PAPER Ital. J. Food Sci., vol. 27 - 2015 181 - Keywords: Fresh red meat consumption, Principal Component, K-Means Cluster and Multiple Regression Analyses, Sensory and hedonic quality attributes - HOW SENSORY AND HEDONIC QUALITY ATTRIBUTES AFFECT FRESH RED MEAT CONSUMPTION DECISION OF TURKISH CONSUMERS? Y. TOPCU*, A.S. UZUNDUMLU and D. BARAN Department of Agricultural Economics, College of Agriculture, Ataturk University, 25240-Erzurum, Turkey *Corresponding author: Tel. +90 442 2311393, yavuztopcu@atauni.edu.tr; ytopcu25@hotmail.com, ytopcu025@gmail.com ABSTRACT The aim of the study is to explore how the sensory and hedonic quality attributes of the fresh red meat affect its consumption preference and amounts of Turkish consumers. The data obtained from 385 households in Erzurum were used for Principal Component (PCA), K-means Cluster and Multiple Regression/Correlation (MRC) Analyses. The results of the study highlighted considera- bly that the sensory quality attributes on their consumption preference had a much bigger effect than the hedonic ones for each cluster. However, its price and their income accepted as the im- portant indicators of the hedonic ones, but a much lower impact on all clusters. mailto:yavuztopcu@atauni.edu.tr mailto:ytopcu25@hotmail.com mailto:ytopcu025@gmail.com 182 Ital. J. Food Sci., vol. 27 - 2015 INTRODUCTION In recent years, the increases of the red meat prices due to the contractions in the meat supply covering the production, processing and market- ing of the meat and the meat products obliged the political units of the government to import live red meat materials, and thus they have pro- vided stabilization at some levels for the meat supply and price (YAVUZ et al., 2013). Howev- er, the origins of the meat sources among the consumers consuming the fresh red meat have been considered as an important decision fac- tor. A few portion of them, therefore, exhibited the buying attitude and behaviors within a mod- erate consumption trend of those preferred the meat sources imported while the others formed the purchase models with a positive motivation in the red meat consumption preferences based on the national meat sources. There was a string relationship between the origin of the red meat sources as an indicator of the consumers’ purchase decisions and both the sensory quality including in its intrinsic qual- ity attributes and the hedonic quality consist- ing of its extrinsic quality attributes (BERNUES et al., 2003). The strong motivational effects of their individual, demographic, socioeconomic, psychological and health characteristics, on the other hand, are among the primary preference factors, as well (TOPCU, 2012). Turkish consumers’ purchase decisions based on the food safety and quality of the fresh red meat have changed considerably due to its con- fidential and sensory quality attributes affected directly by microbiological insecurity (dioxin ef- fect, BSE and salmonella diseases) of foreign- origined red meat sources, and by the contam- ination risks with various pollutants (antibiotic and hormones) in the last years (TOPCU, 2015; TOPCU and UZUNDUMLU, 2012). They have also concerned about deterioration of their health conditions with the meat consumption of the imported animals being exposed to genetic ma- nipulations, and then about inheriting to the future generations of the negative phenomenon (TOPCU, 2012). In addition to imported live animal sources, presence of unhealthy fresh red meats penetrat- ing to domestic markets uncontrollably and il- legally, unknown origin, exposing to some addi- tives at levels threatening human health, and the pollution, contamination and microbiological in- security of the manufacturing meats in the live- stock farms and meat processing facilities un- der unhygienic conditions have caused the con- sumers to exhibit more sensitive purchase atti- tude and behaviours by considering the sensory quality attributes (YAVUZ et al., 2013). On the other hand, not only digestive and cor- onary heart diseases caused by obesity based on the passive lifestyles revealed the communi- cation and information age but also chronic ill- nesses resulting from nutrition with the inten- sive protein diets consisting of the red meats and the impacts of the life circle of the mature peo- ple have gradually reduced the consumers’ red meat consumption tendencies, and thus nowa- days their purchase decisions towards the red meats have been changing habitually (REALINI et al., 2013; TOPCU, 2012). The analyses of the customers reviews/as- sessments about the intrinsic quality attributes having a direct relationship between the senso- ry quality and confidential attributes (food qual- ity measured by the chemical and microbiolog- ical tests) of the red meats such as the struc- tural and visual characteristics (REALINI et al., 2013; McCARTY et al., 2003) are of a much more importance. On the other hand, they also fo- cus on the extrinsic quality attributes correlat- ing with the hedonic quality attributes such as the actual product image, the cost to custom- er, the origin of the meat source, disposable in- come (TROY and KERRY, 2010; BURNUES et al., 2003; McCARTY et al., 2003), and reflecting the consumers’ individual characteristics and de- termining their purchase decisions about the fresh red meat. Therefore, eliminating the fac- tors affecting negatively their consumption sat- isfaction and loyalty accepted at the focal point of the production, consumption and marketing activities, determining the main attributes