Consumers’ evaluation and intention to buy traditional seafood: The role of vintage, uniqueness, nostalgia and involvement in luxury

This study extends the understanding of the relationship between consumer motivation, involvement and evaluation of traditional food products (TFP). One important research issue was to explore whether consumers' perceived uniqueness could be one important quality or value that theoretically and empirically could differentiate TFP from ordinary, usual and "everyday" traditional food. A central location test of two different vintages (maturation times) of salt-cured clipfish in combination with a survey to assess individual motivational constructs was used to test the conceptual model. Our expectations were confirmed. Perceived uniqueness influences future intention to consume and is motivated by involvement in TFP. The theoretical arguments of the relevance of the perceived uniqueness of traditional food was also supported by the positive relationship between general involvement in luxury (prestige) and involvement in TFP. In addition, this study found a significant positive relationship between product-specific nostalgia and involvement in TFP. To our knowledge, no previous study we are aware has discussed, designed or tested these theoretical relationships. Our findings have promising implications for the seafood industry in order to achieve unique product value and increase consumers' willingness to pay, by promoting uniqueness, prestige and nostalgic product features of vintage salt-cured clipfish.


Introduction
Research on what traditional food is, how it is evaluated by consumers and what explains consumers' preferences, attitudes and consumption of traditional food products (TFP) is emerging (Kristbergsson & Oliveira, 2016). Traditional foods are associated with products that are consumed frequently as a part of daily life or habit, as well as on a seasonal basis on specific occasions like Christmas, Easter or even very infrequently in connection with travelling experiences. TFP not only has to contain traditional ingredients, but also must be processed in a traditional way, according to traditional recipes (Verbeke, Guerrero, Almi, Vanhonacker, & Hersleth, 2016). Like most foods, TFP has its roots in a locality, region, country or geographic area (Guerrero, Claret, Verbeke, & Enderli, 2010). Thus, defining TFP is difficult because of the many different and conflicting attributes or features associated with the construct.
One salient feature of TFP is its associations with the past, and its transmission from one generation to another, the transfer of the knowhow, culinary arts, food preferences and emotions across generations constitutes gastronomic heritage. Thus, "nostalgia" -as an individual's feelings and association with the past in different behavioural contexts (Sedikides, Wildschut, Arndt, & Routledge, 2008) -should be relevant for our understanding of TFP (Renko & Bucar, 2014). Another interesting finding from Guerrero et al. (2010) is that consumers' associations with TFP are related to specific events and specific gastronomic heritage in highly prestige surroundings or contexts (e.g. specific collaborations at an upscale restaurant). Thus, TFP could be associated with uniqueness, exclusivity, luxury or prestige (Jaeger et al., 2017;Vigneron & Johnson, 2004). From previous studies (Almli, Verbeke, Vanhonacker, Naes, & Hersleth, 2011;Pieniak, Verbeke, Vanhonacher, Guerrero, & Hersleth, 2009), we can infer that attitudes and familiarity are the most salient antecedents toward traditional food consumption, and that convenience in some countries is a barrier for consumption. Several other proposed antecedents (e.g. price, health attributes, ethical concern, sensory appeals) are not significantly related either to attitudes or to consumption of traditional products. Contradictory and non-significant proposed findings encourage research on motivational forces (antecedents) for preferences and attitudes towards, as well as engagement in and consumption of, TFP.
In order to overcome the problem with differences in what individuals can associate with their evaluation of "traditional products", Fernandez-Ferrin, Turrientes, Bande, Artaraz-Minon, and Galan-Ladero (2018) used four food brands corresponding to three product categories (cheese, wine and dry-cured ham) as indicators for traditional and local products. However, no physical products were used in their survey. This study will extend their research design, so as to explore consumers' evaluations of TFP by integrating survey questionnaires to assess individual differences (e.g. nostalgia and luxury values) with a central location test of perceived quality, taste and visual attributes of two different traditional food products. In summary, the purpose of this study is threefold. Our first contribution to the existing literature is to test the general and attributespecific evaluation of a traditional seafood product with two different degrees of maturation time of salt-cured clipfishbachalao. A second, but related, contribution is to investigate whether perceived uniqueness as a general product image is related to consumer intention to buy (consequence) and general involvement in TFP (antecedent). We propose that perceived uniqueness can be an important feature that theoretically separates TFP from the ordinary, usual and "everyday" traditional meals. Third, this study extends previous studies (Almli et al., 2011;Pieniak et al., 2009), and it follows up (e.g. Fernandez-Ferrin et al., 2018;Wang, De Steur, Gellynck, & Verbeke, 2015) by exploring whether consumers' nostalgia and luxury or prestige values motivate their involvement in TFP. Combining survey constructs of individuals motivational forces (traits, values and involvement) with quality and taste-evaluations of actual meals can produce more scientifically sound and transferable results for industrial implementations. Results from this study can be used to validate the potential for product differentiations based on the length of the curing, aging or maturing process. The duration of the maturing process increases consumers' preferences and willingness to pay for meat, cheese and wine products (Cilla, Martínez, Beltrán, & Roncalés, 2006) -and hopefully for some seafood products, as well.

