AKADEMIJA
DETECTING CONSUMERS WITH LOW FAMILIARITY IN COUNTRY-OF-ORIGIN RESEARCH
Vesna Kuralt
University of Ljubljana kuralt.vesna@gmail.com
Dr. Anja Mohorko
Oxford University Press anja.mohorko@oup.com
Dr. Mihael Kline
University of Ljubljana mihael.kline@fdv.uni-lj.si
Izvirni znanstveni članek Original Scientific Paper
Abstract: This research aims to detect the attitudes of consumers who are not familiar with the country of origin by studying the image of an unfamiliar country of origin. To the best of our knowledge, there are no studies of country-of-origin effect in the context of low familiarity. In order to detect such consumers, we have added the 'I don't know' option to the Likert scale. This has enabled us to divide the respondents into two groups according to their familiarity with the target country, which is an unfamiliar country of origin. Since this pre-assumes a considerable lack of consumer knowledge, we based our model on their emotions. For this reason, we referred to consumer affinity research when building the model of unfamiliar country of origin effect. An indirect impact of consumer affinity on their willingness to buy was fully mediated by perceived risk in both groups of consumers with high and low familiarity. Hence, we encourage researchers to include perceived risk in their study design.
Keywords: familiarity, unfamiliar country of origin, perceived risk, consumer affinity, consumer knowledge, I don't know
ZAZNAVANJE POTROŠNIKOV S ŠIBKIM POZNAVANJEM DRŽAVE IZVORA
Povzetek: Ogromno študij na področju marketinga za kvantitativno metodo merjenja stališč uporablja Likertovo lestvico, vendar takšna metoda iz vzorca anketirancev izloči tiste, ki imajo omejeno znanje o objektu. Ti na postavljena vprašanja niso sposobni odgovarjati ali pa se ne čutijo dovolj kompetentne za takšne sodbe. Z namenom vključitve teh anketirancev smo Li-kertovi lestvici dodali možnost odgovora ne vem. Tako jim omogočamo sodelovanje v anketi kljub šibkemu poznavanju nekaterih izmed preučevanih latentnih spremenljivk. Navedeno metodo merjenja smo uporabili na primeru študije države
53
AKADEMIJA
izvora in tako poskušali zaznati stališča anketirancev z omejenim znanjem o državi izvora.
Poznavanje je eden od glavnih moderatorjev učinka držav izvora, zato je izbira takšnega konteksta študije smiselna. Praviloma so študije izvedene na primerih zelo prepoznavnih držav izvora, a bi to onemogočalo doseganje osnovnega cilja te študije. Prav zato smo na podlagi postopka prepoznavanja utemeljili tako imenovano neuveljavljeno državo izvora. Ta se od dobro prepoznavne oziroma uveljavljene države izvora razlikuje v količini predznanja anketirancev in njihovih izkušenj z njo. Pri tem je treba poudariti, da neuveljavljena država ni nujno slabo prepoznavna med vsemi skupinami deležnikov. Nekateri izmed njih imajo morda številne izkušnje z njihovimi produkti, medtem ko drugi njenih tržnih znamk sicer ne poznajo, toda so dobro seznanjeni z drugimi vidiki države in zato nimajo težav pri tvorjenju eksplicitnih stališč do neuveljavljene podobe države.
Neuveljavljeno državo izvora torej nekateri deležniki dobro poznajo, drugi pa slabše. Tako tudi neuveljavljeno državo izvora ločimo od neznane države, ki je izbranim skupinam deležnikom popolnoma nepoznana in zato do nje ne morejo oblikovati stališča. Če za razlago teh skupin uporabimo postopek prepoznavanja, izbrane skupine deležnikov v primeru visoke stopnje poznavanja v postopku oblikovanja stališča enostavno prikličejo informacije iz spomina, medtem ko je ta v primeru nizke stopnje poznavanja omejen ali celo nemogoč, čeprav je objekt preučevanja anketirancem razmeroma poznan. V kontekstu naše študije tako ti niso zelo dobro seznanjeni s podobo države izvora oziroma lastnostmi njenih produktov, toda imajo nekaj asociacij na njeno splošno podobo. Ti deležniki se zaradi omejenega znanja pri oblikovanju stališča pogosto zanašajo tudi na čustva kot informacije. Zaradi omejenih informacij nekaterih deležnikov raziskovalni model neuveljavljene države izvora tako osnujemo s čustvenimi vezmi do države.
Naklonjenost do države je eden od novejših kon-struktov na področju preučevanja držav izvora. Pred kratkim so nekateri avtorji definirali različne razsežnosti kognitivne naklonjenosti do države, ki pogojujejo čustveno oziroma afektivno oblikovano stališče naklonjenosti do države. Raziskave kognitivne naklonjenosti v nasprotju z afektiv-no zasnovano študijo naklonjenosti do države ugotavljajo njen posreden vpliv na pripravljenost k nakupu. Vzrok različnih rezultatov posameznih študij je verjetno metoda raziskovanja, saj so anketiranci v primeru ugotovljenega neposrednega vpliva naklonjenosti do države na intenco nakupa
posamično izbrali državo, do katere čutijo največjo naklonjenost. Rezultati omenjene študije kažejo, da v primeru velike naklonjenosti do države ta na pripravljenost k nakupu vpliva neposredno ali pa jo mediira njihova zaznava tveganja.
Zaradi konteksta raziskave neuveljavljene države izvora je vključitev konstrukta zaznave tveganja zelo smiselna. Anketiranci s šibkejšim eksplicitnim znanjem bi se namreč lahko na podlagi svoje zaznave tveganja odločali o nakupu produkta iz neuveljavljene države izvora. Prav zato v model naše študije poleg hipotez v skladu z literaturo o naklonjenosti do države vključimo tudi to hipotezo. V okviru naše študije smo torej preverili neposredni in posredni vpliv naklonjenosti do države na težnjo po nakupu v kontekstu neuveljavljene države izvora.