ac- celerating the consumption trends of the homog- enous target consumers masses, and then de- signing the marketing tactic and strategies for them could provide the important advantages in terms of the efficiency of the production re- sources and the maximization of the expected utilities on all market dynamics. The diet meeting the main requirements of the people by providing the metabolic energy needed for biological organisms consists of the plant and animal foods, and thus they must be consumed to sustain a healthy and balanced life of the peo- ple at an adequate level (TOPCU and UZUNDUM- LU, 2012). In particular, the animal-derived pro- teins such as the meat, milk, eggs, cheese being of a sufficient level for the essential amino ac- ids are also suitable in terms of digestion. About 75-80% of high-quality animal-derived proteins are transformed into the body proteins. In con- trast, plant-derived proteins classified as low- quality protein have an insufficient level of some essential amino acids, and it is benefited from only 40% of their proteins since their digestion is fairly difficult (TOPCU, 2012). A person must consume 70 gr proteins per day for an adequate and balanced diet; therefore, it must be at least half of animal origin protein (SEKER et al., 2011). The consumption amounts of the animal origin foodstuffs per capita are very low in Turkey, but that of the plant-derived foods is greater than that of the developed coun- tries (KARKACIER, 2000). If people fed with ce- real-based food products give more weight into Ital. J. Food Sci., vol. 27 - 2015 183 the animal origin ones in Turkey, and they gain the proper eating habits, it could be contributed significantly to the improvement of the life qual- ity of the community. Today, the consumption levels of the ani- mal-origin products are considered as a devel- opment indicator of the countries, and thus as their socioeconomic structures improves, the consumption amount of the protein foods in- creases in contrast to reduce that of the car- bohydrate foods (YAYLAK et al., 2010). As a re- sult of all this, the meat includes an impor- tant part of people’s daily diets in USA, EU and other developed countries, and it meets 15%, 40% and 20% of the energy, protein and fat re- quirements per day, respectively, but its con- sumption fluctuates significantly in various regions of the world (DANIEL et al., 2011). For example, while the annual meat consumption amounts per capita in 2011 were calculated as 142, 125, 82 and 80 kg in Austria, USA, Ger- many and UK, representing the leader coun- tries in the meat production and consumption, respectively, it was only 12 kg in Turkey. Ac- cording to the data obtained from the research area (Erzurum), on the other hand, the annu- al meat consumption amount per capita calcu- lated as 17.8 kg in 2008, 12.5 and 11.5 kg in 2011 and 2012 were found (TOPCU and UZUN- DUMLU, 2012). The substitution of the imported red meats instead of the red meat derived from the animal husbandry based on a quality pastures and a wide variety flora having a significant impact on the red meat quality caused the consumers’ preferences shift to other meat groups in the research area, and their red meat consumption trends decreased significantly by influencing negatively its sensory and hedonic quality at- tributes (TOPCU and UZUNDUMLU, 2012). As stated above, the annual red meat con- sumption amount per capita in the research area and Turkey was about 5-8 times less than that in developed countries, but this difference continued to increase steadily, in the last five years. Especially, there has been an increasing trend towards the poultry meat consumption due to the dissatisfactions of not only the sen- sory quality attributes at the meat production, manufacture and retail levels based on the im- ported meat materials but also the hedonic qual- ity attributes pointing the actual product imag- es. However, it has been experienced the signif- icant changes such as the region origin and the types of the red meat sources in the consump- tion preferences of the consumers obtained from the red meats about 58, 34 and 36% of the total meat consumption in the developed countries, Turkey and the research area, respectively (McA- FEE et al., 2010; LICHTENSTEIN et al., 2006). The red meat consumption amounts at much low levels in Turkey and the study area could result from not only the sensory and hedonic quality attributes of the fresh red meat but also the socioeconomic and individual attitudes of the consumers and the confidence to the man- ufacturer, marketer and retailers (REALINI et al., 2013; TROY and KERRY, 2010; McCARTY et al., 2003). However, the supply amount of the fresh red meat in the research area is higher than its demand amount due to its high sen- sory quality attributes resulting from the ad- vantages of the organic cattle fattening farm- ing based on the quality pastures and rich flo- ra varieties under the agroecological and topo- graphical properties at the high mountainous areas of the region (TOPCU and UZUNDUMLU, 2012). The consumers, therefore, considering the advantages of both the hedonic quality at- tributes reflecting the images of the core and actual products focused on the numerous fac- tors of the marketing mix and higher sensory quality attributes have preferred and consumed intensively them. As a result, the consumers residing in the research region and migrating from the research area to the different regions of Turkey have not only maintained their de- mands for the