Perceived uniqueness
According to Guerrero and colleagues (2010) and Almli et al. (2011), TFP need not only to be tasteful, but it must also have a "distinct" taste. We noted with interest that out of 15 product characteristics tested, Almli et al. (2011) found that only general "high" quality and "special" taste were the attributes that were consistently and significantly associated with TFP across six countries (see Table 4, p 135). "Specificity" -a form of uniquenessis also included several times in the formal definition of TFP (Vanhonacker et al., 2010, p. 454). Furthermore, some consumers feel that TFP are associated with "specific situations" (e.g. celebrations or seasons) made in a "specific way". Product uniqueness is approached from different perspectives and need not always induce associations toward positive consumer values and commercial success (Snyder & Fromkin, 1980). In the area of food uniqueness, Jaeger et al. (2017) found that uniqueness of chocolate was closely associated with "novel" and "unusual". They conclude that the meaning associated withor definition offood product uniqueness is "the quality of being particularly remarkable, special, or unusual" (p. 70). Consumers often express some need for uniqueness when they are searching for something that is difficult to obtain (Vigneron & Johnson, 2004), and perceived product uniqueness is positively related to attitudes, intention and choice of luxury or prestige brands (Chen & Peng, 2014;Ko, Costello, & Taylor, 2017;Miller & Mills, 2012;Shukla, 2012). The complexity of food uniqueness and its associations with luxury and elegance is also confirmed by Favalli, Skov, and Byrne (2013).
Thus, the first research issue in this study investigates whether consumers perceive traditional food as unique, special and unusual (H1a), as well as hypothesize that perceived product uniqueness is positively related (H1b) to their intention to consume after performing a quality and taste evaluation of the two different seafood products. "Intention", in this study, is defined as the consumer's motivation to buy and consume the products in the future (Ajzen, 1991).

Salient quality and sensory attributes of vintage food
One production method to influence consumer quality and sensory evaluation, expectations and value is through vintage, or ageing or maturation time when its meaning is transferred to a food context. The construct of vintage has its origin in wine, especially wine of high quality, characterizing the year and location in which a particular wine was produced (Niemeyer, 2015). Over the years, the construct has been used to describe furniture, cars, clothing and other products that produce a connection with the past and nostalgia (Sarial-Abi, Vohs, Hamilton, & Ulqinaku, 2017). For example, Amatulli, Pino, De Angelis, and Cascio (2018) propose a positive relationship between perceived uniqueness and nostalgia for luxury vintage products in their qualitative study. Vintage items also symbolize uniqueness and authenticity.
In the area of food, a product can be perceived as being vintage by using production or promotion processes that reminds the consumers of the past. Examples of this include the long aging or maturation time for certain dry-cured hams, meats and cheese products. Vintage attributes are also used as indicators of high quality, high price and exclusivity among food consumers (Hersleth, Lengard, Verbeke, Guerrero, & Naes, 2011). The most exclusive Spanish dry-cured ham, Grand Reserva, is cured for more than 38 months.
Hersleth, Monteleone, Segtnan, and Naes (2015) identified 25 descriptors or attributes related to appearance, odour/smell, flavour/taste and texture/mouth feel. With inspiration from marketing theories about vintage, authentic or nostalgic products (e.g. Sarial-Abi et al., 2017), as well as sensory studies of vintage and mature, aged or cured food products (Andrade et al., 2017;Hersleth et al., 2011Hersleth et al., , 2015, this study uses salt-cured and dried clipfish of cod as the object for evaluation (Oliveira, Nunes, Vaz-Pires, & Costa, 2016). Salted and dried cod has a long shelf life and a unique sensory taste, odour, colour and texture. Before consumption, the product is subject to desalting, which involve soaking the fish in water. Thus, this study combine the relevant general attributes of TFP (e.g. quality, taste, natural, appearance, health and nutrition) with specific sensory studies of dry-cured food (texture, smell, colour and saltiness) (Almli et al., 2011;Fernandez-Ferrin et al., 2018;Hersleth et al., 2011;Pieniak et al., 2009). For example, the white/yellow colour and saltiness are included as specifically relevant for salt-cured clipfish of cod. Thus, our second research issue in this study is to test if "vintage" (maturing time) cause quality, taste and visual sensory differences (H2) in the two seafood products.