Ob pomoči odgovorov ne vem smo anketirance razvrstili v skupini visoke in nizke stopnje poznavanja podobe države izvora. Za obe skupini smo testirali vse hipoteze z regresijo in modeliranjem strukturnih enačb. Rezultati pokažejo različna modela učinka (neuveljavljene) države izvora za deležnike z visoko in nizko stopnjo poznavanja, saj je zaradi pomanjkljivega znanja o podobi države izvora ta celoten konstrukt izpuščen iz modela anketirancev z nizko stopnjo poznavanja. V naši študiji tako prikažemo možnosti raziskovanja izbranih skupin deležnikov z omejenim predznanjem in izkušnjami, in sicer na podlagi omogočanja odgovora ne vem. Seveda moramo pri vključevanju takšnih anketirancev preveriti, ali se njihov proces oblikovanja stališč razlikuje od drugih. Ne nazadnje pa rezultati naše študije ponujajo svež pogled na raziskovanje držav izvora in vnovič potrdijo izjemen pomen čustev pri ocenjevanju podobe države izvora.
Ključne besede: prepoznavnost, neuveljavljena država izvora, zaznava tveganja, naklonjenost potrošnikov, znanje potrošnikov, ne vem
54
AKADEMIJA
1. INTRODUCTION
The country-of-origin or product-country image effect is one of the most researched fields in international marketing. This is probably due to the value of understanding consumer reactions to products for managers of international brands. However, some doubts have recently been raised about the relevance of these studies because of the unrealistic conditions under which they are conducted (Usunier, 2006; Samiee, 2011). This criticism is based on conscious or explicit use of the country-of-origin cue, while implicit attitudes are still largely neglected in country--of-origin research, even though some studies already demonstrated the automatic influence of product-country images in consumer evaluations (Liu & Johnson, 2005; Herz & Diamantopou-los, 2013). This can also occur unintentionally. Hence, even consumers with low familiarity are influenced by the country-of-origin cue, whether they are aware of it or not.
The second major criticism of country-of-origin research is also closely related to consumer knowledge. In a highly globalized world, it is increasingly difficult to pinpoint the exact origin of a brand and even harder to assess the origin of a product, which leads to low awareness of brand origin recognition accuracy among consumers (Samiee et al., 2005). This should not be surprising, even though it is supposed to be vital for the country-of-origin effect. It is true that brand recognition—or rather consumer knowledge—seems to be key in this discussion, since knowing the brand origin has a diagnostic role in purchase situations (Samiee, 2011). At any rate, due to excessive information in the modern age, consumer learning is thus more often incidental than intentional and misinformation is highly possible, if not intended. In any case, correct or false, consumers' knowledge affects their evaluations, as long as they are confident in it (Magnusson et al., 2011).
Nevertheless, the focus of country-of-origin research is changing. In the last decade, the interest in studying effects is growing. Research is slowly moving away from traditional country-of-origin studies, while the number of studies on consumer ideologies is increasing. Instead of dealing with cognitive information processing, the focus is on socio-psychological processes stemming from individual mindset and consumer social context (Dmitrovic & Vida, 2010). In this case, the issue of taking the country-of-origin cue into consideration is irrelevant, since the study is interested in how this information is incorporated into their evaluation when perceived as diagnostic.
Everyone agrees that knowledge and, consequently, recognition does play an important role in the evaluation of the country of origin, but its source can be either cognitive or affective information. As mentioned above, the latter are only recently the focus of research, although the primacy of affect over cognition has long been recognized (Zajonc, 1984). Due to the probable lack of cognitions in the case of low familiarity, our research model is based on affects as the source of information for country-of-origin evaluation. Since it would be hard to find a target country that is unfamiliar and induces negative emotions, we have decided to test another poorly researched construct from country-of-origin research in our model. That is consumer affinity, where some mixed results have recently been found in the literature (Oberecker & Diamantopo-ulos, 2011; Nes et al., 2014). Hence, our study will address two gaps in the literature. Most of all, we aim to demonstrate the country-of-origin effect when consumer knowledge is limited. Secondly, we will retest the established hypotheses of consumer affinity research.
2. LITERATURE REVIEW
Stereotypes and categorization cognitions were viewed as the main source of country-of-origin effect for a long time, even though some differences in product-country image evaluation between novices and experts were brought to attention (Maheswaran, 1994). This also shows different levels of explicit knowledge and familiarity among consumers. Interestingly, almost as a rule, only familiar brands and countries of origin are used as examples in studies. This certainly enables researchers to investigate specific effects in consumer perception, such as consumer animosity, since each highly familiar product-country image is quite distinct. Of course, this is rather unavoidable and many studies of this type are quite specific in their results. All the same, new avenues of country-of-origin research promise fresh insight into the same subject with a closer look into implicit and affective attitude.
2.1. UNFAMILIAR COUNTRY OF ORIGIN
This study aims to detect consumers with low familiarity in country-of-origin research. Therefore, we are introducing a new concept called 'unfamiliar country of origin', which has lower familiarity among consumers, but it is not unfamiliar to all of them. A totally unfamiliar country is unknown to the consumers, i.e. they are not aware of it. It would thus be impossible to measure their attitude to an unknown country of origin, be-
55
AKADEMIJA
cause they have none. Awareness enables them to form perceptions of quality, country-of-origin associations, and even loyalty toward a country (Pappu & Quester, 2010). According to the theory of country equity, awareness is the precondition of country-of-origin effect, on which other facets are built. Although the authors define awareness in a strictly explicit manner as a "consumer's ability to recognize or recall that the country is a producer of a certain product category", we understand awareness in its original implicit way, where retrieval of information from memory is not necessary. Hence, an unfamiliar country of origin is a country that consumers are aware of, even though their explicit knowledge of the product-country image is low or nonexistent.
Compared to unknown countries, it is not impossible to measure the attitude to an unfamiliar country of origin, since each consumer has their own experience and prior knowledge, which adds to their general perceptions of particular countries. If we continue to explain this in terms of the country equity theory, these are associations of the unfamiliar country of origin. The bare minimum would be some general macro image associations related only to country image, which can be attributed to product-country image or micro image, as Pappu and Quester (2010) call it. Instead of a halo effect of a simple or even vague country image (Kleppe et al., 2002), some might have product knowledge. This enables them to use the summary construct to form their product-country image (Han, 1989). These long established effects from the literature are the same in the case of unfamiliar countries. Its images only vary in form from conscious explicit attitude to unconscious implicit attitude, and we would like to uncover the possibility of researching the latter.