red meats with the region of the origin to provide much higher main utility but also contributed to the regional/rural develop- ment through the improvement of the farmer families’ life qualities. On the other hand, the inefficient policies im- plemented to meet the fresh meat deficit without effect on the rural/regional developments and the sensory quality attributes affecting negative- ly the consumer satisfaction with the penetra- tion of the imported red meat to domestic mar- kets caused by the supply contraction of the red meat have led the consumers with dissatisfac- tions to reflect the different purchase decisions, in the last decade. All these individual responses of the consumers about the fresh red meat and their purchase decisions have emerged as the acceptable results of the relationships between the sensory and hedonic quality attributes. This study, therefore, was designed to reach all the objectives mentioned above. In this scope, the main aims of the study are to determine how the sensory and hedonic quality attributes af- fect the fresh red meat consumption decisions of Turkish consumers; and then to construct the target homogenous consumer segments based on their consumption frequencies, and finally to analyze the effectiveness of the factors effecting on their consumption amounts. MATERIAL AND METHODS Material The preliminary data used in this study were obtained from a survey consisting of Turkish consumers’ attitude and behaviors with re- spect to the fresh red meat consumption de- 184 Ital. J. Food Sci., vol. 27 - 2015 cisions conducted in Erzurum,1 Turkey. In or- der to determine the sample size (selected sta- tistically from the local household consum- ers through simple random sampling method) population minimizing sample bias and repre- senting the main population correctly, the city center was divided into three districts covering Aziziye, Palandoken and Yakutiye districts with 6.562, 30.022 and 44.075 households at the west, south and north-east parts of Erzurum, respectively (ANONYMOUS, 2013). METHODS Method used in determination of the sample size In order to calculate the sample size for each district, the following formula was used (TOPCU et al., 2010). 385 )1(** 2 2 = − = c ppZ n Where, n = sample size, Z = z value (1.96 for 95% confidence level), p = percentage making a choice (0.5 used for sample size needed), c = con- fidence interval (used 0.05 = ±5). Then, based on the population of each district, the weight- ed sample size and distribution of the surveys for each district were determined proportional- ly. A total of 385 surveys were distributed with 52, 136 and 197 questionnaires allocated to the Aziziye, Palandoken and Yakutiye Districts, re- spectively. Methods used in the preparation of the questionnaires Participants of the survey were asked to re- spond to each statement indicating the signifi- cance level of the fresh red meat attributes for them. A liker-type scale was used (where 1 re- fers to the least important and 5 refer to the most important attribute). Table 1 reported that 35 attributes of the fresh red meat consisted of 17 sensory quality attributes and 18 hedonic quality ones (TOPCU, 2015; TOPCU and UZUN- DUMLU, 2012). Of five numeric variables refer- ring to the consumers’ socioeconomic character- istics, on the other hand, two referred to their monthly fresh red meat consumption amount and purchase frequency and three included in their monthly income ($), the price per kg of the fresh red meat ($/kg), and the share of its ex- penditure within total food one. Methods used in the statistics analyses After editing and coding, the primary data were first used in Principal Component Analy- 1 Erzurum (39°45’N latitude and 41°15’E longitude) is lo- cated in the north-east region of Turkey and is on the Silk Road. Erzurum is one of the biggest provinces in the re- gion with a total population 778.195, and the city centre population is 509.474. 2 A factor extraction method used to form uncorrelated lin- ear combinations of the observed variables. The first com- ponent has the maximum variance. Successive compo- nents explain progressively smaller portions of the var- iance and are all uncorrelated with each other. PCA is used to obtain the initial factor solution. It can be used when there is a single correlation matrix. 3 This method is an orthogonal rotation method that mini- mizes the number of variables that have high loading on each factor. It simplifies the interpretation of the factors. sis (PCA)2 to determine the main factors relat- ed to the product attitudes influencing on the consumers’ fresh red meat purchase patterns. PCA is a data reduction technique that reduc- es the number of variables used in an analysis by creating new variables (called factors) that combine redundancy in the data (SPSS 15.0, 2006). The first step in PCA is to determine the number of relevant factors. This was conduct- ed by PCA using the varimax rotation method (VRM)3. PCA was used initially to identify under- lying aspects explaining a correlation among a set of the food product attributes. The purpose of PCA was to identify those attributes account- ing for a relatively large proportion of the vari- ance in the sample. In the second and final step of the statistical analyses, the main factors obtained from PCA were used for k-means cluster and multiple re- gression/correlation MRC Analyses, respective- ly. In the second step, according to the fresh red meat consumption frequencies, therefore, the target consumers were separated