Antecedents of consumer involvement in traditional food products
Individual eating behaviour, food consumption and choice are influenced by multiple and complex motivational forces and traits (Renner, Sproesser, Strohbach, & Schupp, 2012;Steptoe, A., Pollard, T. M., & Wardle, J., 1995;Symmank, C., Mai, R., Hoffmann, S., Stok, M., Renner, B., Lien, N., & Rohm, 2017). The third main aim of this study is to investigate whether two of the most relevant motivational forces, health and convenience, influence consumer involvement in TFP (Pieniak et al., 2009). In addition, we extend previous research by exploring whether luxury and perceived uniqueness could motivate consumers involvement in TFP.
Some studies use the term "general image" (e.g. Almli et al., 2011) in order to make a clear distinction between the general evaluation and the evaluation or perception of sensorial and other specific attributes or characteristics. For example, Pieniak et al. (2009) use attitude (happy, exiting and delightful), and Almli et al. (2011) use "personal opinion/feelings" (negative/positive) to assess consumer's global evaluations or general images of TFP. This study uses involvement in TFP as evaluation of "the general image of the category" in the survey section. Consumer involvement is frequently defined as a consumer's personal interests, importance, concern and significance attached to an attitude object (Zaichkowsky, 1994). Involvement as a theoretical motivational construct is closely related to engagement (Brodie, Hollebeek, Juric, & Ilic, 2011) and attitude (Mittal, 1995).
In Pieniak et al. (2009), familiarity was the only motivational construct (out of eight) that was significant associated with attitudes toward TFP across all countries. For example, convenience orientation was negatively related to both general attitudes toward TFP and traditional food consumption (TFC) in three out of six countries. Health-related motives (importance of natural content, healthiness and weight control), sensory appeal, ethical concern and price were not significantly related to attitude or consumption of TFP in most countries. Fernandez-Ferrin et al. (2018) found that ethnocentrism is sometimes, but not always, related to the actual purchase of local-regional-traditional food products. Divergent and unexpected results across studies, countries and products allows for inclusion of other motivational constructs, as well as replication of previous studies.
Thus, this study will include three of the most relevant motives for explaining or predicting attitudes and consumption of TFP (Pieniak et al., 2009): familiarity, health and convenience. In addition, this study includes involvement in luxury as a new motive for involvement in TFP. However, our theoretical approach to familiarity, health and convenience differs from the TRUFOOD study. The first motive, familiarity, was in Pieniak et al. (2009) defined as a combination of preference or involvement in consuming familiar ("what I usually eat") and nostalgic (traditional) food ("I ate it when I was a child").
Nostalgia involves a contrast between the present and the past and has been studied in a variety of theoretical areas and contexts (Routledge, 2015). Nostalgia can be "real", "affective/emotional", "sentimental", "simulated/vicarious", "personal", "collective", "historical" or "cultural". Nostalgia literature within food anthropology (Holtzman, 2006) and food psychology (e.g. Vignolles & Pichon, 2014;Wang, Keh, & Chao, 2018) is relatively rich. Less than a handful studies we are aware of are dealing with nostalgia-related specificity toward traditional food. With high relevance for our study, Renko and Bucar (2014) used both qualitative (focus groups) and quantitative (survey) methods to approach nostalgia among chefs and consumers in Croatia. They found that consumers evaluated traditional food not only positively in terms of health, taste and nutritional value, but also associate it with their fond childhood memories, collaborations and pleasures of family togetherness. This study uses a general approach of nostalgia as "a sentimental longing for one's past" (Sedikides, Wildsehut, Arndt, & Routledge, 2008), but framed towards a specific context of consuming traditional food (e.g. "eating meals of clipfish reminds me of the past"). We expect that nostalgia is positively related (H3) to involvement in traditional food.
Health, was in Pieniak et al. (2009) associated with several health-related constructs such as health, natural content and weight control. This study uses a more general health concern approach based on the consumer involvement construct (Zaichkowsky, 1994) and adapted to health involvement in food consumption behaviour literature (Olsen, 2003;Pieniak, Verbeke, Olsen, Hansen, & Brunsø, 2010). Traditional food is often perceived as natural and healthy (Almli et al., 2011). However, the relationship between health attributes and attitudes toward traditional food was insignificant in Pieniak et al. (2009), except for in France, where health was both positively (natural content) and negatively (healthiness) related to traditional food consumption. Dispute the contradictory empirical findings, this study follows the expectation discussed above and assumes the finding of a positive relationship (H4) between health involvement and involvement in TFP.
The third motive we replicate and extend from Pieniak et al. (2009) is perceived product convenience: the degree to which individuals perceive TFP as easy and fast to plan, buy and prepare. This study defines convenience as a more general individual trait or value ("food convenience orientation"), meaning the degree to which individuals evaluate themselves as preferring meals that are fast and easy to plan, buy and prepare (Candel, 2001;Olsen, Scholderer, Brunso, & Verbeke, 2007). Pieniak et al. (2009) found perceived product convenience to be negatively associated with general attitudes towards TFP in three out of six countries (Belgium, France and Italy). Based on those results, and the fact that traditional food meals are often reserved for special occasions (Easter and Christmas), seasons and highly-motivated situations, this study expects to find a negative relationship (H5) between convenience orientation and involvement in TFP.
Finally, TFP is associated with high quality, expensiveness, handcrafting and perceived uniqueness Verbeke, Guerrero, Almli, Vanhonacker, & Hersleth, 2016). As noted above, these product attributes are related with prestige or luxury-seeking consumer values or behaviours (Vigneron & Johnson, 2004). Prestige and luxury are often synonymously used in the marketing literature when trying to define the meaning of luxury (Ko et al., 2017;Miller & Mills, 2012). Most definitions of luxury deal with assessing consumers' perceived value of different attributes with a prestigious brand or product, such as very high price/quality, exclusivity, authentic value, craftsmanship, hand-crafting, history, rarity and uniqueness, among others. This study uses the involvement construct (Zaichkowsky, 1994) as a basis for defining and measuring consumers' general concern and interests in prestige or luxury products (Dubois, Czellar, & Laurent, 2005;Heide & Olsen, 2017;Shukla & Purani, 2012). Traditional food products share some of the same attributes as luxury or prestigious products. Thus, it is reasonable to expect that individuals' interests in prestige or luxury products can be positively related (H6) to their involvement in TFP.
A visual presentation of our integrated conceptual framework is given in Fig. 1. The left side deals with the relationship between the motivational forces based on a survey questionnaire performed as the first part of our central location test. The right side deals with the relationship between perceived quality, taste and visual evaluation of uniqueness and intention to buy based on an evaluation of two traditional seafood products with different maturity times (ordinary and mature).