The difference between explicit and implicit attitude is in the process of recognition, since recollection or retrieval is a consciously controlled process and mere familiarity is an unconscious process produced by perceptual fluency (Jacoby et al., 1993). Therefore, in the case of high familiarity with the object, consumers are expected to be able to retrieve some relevant information from memory; whereas objects with low familiarity will feel familiar, while any recollection will be difficult. However, these distinctions are not always clear-cut, as we cannot simply link explicit memory to cognitive attitude and implicit memory to affective attitude, because some cognitions are automatic (e.g. stereotypes) and affects can be stored in memory as information. Note that our study focuses on measuring explicit attitudes and familiarity, although we acknowledge that
fluency of implicit memory is its underlying process (Wagner & Gabrieli, 1998).
We have already mentioned country-of-origin studies presenting the automatic stereotype effect, which is evident in the cognitive and affective dimensions of attitude (Herz & Diamanto-poulos, 2013). In order to detect such an unconscious process, one can look at its results with implicit measures, such as an implicit association test (Martin et al., 2011), or explicit measures, such as the Likert scale. Knowledgeable consumers will have no problem forming explicit attitudes and will use the most accessible associations from their memory in their judgment, whereas consumers with limited knowledge have fewer association nodes in their memory, which is why these have a higher chance of being activated (Czellar & Luna, 2010). For this reason, each personal experience or contact with an unfamiliar country of origin can have a great impact, and frequency makes associations more accessible and easier to recollect from memory.
Considering some consumers might have no explicit knowledge of a product-country image, our study focuses on consumer affinity as the basis of an unfamiliar country of origin. This is due to the possible lack of cognitions in consumers who can resort to affect as information in their evaluations (Clore et al. 2001). While both cognitions and emotions can occur without awareness, affective responses in particular can occur without cognitive encoding and in total absence of recognition memory (Zajonc, 1984). Hence, affects in particular can be formed automatically and unconsciously, which makes assessing an emotional connection to any country possible regardless of the level of familiarity.
Academics agree that there is a severe lack of studies on affective attitude in the field of country-of-origin research (Roth & Diamanto-poulos, 2009; Maheswaran et al., 2013), even though an affective image has a stronger impact on the intention to buy (Wang et al., 2012; Li et al., 2014). So far, negative affects toward foreign countries have received considerable attention in the literature (e.g. ethnocentrism, animosity), unlike favorable attitudes, which have largely been neglected. This further encourages us to study favorable facets of an unfamiliar country of origin. Therefore, we refer to the only scale using positive emotions toward a country available in the literature. That is the consumer affinity scale, which is based on feelings of sympathy and attachment to a country (Oberecker & Diamanto-poulos, 2011).
56
AKADEMIJA
2.2. CONSUMER AFFINITY
Oberecker, Riefler, and Diamantopoulos (2008) were the first to conceptualize consumer affinity as being expressed by "a spontaneous or natural liking or sympathy" and by stronger emotions, such as feelings of "cohesiveness and identification" or "attraction". However, the first definition of consumer affinity in the area of country-of-ori-gin research was provided by Jaffe and Nebenzahl (2006). They define it as a favorable and primarily affectively based attitude toward a focal foreign country, which might affect consumer behavior toward products and brands from the affinity country.
When Oberecker et al. (2008) conceptualize consumer affinity, they distinguish idiosyncratic from normative affinities. Idiosyncratic affinities occur at individual level only, while normative affinities depend on cultural influences and occur at the national level. Idiosyncratic consumer affinity is the source of affinity toward an unfamiliar country of origin, whereas normative consumer affinity is more typical of a familiar country of origin, since it is based on knowledge shared with others via mass media or social networks. Moreover, they base their micro and macro drivers of consumer affinity accordingly.
As Nes et al. (2014) explain, their macro drivers seem to express what respondents like about the affinity target and the micro drivers seem to express how they developed this affinity. The latter are based on direct personal experience with the affinity country and/or its citizens, such as personal contacts, travel, and stays abroad. In contrast, macro driver factors are information on country characteristics, which may also be gained indirectly through media, books, or school education, such as lifestyle, scenery, culture, politics, and economics (Oberecker et al. 2008). These factors overlap to a greater or lesser extent with recently developed cognitive consumer affinity dimensions: Wongtada et al. (2012) propose education affinity, business affinity, and people affinity as latent dimensions of consumer affinity; Nes et al. (2014) suggest general affinity, culture/ landscape, music/entertainment, people, and politics dimensions; Bernard and Zarrouk-Karoui (2014) identify personal experience, beauty of nature, and culture as drivers of consumer affinity.
On the other hand, Oberecker and Diamantopoulos (2011) develop a higher-order construct with affectively based sympathy and attachment. Their scale measures affective consumer affinity, while other authors focus on cognitive consumer affinity (Wongtada et al., 2012; Nes et al., 2014; Bernard & Zarrouk-Karoui, 2014). At any rate,
the results consistently show that the cognitive factors of consumer affinity are the drivers of the more general items proposed for affective consumer affinity (Oberecker et al., 2008; Nes et al., 2014; Bernard & Zarrouk-Karoui, 2014). This enables us to simultaneously retest results of both cognitive indirect impact and affective direct impact of consumer affinity on the willingness to buy.
Note that respondents were free to choose their affinity country in the study developing the affective scale of consumer affinity, while other studies had assigned a target country in their survey (Wongtada et al., 2012; Nes et al., 2014). This has probably resulted in strong emotions by respondents (Oberecker & Diamantopoulos, 2011) and might directly impact their intention to buy. Where strong emotions are present, consumers do not particularly care about the product-country image, since their decision is swayed by the strength of their emotions. This is probably why no effect on product-country image was hypothesized in their study. This is where the discrepancy in the literature comes from. Very strong consumer affinity suggests its direct impact on the willingness to buy, whereas weaker levels of affinity take an indirect path in decision-making, mediated by either perceived risk (Oberecker & Diamantopoulos, 2011) or product-country image (Wongtada et al., 2012; Nes et al., 2014; Bernard & Zarrouk-Karoui, 2014).
This distinction can partly arise from the level of product knowledge at the heart of the issue of the unfamiliar country of origin. Some consumers have sufficient product knowledge and thus have no problems forming a product-country image. Those who lack such knowledge can rely on country image associations, also known as the halo effect. Note here that cognitive country image is fully mediated by the product-country image (Li et al., 2014), while the affective country image can directly affect behavior, most commonly measured by the willingness to buy (Wang et al., 2012). Naturally, consumers with limited product knowledge cannot use their nonexistent product-country image in their evaluations. This is why their strong affinity either impacts their willingness to buy directly or they consider the perceived risk in such purchase decisions (Oberecker & Diamantopoulos, 2011).