to three ho- mogeneous clusters including in light users (2-3 times per month), medium users (2-3 times per 15 days) and heavy users (2-3 times per week) (TOPCU, 2012), and then the main factors were allocated to the homogeneous consumer clus- ters based on the meat consumption frequen- cies of the target consumers by k-means clus- ter analysis. In the final step, MRC analysis was used for the measurements of the efficiency of the main factors and socioeconomic variables (independ- ent variables) effecting on the consumers’ meat consumption amounts (dependent variable). In the MRC model taken into consideration the var- iables related to the main factors obtained from the PCA and some socioeconomic characteris- tics of the consumers, therefore, it was tried to measure the relationships between monthly consumption amount of the fresh red meat per household as a dependent variable and the main factors including in their sensory and hedonic quality attributes as long with the consumers’ socioeconomic characteristics called as the in- dependent variables. In order to test whether or not the normal dis- tribution of the main factors’ group scores de- Ital. J. Food Sci., vol. 27 - 2015 185 rived from PCA and socioeconomic variables col- lected from the consumers exhibited, was ap- plied the various transformations techniques, and finally logarithmic transformation one pro- viding the closest distribution to the normal was chosen. On the other hand, the parameter coef- ficients were estimated by using ordinary least squares (OLS). Individual and group significance of these coefficients were tested using t and F tests, respectively. In order to evaluate wheth- er to be any econometrical problem among the variables, it was tested the overall multicollinar- ity and auto-correlation problems by consider- ing the variance-inflating factor (VIF) and Durbin- Watson d statistics, respectively. Multicollinear- ity among variables was detected by calculating (VIF) (GUJARATI, 2005; SPSS 15.0, 2006). MRC model could be written as following equa- tion: ),,,,,,,,,,,,( iDISINCMEPRCFORORGORGLOCMESDURCOSCONNUTVALIMAGVISQLYCRERISMICSAFAROMAfFMCONS e= Dependent Variable FMCONS : Monthly fresh red meat consumption amount (kg) Independent Variables AROMA : Aroma of the fresh red meat (sensory quality attribute) MICSAF : Microbial safety (associated with sensory quality attributes) CRERIS : Chemical residue-related risk (associated with sensory quality attributes) VISQLY : Visible quality (sensory quality attribute) IMAG : Accual fresh meat image (hedonic quality attribute) NUTVAL : Nutritional value (associated with sensory quality attributes) COSCON : Cost to the consumer (hedonic quality attribute) MESDUR : Meat sources and its durability (associated with hedonic quality attribute) ORGLOC : Local-origined organic meat willingness (hedonic quality attribute) FORORG : Foreign-origined meat willingness (hedonic quality attribute) MEPRC : The price of the meat ($) (hedonic quality attribute) DISINC : The disposable income (hedonic quality attribute) All analyses were run in sequence with the SPSS 15.0 statistical software. These techniques were used widely to analyze the meat attributes playing the important roles on the consumers’ consumption preference and amounts in mar- keting studies (TOPCU and UZUNDUMLU, 2012; KRYSTALLIS et al., 2007; DIEZ et al., 2006; OLI- VER et al., 2006; VERBEKE and VACKIER, 2004; BERNUES et al., 2003). RESULTS AND DISCUSSION Results of PCA related to the sensory and hedonic quality attributes of fresh red meat Kaiser Normalization (KMO) which compares partial correlation coefficients with observed ones including in the sensory and hedonic quali- ty attributes effecting on the fresh red meat con- sumption preferences of the Turkish consumers was calculated as 0.85, and this means that the data set for the PCA was at a perfect level since the test score was greater than 0.5 (Table 1). On the other hand, the chi-square value calcu- lated for Bartlett’s test of Sphericity statistics of their main factors with respect to the consumers’ meat consumption preferences was calculated as 5060.68 (p: 0.000), and thus the unit matrix hypothesis was rejected (p<0.01). The PCA us- ing varimax rotation method grouped the thirty- seven variables related to the sensory and he- donic quality attributes for fresh red meat into the ten main factors with Eigen-values greater than 1, and the main factors explained the 71% of the total variance (Table 1). As shown in Table 1, the results of the study showed that aroma (F1) and visible quality (F4) affecting the fresh red meat consumption prefer- ences of Turkish consumers by explaining about 29.6 and 5.3% of total variance constituted di- rectly the first and second sensory quality fac- tors, respectively. By having an important impact on the quality attributes of F1 and F4; moreo- ver, microbial safety (F2) and chemical residue-re- lated risk (F3) associated much closely with the sensory quality attributes of the fresh red meat were interpreted as the second and third senso- ry quality factors indicating about 7.5 and 7.1% of total variance. The same main factors for the sensory quality attributes based on the various similar variables taken into consideration by TOPCU and UZUNDUMLU (2012), KERGOAT et al. (2010), TROY and KERRY (2010), AASLYNG et al. (2007), KRYSTALLIS et al. (2007) were reported in their previous researches, as well. On the other hand, the results of the present study indicated that actual product image (F5); nutritional