Products
Two types of salt-cured and dried cod (clipfish) were made for this study. The first type of clipfish was produced according to normal industry standards, that is, it was salt cured for about 6 weeks and then dried. The second type was produced with a more traditional method, and was salt-cured for 6 months and then dried. Before consumption, the products were subject to desalting, which involves soaking the fish in water.

Participants and procedure
Using a nonprobability sampling method, participants were recruited from parents, relatives and players on local sport teams. In order to recruit relevant participants for the study the following criteria were used: participants had eaten fish during the last year and have no allergies to seafood. A convenience sample total of 117 subjects (43% female; 57% male) between 18 and 77 years of age (mean age 48.6) were recruited to participate in the test in Tromsø, Norway. 80% of the participants had attended college or university. The reason for the high number of participants with higher education could be because the test was conducted on a university campus.
In the present study, a central location test was used in order to create a more realistic meal setting (King, Weber, Meiselman, & Lv, 2004). The test was conducted during the dinner period, between 3 and 6 p.m., over 3 consecutive days. Participants were asked not to eat dinner before the test. Up to 15 respondents participated in each session, which took about an hour. No social interaction was allowed. Participants first completed a survey questionnaire with items expressing the latent content of nostalgia, convenience orientation and involvement in health, luxury and TFP. After completing the first part of the questionnaire, the participants were served a meal including clipfish, potatoes, pea purée and bacon and vinaigrette. All consumers were served the same meal twice (within-subject design), with only one difference: the type of clipfish (normal or traditional). The consumers received no information about how the clipfish was produced. The order of the meals was randomized, and the participants performed the evaluation of the first meal during and after eating, but before they received and evaluated their second meal. The evaluation part of the questionnaire was intended to cover perceived uniqueness, quality, sensory attributes and intention to buy such a clipfish in the future.