3. RESEARCH MODEL
Most of the decisions for selecting a particular affinity country were based on respondents' personal attachment to the chosen country due to their vacation experience, family lineage etc.
57
EE3
AKADEMIJA
This indicates the possibilities of analyzing consumer affinity in the case of an unfamiliar country of origin, since some consumers have a strong emotional bond to unfamiliar countries of origin and, consequently, even product knowledge, despite their generally perceived unfamiliarity in the international market. Since our study will include consumers with both high and low familiarity, we will test both the direct and indirect impact of consumer affinity on the willingness to buy in accordance with results from previous research. Hence, hypotheses 1-3 are formed according to the assumption of direct impact of consumer affinity on the intention to buy (Oberecker & Diamantopoulos, 2011), and hypotheses 4 and 5 are formed according to results from the literature on cognitive consumer affinity (Wongtada et al., 2012; Nes et al., 2014). In addition, we have posed another hypothesis in accordance with the context oflow familiarity. The more unfamiliar the country of origin is, the more uncertain the results of the purchase are. Therefore, H6 sugge-sts the mediation or (consumers' perceived risk, because the willingness to buy is determined by the level of uncertainty thaf consumers feel due to low familiarity.
The contribution of Oberecker and Diamanto-poulos (2011) is the most valuable for our study, not only because we base our model on their affective scale of consumer sffinity,but also due to the implementation of perceived risO in the model. Perceived risk, according to them, fully mediates the imcactof cons umer affinity/ on t he willingness to buy. The authors incorporate this construct into their study on the assumption that this is how consumers reduce tCeir on certainty about the purchase oaa product, by preferring familiar options to unfamiliar ones in risky consumptioa situations (Oberecker & Diamantopoulos, 2011).It is tru e that con sumers with no or poor explicit knowledge will use heuristics, since other information is not available to them, and prior research has shown that extrinsic cues,
such as the price or country of origin, can act as risk-relieving information (Aqueveque, 2006).
Hence, in the case of low familiarity, consumers might even simply assess the perceived risk instead of evaluating the product-country image, when information on the latter is unavailable to them. For this reason, we predict an alternative path to indirect impact of consumer affinity, when consumers have scarce data on the product-country image or lack confidence in it. Consumers with low familiarity are unable to evaluate the product-country image and do not have strong affinity toward an unfamiliar country of origin. This is why they rely on their evaluation of the level of risk in their decision on buying products from an unfamiliar country of origin. Thus° we predict that consumer affin ity will affect the perceived risk and the latter will impact the willingness to buy.
Hypothesis 1: The higher the affinity consumers have, the lower the risk they perceive.
Hypothesis 2: The higher the risk consumers perceive, foe less willing they are to buy
As noted above, the mixed results in the literature are probably caused by research design, where respondents either chose rheif own affinity country or are assigned a target country in the survey. The direct impact in the rare of an unfamiliar country of ori gin is not im possible, b ut is highly unlikely due to the lack of experience and prior knowledge of most consumers. Regardless, we will retest it, even ii it seems onlikely that the respondents would have such strong feelings toward an unfamiliar country of origin that is assigned to them in the survey. At ant rate, this would certainly be possible if we looked for a sample with ctrong positive feelings toward a target corntry oc let them choose their own affinity country (Oberecker & Diamantopoulos, 2011). It is also aossible that a memorable experience would generate strong feelings of sympathy and/
Figure1: Proposedresearchmodel
58
AKADEMIJA
or attachment, which would be recollected at the time of the evaluation using affect as information (Clore et al., 2001) in the absence of product knowledge.
Hypothesis 3: The higher the affinity consumers have, the more willing they are to buy.
More recent research focuses on developing scales for cognitive consumer affinity and has consistently demonstrated an indirect impact of consumer affinity mediated by the product-country image (Wongtada et al., 2012; Nes et al., 2014). This is probably due to research design with a target affinity country, which measures the image of a particular country of origin. Thus, the strength of consumers' emotional bond to this target country varies. The same will be the case in our study. For this reason, we expect the impact of consumer affinity on the product-country image (Wongtada et al., 2012; Nes et al., 2014) and the impact of the product-country image on the willingness to buy, which has long been established in the literature (Verlegh and Steen-kamp, 1999; Maheswaran et al., 2013). Under the condition that consumers have the required product knowledge for such retrieval from memory or they are at least sufficiently familiar with the target country in order to form evaluations by halo effect (Han, 1989).
Hypothesis 4: The higher the affinity consumers have, the better the product-country image they hold.
Hypothesis 5: The better the product-country image consumers hold, the more willing they are to buy.
Perceived risk serves as an additional variable to measure outcomes in consumer research and is part of the conceptualization of consumer affinity (Oberecker et al., 2008), but it is not part of the recent analyses of consumer affinity or country-of-origin research. In the case of the latter, there are a few exceptions (see Aqueveque, 2006; Michaelins et al., 2008). We would like to shed some light on this overlooked construct, which could play an important role in unfamiliar country of origin research. Due to the high probability of limited experience and product knowledge, perceived risk should be an important predictor of the willingness to buy. For this reason, we are adding the sixth hypothesis to our retest of the mixed results in the area of consumer affinity.
Hypothesis 6: The better the product-country image consumers hold, the lower the risk they perceive.
4. METHOD
This research design is not a standard one, but it aims to consider all levels of consumer knowledge in order to obtain the highest quality data that can be collected by a survey. Its design is acknowledging that an unfamiliar country of origin might demand a different approach due to a lack of explicit knowledge and confidence in it. We understand that the Likert scale is designed to measure explicit attitudes, namely highly conscious and controllable memory processes. As we attempted to detect the consumers with low familiarity in addition to consumers with high familiarity, we adapted the Likert scale, adding the 'I don't know' option, in order to understand where explicit consumer knowledge is insufficient in the case of low familiarity. At the same time, we changed the typical odd-numbered scale to an even-numbered six-point Likert scale,1 which enables us to detect even the slightest tendencies to negative or positive valence in attitude. This even-numbered scale with the added 'I don't know' option is valid, since the middle neutral option is interchangeable with the added 'I don't know' (Dolnicar & Grün, 2013). Actually, this type of scale has already been tested in the context of low familiarity of brands, because these are very well known among some consumers and quite unfamiliar to others, which is also the case with the unfamiliar country of origin.