value (F6), cost to the consumer (F7) and meat sources and durability (F8) of the fresh meat; local-origined organic meat willingness (F9) 186 Ital. J. Food Sci., vol. 27 - 2015 Table 1 - The measurement results of PCA with regard to the sensory and hedonic quality attributes based on the fresh red meats consumption decisions of Turkish consumers. Factos interpretions and the variables Factor loadings* F1 F2 F3 F4 F5 F6 F7 F8 F9 F10 AROMA (F1: AROMA) Flavour 0.841 0.078 0.121 0.127 0.032 0.127 0.102 0.039 0.015 0.029 Succulence 0.831 0.076 0.061 0.138 0.047 0.076 0.014 -0.007 -0.035 -0.014 Smell/odour 0.750 0.083 0.115 0.083 0.205 0.051 0.026 0.165 -0.035 -0.038 Juiciness 0.705 -0.003 0.086 0.146 0.017 0.178 0.054 0.120 0.099 0.277 Visible fat 0.530 0.251 -0.021 0.070 -0.014 0.246 0.217 0.068 0.355 -0.333 MICROBIAL SAFETY (F2: MICSAF) Dioxin scare 0.079 0.880 0.192 0.088 0.103 0.071 0.037 0.008 0.112 -0.019 Salmonella disease 0.088 0.875 0.096 0.120 0.096 0.162 0.053 0.040 0.049 0.006 BSE 0.081 0.802 0.133 0.151 -0.004 0.130 0.119 0.050 0.044 0.011 Trustworthiness 0.090 0.666 0.073 0.119 0.209 0.262 0.163 0.122 -0.188 -0.242 CHEMICAL RESIDUE-RELATED RISK (F3: CRERIS) Meat manufacturing methods 0.139 0.161 0.698 0.090 0.178 0.108 0.087 0.086 0.059 -0.013 Antibiotic residues 0.128 0.352 0.687 0.022 0.220 0.172 0.010 0.064 -0.111 0.193 Hormone residues 0.017 0.449 0.603 0.153 0.092 0.148 0.065 0.156 -0.163 0.122 VISIBLE QUALITY (F4: VISQLY) Freshness of the meat 0.346 0.164 0.010 0.789 0.028 0.155 0.085 -0.038 0.032 0.039 Meat part obtained from the animal 0.038 -0.007 0.077 0.772 0.084 0.091 0.098 0.211 0.013 0.080 Natural structure of the meat 0.197 0.103 0.138 0.762 0.161 0.164 0.074 -0.015 0.064 0.065 The meat colour 0.104 0.234 0.093 0.740 0.149 0.121 0.119 0.154 0.001 -0.078 Tenderness 0.140 0.053 0.342 0.527 0.077 -0.012 0.210 0.231 0.237 0.244 ACCUAL FRESH MEAT IMAGE (F5: IMAG) Brand name 0.179 -0.020 0.239 0.113 0.769 -0.006 0.041 0.021 0.145 0.016 Package 0.029 0.142 0.012 0.130 0.672 0.133 0.306 0.274 -0.006 0.080 Advertisement -0.066 0.002 0.141 0.181 0.647 0.055 0.152 0.251 0.398 -0.114 Free of additives/natural product 0.264 0.144 0.190 0.091 0.627 0.193 0.173 -0.044 -0.030 0.005 Labelling 0.134 0.134 0.156 0.068 0.605 0.151 0.474 0.131 -0.151 0.055 NUTRITIONAL VALUE (F6: NUTVAL) Brand name 0.112 0.111 0.200 0.216 0.153 0.832 0.000 0.091 0.070 0.072 Package 0.206 0.153 0.230 0.183 0.115 0.820 0.056 0.097 0.114 0.109 Advertisement 0.248 0.254 0.147 0.139 0.072 0.764 0.168 0.018 -0.064 0.021 Organic product 0.172 0.029 0.188 0.100 0.175 0.754 0.171 0.095 0.064 -0.044 COST TO THE CONSUMERS (F7: COSCON) Discounts in the meat prices -0.019 0.136 0.220 0.158 0.143 -0.031 0.707 0.065 0.295 0.006 The initial price of the meat 0.120 0.114 0.040 0.069 0.253 0.098 0.702 -0.030 -0.103 -0.029 Outlets/sales points of the meat 0.187 -0.034 0.097 0.146 0.154 0.087 0.693 0.238 0.196 0.041 WILLINGNESS TO BUY MEAT SOURCES AND ITS DURABILITY (F8: MESDUR) Beef 0.180 0.068 0.144 0.212 0.002 0.064 0.113 0.777 0.030 -0.037 Sheep and goad meats 0.052 0.024 0.036 0.026 0.253 0.046 -0.017 0.747 0.221 0.049 The shelf life of the meats 0.330 0.050 0.245 0.063 0.105 0.104 0.327 0.551 -0.128 -0.024 LOCAL-ORIGINED ORGANIC MEAT (F9: ORGLOC) Willingness to buy local-orig. meat 0.301 -0.055 0.159 0.124 0.094 0.309 0.020 -0.007 0.593 0.127 Willingness to buy organic meat 0.068 0.192 0.076 0.222 0.136 0.239 0.137 0.235 0.582 0.130 WILLINGNESS TO BUY FOREIGN-ORIGINED MEAT (F10: FORORG) Willingness to buy foreign-orig. meat -0.055 0.145 -0.019 0.029 0.107 0.063 0.118 0.113 0.062 0.803 Eiqen-values 10.655 2.689 2.554 1.914 1.545 1.452 1.233 1.173 1.119 1.034 Share of explained variance (%) 29.598 7.468 7.095 5.318 4.292 4.033 3.426 3.263 3.109 2.872 Cumulative share of that (%) 29.598 37.066 44.160 49.478 53.763 57.803 61.229 64.492 67.600 70.472 KMO (Kaiser-Meyer-Olkin) statistic 0.848 Bartlett’s test of Sphericity [Chi - square (2, df:630):5060.68](p:0.000) * Bold numbers indicated the largest loading for each variable. Ital. J. Food Sci., vol. 27 - 2015 187 and foreign-origined meat willingness (F10) con- structing the hedonic quality attributes of the fresh red meat consisted of about 4.3, 4.0, 3.4, 3.3, 3.1 and 2.9% of total variance, respectively (Table 1). When the obtained main factors relat- ed to the hedonic quality attributes in the pre- sent study were compared with the others pre- sented by the results of the previous research- es; while the F5, F6 and F7 factors with respect to the hedonic quality were supported direct- ly with the results of the studies conducted by FURNOLS et al. (2011), KRYSTALLIS et al. (2007), AUDEBERT et al. (2006), BERNUES et al. (2003) and VERBEKE and VACKIER (2004), the F8, F9 and F10 factors focused on the attributes of the hedonic quality were of the strong relationships with the results of those presented by FURNOLS et al. (2011), NAPOLITANO et al. (2010), BERNUES et al. (2003), McCARTY et al. (2003). As a result of all these, the sensory quality at- tributes explaining about 70% of the main fac- tors effecting on the fresh red meat consump- tion preferences of Turkish consumers were of a much bigger influence than the hedonic qual- ity ones. Results of cluster analysis related to the sensory and hedonic quality attributes of fresh red meat The sensory and hedonic quality attributes derived from the PCA and effecting on the fresh red meat purchasing patterns of Turkish con- sumers were separated into three homogeneous consumers segments through