Measurement and scaling
After each meal, the respondents had to evaluate their experiences on a seven-point semantic bipolar differential scale. Respondents were presented with the sentence: "How would you evaluate your experience with the meal you now have tasted"? To assess perceived uniqueness, three bipolar adjectives were used: "common/ordinary -unique"; "for everyday usespecial"; and "usual -unusual/specific". These items are adapted from the uniqueness dimension of perceived prestige/luxury brands (Vigneron & Johnson, 2004) and perceived food uniqueness (Jaeger et al., 2017). After evaluation of the general meal, the respondents were presented with the following sentence: "If you try to eliminate the accessories, how would you evaluate the following attributes with the main seafoodthe clip-fish?" The attributes presented coved perceived quality (bad-good), naturalness (unnatural-natural), taste (bad-good; neutral-mature; saltiness), texture (bad-good; dryjuiciness), smell (bad-good), appearance (bad-correct; bad-nice colour; yellow-white) and health (unhealthy-healthy; low-high nutritional value).
In the measures of intention to consume, the two different products were framed as: "Assuming that the clipfish you now have tasted in this meal were available in your supermarket, how likely would it be that you would consume this product within the coming: (a) 'one month', (b) 'half a year', and (c) 'one year'". The items were coded by a seven-point probability scale, ranging from "very unlikely (1)" to "very likely", which has been previously used in similar studies (e.g. Olsen, Heide, Dopico, & Toften, 2008).
"Involvement" in this study is defined as consumer's personal interests, importance, concern and significance (Zaichkowsky, 1994). We used the same theoretical framework to assess involvement in traditional food (ITF), involvement in health (IH) and involvement in luxury products (ILP). The three were included on a seven-point Likert scale (totally disagreetotally agree) as: "ITF/ILP … is important for me"; "I am very interested in. ITF/ILP"; and "ITF/ILP. means a lot to me". These items are used to assess food/seafood involvement (Marshall and Bell, 2003:;Olsen, 2001), attitudes toward luxury (Dubois et al., 2005) or overall luxury values (Shukla & Purani, 2012). Health involvement is assessed in the same format, with three items: "It means a lot to me to have good health"; "Good health is important to me"; and "I often think about my health". These items are frequently used to assess health involvement in food psychology (Olsen, 2003;Pieniak et al., 2008).
Nostalgia was assessed by three items adopted from the Evoked Nostalgia Scale (NOST) developed by Pascal, Sprott, and Muehling (2002). A seven-point Likert scale featured (totally disagreetotally agree) the heading statements "Eating clipfish for dinner" and the three following questions: "Reminds me of the past"; "Brings back good memories of the good time from the past"; and "Makes me think about old traditions". Convenience orientation was measured using the following three items on a seven-point Likert-scale: "The less thinking I need to plan, buy, prepare and cook a meal, the better"; "I prefer meals that are easy and quick to plan, buy, prepare and cook"; and "The less physical effort I need to buy, prepare and cook a meal, the better". The scale is based on Candel (2001) and is used in several previous studies to assess food convenience orientation (e.g. Olsen et al., 2007).

Analytical procedure
A paired sample t-test was used to determine whether there was a significant difference in the sensory evaluation of the two types of mature clipfish. Furthermore, the data analysis employed the two-step approach recommended by Anderson and Gerbing (1988). We first conducted a confirmatory factor analysis using AMOS 24. This analysis was used to test how well the measures of the constructs represent our understanding of the nature of the constructs, that is, the validity and reliability of the constructs. The measurement model that we estimated proposed that each measure should reflect the appropriate constructs underpinning our conceptual model. The second stage of the analysis used structural equation modelling to test the proposed model concerning the drivers of preference for local food by maximum likelihood estimation in AMOS 24. A χ 2 /df ratio lower than 2 indicates a good fit (Tabachnick & Fidell, 2001). The analysis included three other indices: the root mean square error of approximation (RMSEA); the incremental fit index (IFI); and the comparative fit index (CFI). Acceptable model fits are indicated by IFI. CFI values exceeding 0.90 and RMSEA values below 0.08 represent a moderate fit, while values lower than 0.05 indicate a good fit (Browne & Cudeck, 1992).