To collect quality data, we have acknowledged the problem of low familiarity for some respondents. The 'I don't know' option also points to items that did not affect their evaluation and determines which items are relevant in the case of an unfamiliar country of origin. At the same time, we can control the respondents' level of product knowledge as well as their confidence of evaluating perceived risk and willingness to buy. The distinction between subjective and objective knowledge is well established in the literature (Schaefer, 1997). Despite this, the 'I don't know' option was not added to all items. Consumer affinity was measured by a six-point scale only because it is affectively based and the evaluation is thus possible, even if data is very scarce. If such judgment is too difficult to form, we must assume that the target country is an unknown country of origin to this particular respondent. In contrast, we only aim to detect respondents with high and low familiarity with the target country.
We used Slovenia as an example of a young and certainly unfamiliar country of origin. Moreover, low familiarity with Slovenia was established in previous research (see Kline & Berginc, 2003;
1 Strongly disagree, disagree, partly disagree, partly agree, agree, and strongly agree.
59
AKADEMIJA
Brezovec, 2012). Besides, the country selection also enabled researchers to gather a simple sample incorporating respondents of various backgrounds and with different levels of knowledge. Even though the sample is convenient, it is sufficient for this initial demonstration of country--of-origin effect in the context of low familiarity.
4.1. MEASURES
All items used in our survey were adopted from previous research. Our research model is based on affective consumer affinity and its scales, which measure of feelings of sympathy and attachment (Oberecker & Diamantopoulos, 2011). The construct of consumer affinity can be measured on a cognitive or an affective scale. We adopted the affective approach, because consumers with limited knowledge might struggle to respond to several cognitive dimensions or drivers, which were found to affect consumer affinity due to lack of information. In addition to the consumer affinity scale, we also adopted the perceived risk scale used in the aforementioned study. However, we did not use their positively formed scale for the willingness to buy in our study. In order to check if perceived risk is truly a distinct construct, we adopted a negatively formed scale for the willingness to buy from the study of Wongtada and her colleagues (2012), who simultaneously researched consumer affinity and animosity. We eliminated one item, though, because it refers to a particular product purchase ("I would never buy an American car").
The product-country image has many variations of scales in the literature, since it is measured in almost every study in the field of country-of-origin research. Ever since country equity was conceptualized and its scales of the macro and micro image were developed, they have been widely used in research (Nes et al., 2011). Pappu and Quester (2010) themselves used previously established scales. From them, we adopted an extended scale for the micro image, along with five items from Nagashima (1977) and one item from Aaker (1991). The content of the items was altered from "Televisions made in country Y are..." to "Products made in Slovenia are." Since this is the first study using an unfamiliar country of origin as the target country, we retested items for technical advancement and high status of products, which could produce different results to those obtained by Pappu and Quester (2010). In order to leverage this suggestion of products holding a high status despite coming from an unfamiliar country, we introduced a final, sixth item by Aaker (1991): "Products made in Slovenia are usually good value for money."
4.2. SAMPLE
We conducted a web survey among 362 foreign respondents, but 86 of these questionnaires were not filled out completely. Therefore, we could use only 276 for our analysis. The gathered sample consists mainly of young adults (aged 29 on average) from EU-15 who are well educated, with 32% holding a university degree and 49% a master's degree. This is not surprising, since people from higher socioeconomic classes appear to be more familiar with both domestic and foreign brands (Maheswaran et al., 2013). However, some respondents come from other countries: 29 from the new EU members, such as Hungary, Poland, and Romania; 31 from Balkan countries; 19 from Asia; 13 from North America; 10 from Latin America; 9 from Ukraine and Russia; 4 from Africa; 1 from New Zealand and 1 from Australia. The sample consists of 51% female and 49% male respondents. We acknowledge our convenience sample cannot be projected to any population, but it serves the aims of our study.
Besides, this kind of sample represents an additional challenge to the model, since the coun-try-of-origin effect is typically stronger among older and less educated people (Bhaskaran & Sukumaran, 2007), although it is still evident in the generation Y (Zdravkovic, 2013). Above all, a highly educated sample will ensure better quality data, since educated respondents do not tend to hide their lack of knowledge (Lietz, 2010). As expected, the most problematic construct was the product-country image, since many felt unable to evaluate product attributes without any previous experience or prior knowledge. Hence, we use it as a proxy in order to create groups of respondents with high and low familiarity, because consumers with low familiarity will not be able to form an explicit attitude toward the product-country image, or will not even be able to use the halo effect and infer the attitude from their general country image.
In the sample, there were 95 respondents who responded to all product-country image items. They are representing consumers with high familiarity. The rest represent consumers with low familiarity, with 110 respondents choosing the 'I don't know' option for all six items, demonstrating their complete lack of knowledge. This division into two groups was confirmed by a set of control questions verifying that respondents in the high familiarity group indeed have more contacts with the target country (e.g. friends, visits) and better product knowledge.
60
AKADEMIJA
5. RESULTS
First, the reliability of the variables was tested. Four constructs were assumed and shown highly reliable: consumer affinity (a = 0.869), product-country image (a = 0.915), perceived risk (a = 0.903), and willingness to buy (a = 0.829). In our research model, we included all hypotheses previously confirmed by other researchers. Certainly, we did not expect all to be significant in our model, since they were never used in such a combination before. Therefore, we analyzed the data with multiple regressions; firstly, in order to establish which hypotheses were significant. We created factors from items (using the method of maximum likelihood) according to the four constructs described above in order to conduct the
regressions. As mentioned above, we divided the respondents according to their explicit knowledge into groups of consumers with high familiarity and low familiarity. This distinction proved to be valid, since the results of multiple regressions demonstrated some issues in analyzing the product-country image because of too many missing values caused by the 'I don't know' option. Hence, we had to eliminate this construct in the analysis for consumers with low familiarity. The results showed an indirect path of consumer affinity impact for both groups, but it is one that has not been suggested in the literature.
For consumers with high familiarity, we conducted the first multiple regression with willingness to buy as the dependent variable (R2 = 0.