k-means cluster according to their red meat purchasing frequen- cies including the light, medium and heavy us- ers. The main factors reflecting Turkish con- sumers’ attitude and behaviors impacting on each homogeneous consumers segment were given in Table 2. The results of the study showed that the heavy users of the fresh red meat (C1) preferred its con- sumption according to the aroma and visible quality being the main components of the sen- sory quality by taking into consideration the lo- cal-origined organic meat willingness accepted as an indicator of the hedonic quality attributes. Therefore, the homogeneous consumers in the C1 determined their purchase attitude and be- haviors by considering the sensory quality at- tributes of the red meat delivered from the an- imal materials fattened under the advantages of the superior agroecological characteristics of the study area. The fresh red meat presentations positioned at the retailer shelves by highlight- ing their core benefits, and causing no changes on the sensory quality attributes of them with the designation of the local-origined meat un- der the generic brands for the consumers in C1 could have an important impact on their pur- chase patterns. The results of the study also indicated that the medium users of the fresh red meat (C2) fo- cused on the purchase patterns constructed by a combination of the factors affecting the senso- ry quality including in the microbial safety and chemical residue-related risk associated with willingness to obtain the salutary meats mini- mizing the negative effects on human healthy at the feeding and nourishment levels of the live- stocks and the manufacture levels of the red meats. They, moreover, paid a major attention to the hedonic quality related to the nutritional value and the cost of the ownership minimizing their expenditures by being provided easily and Table 2 - Final cluster centres and the number of cases in each cluster. Main factors Clusters* Heavy users (C1)** Medium users (C2)** Light users (C3)** AROMA (F1) 0.465 -0.122 -0.127 MICSAF (F2) -0.751 0.423 -0.174 CRERIS (F3) -0.641 0.342 0.073 VISQLY (F4) 0.156 -0.241 0.076 IMAG (F5) -0.263 -0.034 0.333 NUTVAL (F6) -0.955 0.328 0.215 COSCON (F7) -0.042 0.079 -0.009 MESDUR (F8) 0.117 -0.890 0.473 ORGLOC (F9) 0.330 -0.423 0.108 FORORG (F10) -0.078 -0.320 0.222 Disposable income (DISINC) 3.570 3.280 2.980 The prices of the meats (MEPRC) 1.340 1.330 1.330 Number of total cases in each cluster *** 81 193 111 % of total cases in each cluster 21% 50% 29% * Bold numbers indicate the largest final cluster centre scores for each factor. ** According to F statistics, the final cluster center scors were found very importance (p<0.01). *** The total number of the cases (n): 385. 188 Ital. J. Food Sci., vol. 27 - 2015 comfortably to the consumer of C2. The imple- mentations of differentiated actual product strat- egies based on the price discriminations under own brands of the manufacturers of the red meat with food safety that there was no negative im- pact on their health, therefore, could also affect positively their purchasing motivation by empha- sizing the core benefits of the red meat for C2. The results of the study, furthermore, ex- plained that the light users of the fresh red meat (C3) tended to buy the fresh red meat with the hedonic quality attributes positioned to the focus point of the purchase attitude psychology alien- ated under the actual fresh meat image with the foreign-origined meat willingness, its resources and durability. They, thus, preferred the national or global brands exhibiting the actual fresh red meat image by turning to durable foreign-orig- ined meat sources with long shelf life. It could be implemented the marketing tactic and strat- egies focused on the designs of the actual fresh red meat image under the name of the nation- al and global brands for the consumers in C3. Results of MRC analysis related to the sensory and hedonic quality attributes of fresh red meat The results of the statistical tests in Table 3 reported that the VIF values with 1.06 and 2.47 indicating the scores between 1.00 and 2.50 de- termining the acceptable reference range and 1.85 Durbin-Watson d statistics positioned be- tween d u (1.96) and 4-d u (1.68) referred that there were no the econometrics problems for multicol- linarity and auto-correlation in the MRC model (GUJARATI, 2005; KALAYCI, 2005). According to the results diagnosing the econometrics prob- lems, the data sets could be used directly for the MRC model. The determination statistics, OLS estimates of the parameter coefficients and other statistic measurements such as F and t, collinearity and correlation matrix scores were given in Table 3. The results of the MRC analysis indicated that the determination coefficient (R2) and adjusted (adj.) R2 was calculated as 0.51 and 0.50 in the MRC model, and thus all the independent vari- ables explained the least 50% of the dependent variable. Rejecting the null hypothesis, further- more, referring that there was no the relation- ships between the dependent variable and all the independent variables by making the parame- ter coefficients of them equal to zero, F-statistic was calculated as 6898 (p:0.000). On the oth- er hand, the partial regression coefficients of all the independent variables taking into consider- ation t-statistics, except for those that of the mi- crobial safety (MICSAF), the cost to the consum- ers (COSCON), the meat sources and its durabil- ity (MESDUR) and the foreign-origined meat will- ingness (FORORG), were statistically found to be meaningful (t c(df:12;0.01/0.05) ); and then the signs of their coefficients including in the chemical residue-related safety (CRERIS) and the price of the meats (MEPRC) with negative and the oth- ers with positive ones at the levels (p:0.01 and 0.05) were also found consistent with the eco- nomic theories. The results of the MRC analysis also high- lighted that the visible quality (VISQLY), the aro- ma (AROMA), the nutritional value (NUTVAL), the accual fresh meat image (IMAG), MEPRC, the lo- cal-origined organic meat willingness (ORGLOC), CRERIS, the disposable income (DISINC) were of very important impacts on the fresh red meat Table 3 - The measurement results of the MRC analysis and some statistic tests. n: 385 R2: 0.51 Adj.R2: 0.50 F c(12;361) : 6.898* DW d c : 1.85 Variables MRC model Collinearity statistics Correlations β:Coefficientsa Std. error t c -value p-value Tolerance VIF Zero-order Partial Part Constant 1.478 0.563 2.623 0.009* - - - - - LOGAROMA 0.484 0.189 2.556 0.011* 0.374 2.472 0.027 0.133 0.121 LOGMICSAF 0.023 0.093 0.244 0.807 0.550 1.818 0.037 0.013 0.012 LOGCRERIS -0.161 0.086 -1.868 0.053** 0.559 1.788 0.102 -0.098 -0.088 LOGVISQLY 0.553 0.194 2.844 0.005* 0.487 2.483 0.093 0.148 0.135 LOGIMAJ 0.245 0.086 2.845 0.005* 0.461 2.167 0.115 0.148 0.135 LOGNUTVAL 0.370 0.136 2.716 0.007* 0.413 2.418 0.035 0.142 0.129 LOGCOSCON -0.107 0.084 -1.277 0.202 0.603 1.658 0.016 -0.067 -0.060 LOGMESDUR 0.008 0.081 0.095 0.924 0.682 1.465 0.022 0.005 0.005 LOGORGLOC 0.210 0.109 1.920 0.056** 0.428 2.338 0.078 0.101 0.091 LOGFORORG 0.004 0.026 0.156 0.876 0.778 1.285 -0.019 0.008 0.007 LOGMEPRC -0.240 0.091 -2.633 0.009* 0.947 1.056 -0.140 -0.137 -0.125 LOGDISINC 0.106 0.072 3.791 0.001* 0.886 1.129 0.133 0.078 0.070 a The coefficients consisted of the standardized coefficients. * (p<0.01) **(p<0.05). Ital. J. Food Sci., vol. 27 - 2015 189 consumption amounts of Turkish consumers (p<0.01 and 0.05). In relation to the profiles of the cluster segments based on the underlying dimensions of the sensory and hedonic quality attributes emerged from PCA, the satisfaction of VISQLY and AROMA identifying the sensory quality attributes of the fresh red meat support- ed by ORGLOC of Turkish consumers in C1 in- creased considerably very high effects on their meat consumption amounts with 0.55 (p<0.01), 0.48 (p<0.01) and 0.21 (p<0.01) rates, respec- tively. When considered the credence quality char- acteristics assigned to the sensory quality at- tributes affecting the fresh red meat consump- tion amounts of Turkish consumers in C2, there was an adverse relationship between CRERIS (β: -0.16 p<0.05) and its consumption amounts, but not NUTVAL (β: 0.37 p<0.01). As minimized its CRERIS, and improved its NUTVAL, therefore, it could be increased those at a lower level than all the others. IMAG (β: 0.24 p<0.01) standing for the hedonic quality attributes and NUTVAL (β: 0.37 p<0.01) symbozing the sensory ones, on the other hand, had a moderate impact on those of them in C3. Overall, the average agreement of all the cluster members in relation to the price of the meat (MEPRC) exhibiting an adverse rela- tion with its consumption amounts occurred at a moderate level (β: -0.24 p<0.01), but the dis- posable income (DISINC) having a linear relation with those substantiated at lower levels (β: 0.11 p<0.01) than the others. The results of the present study confirmed briefly that the sensory quality attributes, espe- cially VISQLY, AROMA, NUTVAL, on the fresh red meat consumption amounts had a much higher impact than the hedonic quality attributes such as IMAG, MEPRC, DISINC. In particular, the re- sults emphasized that the pleasure of Turkish consumers provided from its consumption were under not only the effects of the numerous post- purchasing criterion (aroma, nutritional value, microbial safety, etc.) but also the impacts of the various pre-purchasing processes based on the visual quality cues (colour, freshness, natural structure, specific animal parts and meat out- lets, etc.) impacting on their purchase decisions. The results of the present study with regard to the sensory and hedonic quality attributes of the fresh red meat are in line with the results of the previous studies reported that the con- sumers paid particular much more attention to the sensory quality attributes of the red meat inspected visually in a pre-purchasing process (colour, freshness, natural structure, fat content, specific animal parts and meat outlets) (REALINI et al., 2013; TROY and KERRY, 2010; AASLYNG et al., 2007; VERBEKE and VACKIER, 2004), and then to its credence attributes perceived such as nutritional value (vitamin and protein content), microbial freeness (salmonella, dioxins, BSE) (KRYSTALLIS et al., 2007; VERBEKE and VACK- IER, 2004) and chemical freeness (hormones, antibiotics) (KRYSTALLIS et al., 2007; VERBEKE and VACKIER, 2004) rather than the hedonic ones (meat image, price, income, country of or- igin) (FURNOLS et al., 2011; NAPOLITANO et al., 2010; KRYSTALLIS et al., 2007; AUDEBERT et al., 2006; OLIVER et al., 2006; VERBEKE and VACK- IER, 2004; BERNUES et al., 2003). CONCLUSIONS The results of the study confirmed clearly that the heavy users in C1 attributed a much more importance to the sensory quality attributes based on willingness to buy the local-origined organic meat endorsing the aroma as an indica- tor of the post-purchasing satisfaction followed by the fresh red meat consumption and the visu- al quality considered in the