Consumer's evaluation of uniqueness, quality, taste and visual attributes
The results showed no significant difference in how the uniqueness, quality and sensory attributes were evaluated for the two types of mature clipfish (Table 1). Both varieties were evaluated around 4 on a scale from 1 to 7. The quality, taste and visual attributes received a relatively higher score between 5.0 and 5.9, which indicates a favourable impression of the products. Both products were evaluated as healthy. However, the difference in maturation time significantly (p < .05) influenced the consumer's evaluation of five attributes. As expected, the matured clipfish had a more mature taste, a yellow colour and saltiness, as well as a less favourable colour. In addition, it was evaluated as less nutritious compared to the ordinary clipfish. In total, the consumers evaluated the two maturation times (vintage) differently on the same salient attributes without information about the maturation times or the physical difference between the products. Thus, our expectations that "vintage" (mature time) influenced differences in consumer's evaluation of some salient quality, taste and visual attributes in the two clipfish products, are confirmed (H2).

Reliability and validity of the measures
Two initial confirmatory factor analyses (CFA) based on 21 variables were conducted (ordinary clipfish and matured clipfish, respectively). One item measuring health involvement ("I often think about my health") was omitted due to low factor loading (0.45). The final mea-  (see Appendix). First, all composite reliabilities were 0.86 or higher, which indicated a degree of internal consistency among the measures that was far above the recommended level of 0.60 (Bagozzi & Yi, 1988). Cronbach's alphas are also acceptable, ranging from 0.88 for perceived uniqueness of ordinary clipfish to .96 health involvement. Secondly, the average variance extracted ranged from 0.69 to 0.87. Bagozzi and Yi (1988) suggested a target level > 0.50.
The convergent validity of the scales was assessed using three criteria: (1) individual items' lamda coefficients greater than 0.50; (2) a significant t-statistic for each path; and (3) each path loading greater than twice its standard error (Anderson & Gerbing, 1988). All the item-factor loadings are larger than 0.75 and significant, with each path loading being largely greater than twice its standard error.
All correlations are below 0.55 (see Table 2). The discriminant validity of the scales was assessed using a procedure recommended by Bagozzi, Yi, and Philips (1991). Within each subset of measures, pairs of constructs were examined in a series of two-factor confirmatory models. A chi-square difference test was conducted. The results suggested that for all the pairs of constructs, the two-factor solution was better (p < .001) than the single-factor solution. The discriminant validity of the constructs was also tested using the approach suggested by Fornell and Larcker (1981). No square of correlation among the constructs exceeds the square root of the average variance extracted (AVE) from the specific construct. In summary, the measures of the proposed constructs achieve satisfactory reliability (i.e. are not flawed by random errors) and convergent and discriminant validity (i.e. are not flawed by systematic errors, indicating that we are measuring separate constructs with no significant overlap).

Structural analysis and model testing
An SEM methodology was employed to test the latent variable models. The theoretical models fit the data well (χ 2 = 219.14 with df = 161 [χ 2 /df = 1.36], RMSEA = 0.056, IFI = 0.97, CFI = 0.97). Standardised estimates for the various model paths are provided in Table 3. The relationships between nostalgia (H3), involvement in luxury (H6) and involvement in TFP are supported by the data. The relationships between health involvement (H4), convenience orientation (H5) and involvement in TFP are not supported by the data. We expected positive relations between involvement in TFP and the sensory evaluations of the uniqueness (H1a) of the two products. This relation is not supported by the data for the ordinary clipfish; interestingly, the data did support this relation for the mature clipfish product. The relationships between uniqueness and intention to consume (H1b) are supported by the data for both ordinary and mature clipfish. 24.9% of the variance in involvement in TFP is explained for both ordinary and mature clipfish. 4% of the variance is explained in perceived uniqueness of the mature clipfish product. 15% of the variance in intention to consume mature clipfish is explained, and 13.6% of the variance in intention to consume ordinary clipfish is explained.
Alternative models in which the antecedent constructs directly influenced the uniqueness of ordinary and matured clipfish (i.e. not mediated by involvement in TFP) were performed in order to test the robustness of the proposed model. The alternative models showed poor fit and modification indices (Jöreskog & Sörbom, 1996). This provides support for the proposed theoretical models.