Table 1: Estimates of standardized regression weights for variables included in the original model
	high familiarity	low familiarity
PCI ^ CA	-.518	/
PR ^ PCI	.429	/
WTB ^ PR	.837	.630
PR ^ CA	/	-.472
Q1a: I have a feeling of sympathy toward this country.	.640	.474
Q1b: I feel a pleasant feeling toward this country.	.795	.693
Q1c: This country inspires me.	.856	.708
Q1d: I feel attached to this country.	.815	.813
Q1e: I love this country.	.744	.799
Q1f: I like Slovenia.	.837	.686
Q1g: I'm captivated by this country.	.793	.788
Q2a: Products made in Slovenia are usually good value for money.	.662	/
Q2b: Products made in Slovenia are high status.	.653	/
Q2c: Products made in Slovenia are innovative.	.787	/
Q2d: Products made in Slovenia are dependable.	.880	/
Q2e: Products made in Slovenia are technically advanced.	.924	/
Q2f: Products made in Slovenia have quality workmanship.	.868	/
Q3a: Getting products from Slovenia is risky.	.837	.825
Q3b: Buying products from Slovenia can have uncertain outcomes.	.891	.880
Q3c: Getting products from Slovenia can lead to bad results.	.927	.837
Q4a: I would feel guilty if I bought a Slovenian product.	.772	.795
Q4b: I do not like the idea of owning Slovenian products.	-.009	-.052
Q4c: Whenever possible, I avoid buying Slovenian products.	.744	.822
Q4d: Whenever available, I would prefer to buy products made in Slovenia.	.871	.761
CA = consumer affinity (Q1), PCI = product-country image (Q2), PR = perceived risk (Q3), WTB = willingness to buy (Q4)
61
EE3
AKADEMIJA
157, F (3.84) = 5.228, p < 0.05) and discovered that only perceived risk has a significant impact on it (p < 0.01). This is contrary to the standard country-of-origin effect, where the product-country image determines the willingness to buy. Here, the product-country image actually has a weaker impact on the intention to buy than consumer affinity. Again, this shows the importance of affects in evaluations, since they can impact the willingness to buy directly without the mediation of the product-country image (Li et al., 2014). Hence, H2 is confirmed for respondents with high familiarity, while H3 and H5 are rejected. Next, we were interested in the perceived risk as a dependent variable in a multiple regression. The results (R2 = 0.225, F (2.87) = 12.625, p < 0.001) support full mediation of perceived risk in the impact of the product-country image (p < 0.001) on the intention to buy. Therefore, H6 is confirmed, while H1 is rejected for consumers with high familiarity. In the end, we tested the only remaining hypothesis, H4, which was also confirmed (R2 = 0 .223, F (1.91) = 26.183,p < 0.001).
We applied similar regressio ns analyses to the group of consumers with low familiarity, although we had to sk.p the construct of the product-country image. Hence, our first multiple regression tested thewiltngness to buy as a dependent variable (R2 = 0.183, F (2.56) = 6.278 , p < 0.05) and again revealed the significance of the mediation of perceived risk (p < 0.05). Thus, hypothesis 2 is supported fon all consumers regardless of their level of familiarity. On the other hand, the direct impact of consumer affinity on the willingness to buy. (1-13) is rejected in both cases, which is to be expected when the target
country is an unfamiliar country of origin. As we eliminated the product-country image from the analysis of respondents with low familiarity, the hypotheses 4-6 were not tested, so we cannot reject or confirm them. At any rate, we were able to test the impact of consumer affinity on perceived risk. The regression results (R2 = 0.194, F ( 1.78) = 18.828, p < 0.001) demonstrated a strong support for H1 in the case of consumers with low familiarity, even though the same hypothesis was rejected in the high familiarity context.
Moreover, we analyzed the path diagram using the structural equation modelling program AMOS. The proposed research model was run according to the results of the regression analysis. Indeed, the model was not applicable satisfactorily to either of the groups. This is why we followed the rule that indicator variables should have standardized regression weights of 0.75 or higher on the latent variable they represent. At the same time, we decided to include at least three items per construct. All variables and their respective weights can be found in Table 1. The model fit was still weak. Therefore, we eliminated items below 0.8 on the constructs of consumer affinity and prc>duct-countny imag=, which bott still had mofe tha= theer items. Afterwards, then model did fit (x2 (62) = 70.136, p = 0.224; NFI = 0.919, CFI = 0^90, TLI = 0.9852 RIM SEA = 0.037) in the case of respontents with high familiarity. However, it is much weaker in t he caseof resp) ondents with low familiarity (x2(25) = 39.331, p = 0.034; NFI = 0.934. CFI = 0.974, TLl = 0.953, RM-SEA = 0.056), probably due to the explicit measure of their attitude.
Figure 2: Model for consumers with high familiarity
62
AKADEMIJA
Figure 3: Model for consumers with low familiarity 6. CONC LUSION
By defining) and operationalizing the unfamili ar country of origin, our study broadens the field of country-of-origin research by extending the pool of target countries from familiar countries with well-known images to less familiar countries of origin. TNs is possible because every cou ntry has its image, even though it might not have had an intenUionh onset. Consumern have an attitude towards any country (Kotler & Gertner, 2002) that they are aware of, although only a few of them are well-known. Consequently, many consumers have an unfamiliar country-of-origin image, which is why this study is valuable. In our attempt to demonstrate ehe c ou ntry-oh-origen effect among consumers with low familiarity, we propose to renew the Likert scale by simply adding the 'I don't know' option for such purposes. This enables researches to c,|vide respondents into separate groups according to theirlevel of Snowledge. At the same t)me, they are able to control their confidence in knowledge on any latent variable.
Moreover, our study once again confirms the importance of affects in attitudes and contributes to the knowledge of consumer affinity,1 Our results indicate that indirect im pacts oh consumer affinity are expected in the case of low levels of familiarity or an unfamiliar country of origin. This is consistent with the literature. Consumer knowledge or familiarity with the product-country image is a moderator of the halo effect (Maheswaran et al., 2013). Hence, a direct effect is observed when consumers are familiar with the country of origin, whereas indirect effects mark lower familiarity with the country of origin. For example, Chinese consumers who are relatively less familiar with Norway inferred that products made in Norway are of high quality based on their macro image of Norway as a Western country (Kleppe et al., 2002).
