial and conditional logit discrete-choice models in demography Demography 25 3 415 427
Hunt, K., Gold, M.A., Reid, W. & Warmund, M. 2006 Growing chinese chestnuts in Missouri 8 Feb. 2008 <http://www.centerforagroforestry.org/pubs/chestnut.pdf>.
Jekanowski, M.D., Williams D.R. II & Schiek, W.A. 2000 Consumers' willingness to purchase locally produced agricultural products: An analysis of an Indiana survey Agr. Resource Econ. Rev. 29 8 43 53
Kelley, K.M. & Behe, B.K. 2002 Chef's perceptions and uses of colossal chestnuts HortTechnology 12 1 172
Long, J.S. & Freese, J. 2006 Regression models for categorical dependent variables using Stata Stata Press College Station, TX
McFadden, D.L. 1974 Conditional logit analysis of qualitative choice behavior 105 142 Zarembka P. Frontier in econometrics Academic Press New York
McFadden, D.L. 1986 The choice theory approach to market research Mktg. Sci. 5 4 275 297
Nelson, R.G., Jolly, C.M., Hinds, M.J., Donis, Y. & Prophete, E. 2005 Conjoint analysis of consumer preferences for roasted peanut products in Haiti Int. J. Consum. Stud. 29 3 208 215
Patterson, P.M., Olofsson, H., Richards, T.J. & Sass, S. 1999 An empirical analysis of state agricultural product promotions: A case study of “Arizona Grown.” Agribusiness 15 179 196
Punj, G.N. & Staelin, R. 1978 The choice process for graduate business schools J. Mktg. Res. 11 588 598
Smith, B., Behe, B.K., Fulbright, D.F., Walden, R.M. & Kelley, K.M. 2002 Marketing chestnuts to investigate chefs and consumer interests and develop product quality criteria. A final report on the research conducted for Federal State Mktg. Improvement Program Midwest Nut Producers Council and Michigan State Univ 7 Oct. 2008 <http://www.ams.usda.gov/AMSv1.0/getfile?dDocName=STELPRD3247839&acct=gpfsmip>.
University of Missouri Center for Agroforestry 2006 Nutrition and your health. Why chestnuts? 8 Feb. 2008 <http://www.centerforagroforestry.org/pubs/whychestnuts.pdf>.
U.S. Department of Agriculture 2007 U.S. trade imports 8 Feb. 2008 <http://www.fas.usda.gov/ustrade/USTImFatus.asp?QI>.
Wahl, T. 2002 Southeast Iowa nut growers cooperative. Chestnut market opportunities: Assessing upscale restaurant interest in value added chestnut products 8 Feb. 2008 <http://fpc.unl.edu/Reports/chestnut%20Market%20Opportunities.pdf>.