The role of vintage and perceived uniqueness evaluation on intention to consume
Research seeking to identify the core feature of traditional food products (TFP) and what motivate consumers to buy and consume TFP is an emerging research issue in food psychology (Almi et al., 2011;Guerrero et al., 2010;Pieniak et al., 2009;Verhonacker et al., 2010). This study contributes to this research by proposing, testing and exploring whether consumers' perceived uniqueness could be one primary quality or value feature that separates TFP from non-traditional Table 1 Evaluation and comparison of perceived uniqueness, quality and sensory attributes of mature and ordinary clipfish: A paired sample t-test. food or ordinary, usual or everyday food. If consumers perceive traditional food as a continuum from ordinary to unique, the industry could use unique product attributes to develop and differentiate TFP in order to achieve premium brand value and increase willingness to pay. It could also satisfy consumers' needs for uniqueness, luxury and variation in food consumption. A positive relationship between perceived uniqueness and consumer's intention to consume is confirmed in this study (H1b). Such a relationship has been confirmed in studies of perceived luxury values (e.g. Shukla, 2012), but as far as we know, it has not been empirically tested in previous studies in a food context. This study tests two different products of salt-cured traditional clipfish in a central location sensory test. Norwegian consumers evaluated the meal based on the most matured product as different on some expected attributes that physically change during the 6-month saltcuring period without any product information. The most matured product was perceived as more yellow, and of a less good colour than the traditional product (H2). This confirms results of previous studies, which show that consumers prefer white or light-coloured cod products (Esaiassen et al., 2004;Sveinsdóttir et al., 2009). As expected, the taste of most matured product was perceived as being most traditional. Saltiness is not a preferred quality among many consumers. This may be the reason why the most mature product was perceived as less nutritious. However, no difference was found in the evaluation of the general taste (good/bad), indicating that overall, both products tasted equally good.
Both products were evaluated to be similar on perceived general uniqueness and general quality. Thus, this study confirms previous studies of dry-cured ham, for example (Andrade et al., 2017;Hersleth et al., 2011Hersleth et al., , 2015, which show that the consumer is capable of identifying intrinsic product attributes (e.g. prolonged aging time and saltiness) in blind sensory studies of mature products. This is relevant information for the clipfish industry, since perceived differences in intrinsic sensory attributes are important to achieve trust and develop advanced branding, labelling and communication of vintage or nostalgic values, luxury or uniqueness. Inclusion of external product attributes (price, brand, packaging, stories) are even more fundamental than small sensory differences to achieve competitive product brand-advantage and willingness to pay in a commercial situation (Piqueras-Fiszman & Spence, 2015). This is also confirmed in the context of dry-cured ham (Hersleth et al., 2015), but it remains an issue for future studies of other seafood products (e.g. dried cod, stockfish and sardines in olive oil).
Traditional food is not only a perception or sensory issue, but is also based on consumers' expectations, attitudes and interests. This study perform an empirical test of the relationship between interest or involvement in TFP and a taste and visual evaluation of "real" TFP (H1a). An interesting finding is that this relationship was only significant for "aged clipfish" and not for "ordinary clipfish". This indicates that perceived uniqueness has a potential for making a specific TFP more traditional. Theoretically, we suggest that perceived uniqueness can be one feature that contributes to separating "unique" traditional food from "usual" traditional food. Managerially, it contributes to the potential for the industry to profile salt-cured traditional clipfish within the traditional food consumer segments. We will expect that the relationship between involvement in traditional food and perceived uniqueness will be more extreme if the products are presented with different forms of information, labelling, packaging, brand-names and other extrinsic cues. Those expectations are based on the same theoretical and empirical findings as noted in the previous paragraph (Piqueras-Fiszman, 2015), but it remains an issue for future research.

Perceived nostalgia and luxury value affect involvement in TFP
Theoretically, if traditional food shares the same motivations as ordinary food, it is hard to confirm TFP as a theoretical construct. This study confirms some previous studies (Almi et al., 2011;Pieniak et al., 2009) that neither health involvement (H4) nor convenience (H5) are significant antecedents toward involvement/attitude in TFP. It is noteworthy that our study differs from that of, for example, Pieniak et al. (2009) in the way that our study uses involvement in traditional food, not general attitude, and food convenience orientation instead of perceived product convenience. Even though the constructs are not similar, our results support the TRUEFOOD studies (e.g. Pieniak et al., 2009).
Traditional food (e.g. "I ate it when I was a child"; Pientak et al., 2009) share the meaning of feeling nostalgia for the past (Sedikides et al., 2008). Based on an explorative qualitative analysis, Vignolles and Pichon (2014) suggest that nostalgia is instead a negative or ambivalent emotion, but one that can also be positively associated with nostalgic food consumption. This study confirms the positive relationship between product-specific nostalgia and involvement in TFP (H3). Wang et al. (2018) found a positive relationship between nostalgia and preferences for indulgent food, and results from Renko and Bucar (2014)   suggest a relationship between traditional food and positive past experiences. Thus, the seafood industry could benefit from positioning TFP products with nostalgia elements: associations with emotional and sentimental, personal, collective, historical or cultural attributes, in addition to "real" mature, sensory and quality attributes. For example, the Norwegian cod-fish industry achieved a substantially higher price after promoting "skrei" with nostalgia (posturing a thousand-year history of cod as a wanderer from the north) in the European seafood market. Perceived uniqueness (Favalli et al., 2013;Jaeger et al., 2017) shares the same meaning as prestige and luxury (Miller & Mills, 2012;Vigneron & Johnson, 1999. This studyconfirms a positive relationship between involvement in prestige or luxury (overall luxury value/attitude) and involvement in TFP (H6). Guerrero and colleagues (2010) found that consumers' associations with TFP are related to highly profiled surrounding and specific gastronomic heritage. Thus, both the food and the luxury industry can use knowledge from premium and luxury marketing (Wiedermann & Hennigs, 2013) to increase expectations, values and willingness to pay for high-profile TFP. Several TFP are of high natural quality, seasonal (scarcity), handmade, intended for special social occasions, unique and with an authentic history.