Due to limited knowledge, consumers use simple heuristics and rely on their evaluation of perceived risk when deciding to purchase products from an unfamiliar country of origin. One exam-p!e of heuristic judgement is attribute substitution, which is used to replace a nlifficult question (What is the product-country image?) with an easier one (How risky is it to buy products from t°is country?). AUtribufe subst|tution commonly occurs when we lack information on a question and have access to the knowledge required to answer a similar question (Kahneman & Frederick, 2002). T°is is why idiosyncratic affinities are vital in building a positive image of any unfamiliar country, because personal contacty and experience increase the confidence of consumers in their own evaluations (Urbonavicius et al., 2011).
Therefore, a product can be bought due to its familiarity, because what we know is what we like, and we ase certain that products from particular countries are of better quality (Chang, 2004). This is why average products from a familiar country of origin are evaluated better than products from an untamiliar country of origin , because familiarity inmreases htereotyping (Smifh et al. 2006). Stereotype effect is beneficial for countries with posrtive stereotypes, while those with a negative image are advised to conceal the origin of their products in order to avoid potentially unfavorable associations. However, a country image is not static and can be built using product quality (Sun & Paswan, 2011). Hence, unfamiliar countries should not conceal the country-of-origin information on its products. Besides, familiarity breeds liking, since familiarity and positive emotions are interrelated (Garcia-Marques et al., 2010).
In a study measuring how a brand image affects the country image, participants evaluated the country of origin more positively when they had
63
AKADEMIJA
some knowledge of the origin of the brand. In this case, the newly acquired brand knowledge biased their evaluation of the country-of-origin image. The study found that for small, less familiar countries, information about the well-known brand led to more positive beliefs about the country's innovation and overall image (White, 2012). This is a clear demonstration of the importance of any (favorable) information for the product-country image, as well as the general country image. At the same time, it shows consumers' doubts in the quality of products from an unfamiliar country of origin, when no information is provided to the consumer, thus creating uncertainty.
7. LIMITATIONS AND FURTHER RESEARCH
The greatest weakness of this study in the small convenience sample, which does not allow for the generalization of our results. Hence, our model should be tested on other samples and for other countries. In addition, our study only addresses products in general, which is sufficient for a preliminary study of the unfamiliar country of origin effect, but it may have partly influenced our results. Consumers might have different products in mind and many of them could be thinking of hedonic products because of the affective nature of consumer affinity. Future research should, therefore, compare hedonic and utility products in order to determine how consumer affinity affects each of these or any other product category.
The results of our study suggest that the constructs of consumer affinity and perceived risk should not be disregarded in country-of-origin research, since they are key variables in (unfamiliar) country-of-origin research, mainly because they can impact consumer evaluation, even in the case of low familiarity (Oberecker & Diamantopoulos, 2011). On the other hand, recent literature reviews focus on country equity (Bayraktar, 2013) or nation equity (Maheswaran et al., 2013) and factors such as consumer animosity, ethnocentrism, or other normative influences. However, they do not even mention consumer affinity, although there is a trend of investigating affectively based constructs and consumer ideologies in country-of-origin research (Dmitrovic & Vida, 2010).
Last but not least, future research should improve the design of the unfamiliar country of origin study and delve deeper into perceived risk and its sources. This area of research is a fresh insight into the field, since it focuses on studying
several neglected avenues in country-of-origin research, such as implicit attitudes (Herz & Diamantopoulos, 2013) and emotions (Oberecker & Diamantopoulos, 2011). Besides, understanding how to build the feeling of consumer affinity and reduce perceived risk would be very beneficial for practitioners and academics alike.
REFERENCES
1.	Aaker, D. A. (1991). Managing Brand Equity. The Free Press: New York.
2.	Aqueveque, C. (2006). Extrinsic Cues and Perceived Risk: The Influence of Consumption Situation. Journal of Consumer Marketing, 23(5): 237-247.
3.	Bayraktar, A. (2013). When is the country-of-origin of a brand a weakness in global markets? International Journal of Management Research and Review, 31(8): 3199-3210.
4.	Bernard, Y., & Zarrouk-Karoui, S. (2014). Reinforcing Willingness to Buy and to Pay Due to Consumer Affinity towards a Foreign Country. International Management Review, 10(2): 57-73.
5.	Bhaskaran, S., & Sukumaran, N. (2007). Contextual and methodological issues in COO studies. Marketing Intelligence & Planning, 25(1): 66-81.
6.	Brezovec, A. (2012). Podobe turizma - od konceptualnih dilem do dinamičnega soustvarjanja pomenov. Družboslovne razprave, 70(28): 89-108.
7.	Chang, C. (2004). Country of origin as a heuristic cue: The effects of message ambiguity and product involvement. Media Psychology, 6(2): 169-192.
8.	Clore, G. L., Gasper, K., & Garvin, E. (2001). Affect as information. In J. P. Forgas, (Ed.). Handbook of Affect and Social Cognition (pp. 121-144). Mahwah, N. J.: Lawrence Erlbaum Associates.
9.	Czellar, S., & Luna, D. (2010). The effect of expertise on the relation between implicit and explicit attitude measures: An information availability/accessibility perspective. Journal of Consumer Psychology, 20(3): 259-273.
10.	Dmitrovic, T., & Vida, I. (2010). Consumer Behaviour Induced by Product Nationality: The Evolution of the Fields and its Theoretical Antecedents. Transformations in Business & Economics, 9(1): 145-165.
11.	Dolnicar, S., & Grün, B. (2013). Including Don't know answer options in brand image surveys improves data quality. International Journal of Market Research, 56(1): 33-50.
12.	Garcia-Marques, T., Mackie, D. M., Claypool, H. M., & Garcia-Marques, L. (2010). Is it familiar or positive? Mutual facilitation of response latencies. Social Cognition, 28(2): 205-218.
13.	Han, C. M. (1989). Country Image: Halo or Summary Construct? Journal of Marketing Research, 26(2): 222-229.
14.	Herz, M. F., & Diamantopoulos, A. (2013). Activation of country stereotypes: automaticity, consonance, and impact. Journal of the Academic Marketing Science, 41(4): 400-417.
15.	Jacoby, L. L., Toth, J. P., & Yonelinas A. P. (1993). Separating Conscious and Unconscious Influences of Memory: Measuring Recollection. Journal of Experimental Psychology, 122(2): 139-154.
64
AKADEMIJA
16.	Jaffe, E. D., & Nebenzahl, I. D. (2006). National Image and Competitive Advantage: The Theory and Practice of Place Branding. Narayana Press: Copenhagen.
17.	Kahneman, D., & Frederick, S. (2002). Representativeness Revisited: Attribute Substitution in Intuitive Judgment. In T. Gilovich, D. Griffin, and D. Kahneman (Eds.), Heuristics and Biases: The Psychology of Intuitive Judgment (pp. 49-81). Cambridge University Press: Cambridge.
18.	Kleppe, I. A., Iversen, N. M., & Stensaker, I. G. (2002). Country images in marketing strategies: Conceptual issues and an empirical Asian illustration. Brand Management, 10(1): 61-74.
19.	Kline, M., & Berginc, D. (2003). Tržna znamka države: Študija primera Slovenije. Teorija in praksa, 40(6): 1040-1057.
20.	Kotler, P., & Gertner, D. (2002). Country as Brand, Product, and Beyond: A Place Marketing and Brand Management Perspective. Journal of Brand Management, 9(4): 249-261.
21.	Li, D., Wang, C. L., Jinng, Y., Bradley, B. R., & Zhang, H. (2014). The asymmetric influence of cognitive and affective country image on rational and experimental purchases. European Journal of Marketing, 48(11/12): 2153-2175.
22.	Lietz, P. (2010). Research into questionnaire design - a summary of literature. International Journal of Market Research, 52(2): 249-272.
23.	Liu, S. S., & Johnson, K. F. (2005). The automatic country-of-origin effect on brand judgements. Journal of Advertising, 34(1): 87-97.
24.	Magnusson, P., Westjohn, S. A., & Zdravkovic, S. (2011). "What? I thought Samsung was Japanese": accurate or not, perceived country of origin matters. International Marketing Review, 28(5): 454-472.
25.	Maheswaran, D. (1994). Country of Origin as a Stereotype: Effects of Consumer Expertise and Attribute Strength on Product Evaluations. The Journal of Consumer Research, 21(2): 354-365.
26.	Maheswaran, D., Chen, C. Y., & He, J. (2013). Nation equity: Integrating the multiple dimensions of country of origin effects. In N. K. Malhotra (Ed.), Review of Marketing Research (Vol. 10, pp. 153-189). Bingley: Emerald Group Publishing Limited.
27.	Martin, B. A. S., Lee, M. S. W., & Lacey, C. (2011). Countering negative country of origin effects using imagery processing. Journal of Consumer Behaviour, 10(2): 80-92.
28.	Michaelins, M., Woisetschlager, D. M., Backhaus, C., & Ahlert, D. (2008). The effects of country of origin and corporate reputation on initial trust: An experimental evaluation of the perception of Polish consumers. International Marketing Review, 25(4): 404-422.
29.	Nagashima, A. (1977). A comparative "made in" product image survey among Japanese businessmen. Journal of Marketing, 41(2): 95-100.
30.	Nes, E. B., Yelkur, R., & Silkoset, R. (2014). Consumer affinity for foreign countries: Construct development, buying behavior consequences and animosity contrasts. International Business Review, 23(4): 774-784.
31.	Oberecker, E. M., & Diamantopoulos, A. (2011). Consumers' emotional bonds with foreign countries:
Does consumer affinity impact buying intentions? Journal of International Marketing, 19(3): 45-72.
32.	Oberecker, E. M., Riefler, P., & Diamantopoulos, A. (2008). The consumer affinity construct: conceptualization, qualitative investigation, and research agenda. Journal of International Marketing, 16(3): 23-56.
33.	Pappu, R., & Quester, P. (2010). Country equity: Conceptualization and empirical evidence. International Business Review, 19(3): 276-291.
34.	Roth, K. P., & Diamantopoulos, A. (2009). Advancing the Country Image Construct. Journal of Business Research, 62(7): 726-740.
35.	Samiee, S., Shimp, T. A., & Sharma, S. (2005). Brand origin recognition accuracy: Its antecedents and consumers' cognitive limitations. Journal of lnternational Business Studies, 36(4): 379-397.
36.	Samiee, S. (2011). Resolving the impasse regarding research on the origins of products and brands. International Marketing Review, 28(5): 473-485.
37.	Schaefer, A. (1997). Consumer knowledge and country of origin effects. European Journal of Marketing, 31(1): 56-72.
38.	Smith, E. R., Miller, D. A., Maitner, A. T., Crump, S. A., Garcia-Marques, T., & Mackie, D. M. (2006). Familiarity can increase stereotyping. Journal of Experimental Social Psychology, 42(4): 471-478.
39.	Sun, Q., & Paswan, A. (2011). Country branding using product quality. Journal of Brand Management, 19(2): 143-157.
40.	Usunier, J.-C. (2006). Relevance in Business Research: The Case of Country-of-Origin Research in Marketing. European Management Review, 3(1): 60-73.
41.	Urbonavicius, S., Dikcius, V., & Navickaite, S. (2011). Country Image and Product Evaluations: Impact of a Personal Contact with a Country. Inzinerine Ekonomika--Engineering Economics, 22(2): 214-221.
42.	Verlegh, P. W. J., & Steenkamp, J.-B. E. M. (1999). A review and meta-analysis of country-of-origin research. Journal of Economic Psychology, 20(5): 521-546.
43.	Wagner, A. D. & Gabrieli, J. D. E. (1998). On the relationship between recognition familiarity and perceptual fluency: Evidence for distinct mnemonic process. Acta Psychologica, 98(2-3): 211-230.
44.	Wang, C. L., Li, D., Barnes, B. R., & Anh, J. (2012). Country image, product image and consumer purchase intention: Evidence from an emerging economy. International Business Review, 21(6): 1041-1051.
45.	White, C. L. (2012). Brands and national image: An exploration of inverse country-of-origin effect. Place Branding and Public Diplomacy, 8(2): 110-118.
46.	Wongtada, N., Rice, G., & Bandyopadhyay, S. (2012). Developing and validating affinity: A new scale to measure consumer affinity toward foreign countries. Journal of International Consumer Marketing, 14(3): 147-67.
47.	Zajonc, R. B. (1984). On the Primacy of Affect. American Psychologist, 39(2): 117-123.
48.	Zdravkovic, S. (2013). Does country-of-origin matter to Generation Y? Young Consumers: Insight and Ideas for Responsible Marketers, 14(1): 89-102.
65