Limitations and future research
Our study focuses on the possible effects of maturation time (vintage) on the evaluation of traditional food products (salt cured-clipfish of cod) in a central location test. Future experiments could use different products and different forms of methods and experiments such as segmentation, sensory lab experiments, or in-home tests. Our survey is not representative to verify different motivational forces for involvement in TFP. However, it confirmed the findings of a low or non-significant relationship between perceived health/convenience and attitude toward TFP in a representative study of European consumers (Pieniak et al., 2009).
Even though our proposed constructs and model achieved satisfactory reliability and validity using structural equation modelling, our sample was on the lower end with respect to size. However, according to Iacobucci (2010): "It is of some comfort that SEM models can perform well, even with small samples (e.g., 50 to 100). The vague, folklore rule of thumb considering requisite sample size, e.g., "n > 200" can be conservative, and is surely simplistic. This argument is also supported by Wolf, Harrington, Clark, and Miller (2013). Nonetheless, future studies should use a larger representative samples and include other motivational values, personality tracts or emotion as antecedents for attitudes, satisfaction, consumption and willingness to pay (Renner et al., 2012;Steptoe, A., Pollard, T.M., & Wardle, J., 1995;Symmank et al., 2017). Of particular interest, we suggest "need for uniqueness or status" (e.g. Cardello et al., 2019); "regional/ethnocentrism" (e.g. Fernandez-Ferrin et al., 2018); "social forces" (norms, connectedness, modelling); and "sustainability" or "culture". Including extrinsic product attributes, particularly information/labels, packaging and brands often outperforms intrinsic attributes as predictors of general product evaluation, intention to buy or food choice (Piqueras-Fiszman & Spence, 2015;Symmank, 2019). Thus, future research should use information associated with handmade performance, seasons, authenticity, history stories, long maturation time or other cues to signal luxury, prestige, uniqueness, nostalgia or vintage. Finally, heterogeneity in the production (e.g., desalting), preparing, cooking and serving processes are problematic in central location tests.

Conclusion
This article extends the understanding of the relationship among consumers' general nostalgic and prestige values, their general interest/ involvement in TFP and their evaluation (quality, uniqueness, taste, appearance, etc.) of vintage (maturation time) of dry-salted clipfish, a product with a history of more than five hundred years. The main results indicate that perceived uniqueness can be one relevant feature that constitutes a theoretical distinction between traditional and nontraditional food. Perceived uniqueness influences future intention to consume and is motivated by involvement in TFP. The theoretical arguments and findings are strengthened by the positive relationship between general involvement in luxury (prestige) and involvement in TFP. Finally, individual sentimental feelings or longings for the past (nostalgia) stimulate engagement and involvement in TFP. To our knowledge, no previous study has discussed, designed or tested these theoretical relationships.
Beyond the theoretical and methodological contributions, our findings have significant implications for the seafood and clipfish industry. The Norwegian clipfish industry has suffered economically because of their commodity-and production-oriented business philosophy. This study opens up possible product differentiations based on unique sensory qualities through the curing or maturation process, as has been done with wine, liqueur, cheese and ham or other quality meats. The industry may consider using luxury/prestige and nostalgic promotional or communicational tools (e.g. branding, stories and packaging) to increase consumers' interests, attitudes, preferences, consumption and willingness to pay for traditional products. This could also be done by developing and marketing vintage, mature or cured varieties to increase differentiations and strengthen the uniqueness, value and prestige of TFP for segments of consumers.

Declaration of competing interest
We wish to conform that there are no known conflicts of interest associated with this publication and there has been no significant support for this work that could have influenced its outcome.