pre-purchasing pro- cess of it than the others. On the other hand, the medium and light users in C2 and C3 paid also a more attention to the sensory quality attrib- utes including in the credence attributes such as the microbial and chemical freeness and the nutritional value of it than the members of C1 as considered the efficiency scores of the sen- sory and hedonic quality attribute variables ef- fecting on its consumption amounts of Turkish consumers in MRC model. Consequently, the sensory quality attributes had a much bigger impact on its consumption decision and amounts of Turkish consumers than the hedonic quality attributes in the over- all agreements with regard to the importance of all the factors. Therefore, the fresh red meat consumption amount in Turkey maintaining at a very low level of the annual fresh red meat consumption amount per capita (about 12 kg) when compared with the leading countries in the meat consumption could be increased meaning- fully by the innovative red meat and meat prod- ucts designed through the product and price dif- ferentiations under the production, manufac- turing and marketing tactic and strategies tak- ing into account first its core benefit based on the sensory quality attributes such as the visi- ble quality, aroma and credence attribute, and then its actual benefit focused on the hedonic quality attributes (the country of the origin, its actual image and price and their disposable in- come) for each clusters. REFERENCES Aaslyng M.D., Oksama M., Olsen E.V., Bejerholm C., Baltz- er M., Andersen G., Bredie W.L.P., Byrne D.V. and Gabri- elsen G. 2007. The impact of sensory quality of pork on consumer preference. Meat Sci. 76(1): 61-73. Anonymous. 2013. The Annual Report based on the house- holds, the data of documentation centre of Erzurum Met- ropolitan Municipality, Erzurum, Turkey. Audebert O., Deiss V. and Rousset S. 2006. Hedonism as 190 Ital. J. Food Sci., vol. 27 - 2015 a predictor of attitudes of young French women toward meat. Appetite 46: 239-247. Bernues A.A., Olaizola A. and Corcoran K. 2003. Extrinsic attributes of red meat as indicators of quality in Europe: an application for market segmentation. Food Qual. and Pref. 14: 265-276. Daniel C.R., Cross A.J., Koebnick C. and Sinha R. 2011. Trends in meat consumption in the USA. Public Health Nutr. 14(4): 575-583. Díez J., del Coz, J.J., Bahamonde A., Sañudo C., Olleta J.L., Macie S. 2006. Identifying market segments in beef: breed, slaughter weight and ageing time implications. Meat Sci. 74(4): 667-75. Furnols M.F., Realini C., Montossi F., Sañudo C., Campo M.M., Oliver M.A., Nute G.R. and Guerrero L. 2011. Con- sumer’s purchasing intention for lamb meat affected by country of origin, feeding system and meat price: a con- joint study in Spain, France and United Kingdom. Food Qual. and Pref. 22: 443-451. Gujarati, D.N. 2005. “Basic Econometrics” 10th ed., Refer- ence Publish: 33, ISBN 975-7860-99-9, Istanbul, Turkey. Kalayci S. 2005. “SPSS Applications, Multi-variables Sta- tistics Techniques” 1st ed., Asil Publish, ISBN 975-9091- 14-3, Ankara, Turkey. Karkacıer O. 2000. An analysis of import demand for dairy products in Turkey. Turk J. Agric. For. 24: 421-427. Kergoat M., Giboreau A., Nicod H., Faye P., Diaz E., Beetschen M.A., Gerritsen N. and Meyer T. 2010. Psychographic measures and sensory consumer tests: When emotional experience and feeling-based judgments account for pref- erences. Food Qual. and Pref. 21: 178-187. Krystallis A., Chryssochoidis G. and Scholderer J. 2007. Consumer-perceived quality in ‘traditional’ food chains: the case of the Greek meat supply chain. Appetite 48: 54-68. Lichtenstein A.H., Appel L.J. and Brands M. 2006. Diet and lifestyle recommendations revision 2006: a Scientif- ic Statement from the American Heart Association Nu- trition Committee. Circulation 114: 82-96. McAfee A.J., McSorley E.M., Cuskelly G.J. Moss B.W., Wal- lace J.M.W. and Bonham M.P. 2010. Red meat consump- tion: An overview of the risks and benefits. Meat Sci. 84(1): 1-13. McCarthy M., Boer M., O’Reilly S. and Cotter L. 2003. Fac- tors influencing intention to purchase beef in the Irish market. Meat Sci. 65: 1071-1083. Napolitano F., Braghieri A., Piasentier E., Favotto S., Nas- petti S. and Zanoli R. 2010. Cheese liking and consumer willingness to pay as affected by information about or- ganic production. J. Dairy Res. 77: 273-279. Oliver M., Nute G. Font I., Furnols M., San Julián R., Cam- po M. 2006. Eating quality for beef, from different pro- duction system, assessed by German, Spanish and Brit- ish consumers. Meat Sci. 74: 435-442. Realini C.E., Furnols M.F., Sañudo C, Montossi F., Oliver M.A. and Guerrero L. 2013. Spanish, French and Brit- ish consumers’ acceptability of Uruguayan beef, and con- sumers’ beef choice associated with country of origin, fin- ishing diet and meat price. Meat Sci. 95: 14-21. Seker I., Ozen A., Guler H., Seker P. and Ozden I. 2011. Red meat consumption behaviors in Elazıg and the consum- ers’ opinion about animal welfare. Turk J. Vet. Anim. Sci. 17(4): 543-550. SPSS Base 15.0, 2006. “SPSS Base 15 User’s Guide” 15th ed., ISBN 1-56827-370-3, Chicago, IL. Topcu Y, Turhan B. and Uzundumlu A.S. 2010. Analysis of factors affecting Turkish sunflower oil consumer be- havior: the case of Erzurum province. Ital. J. Food Sci. 22 (2): 239-247. Topcu Y. and Uzundumlu A.S. 2012. Analysis of the factors affecting the consumers’attitude and behaviors towards the meat consumption: the case of Erzurum. Paper No. 24, presented at 10th Annual Congress of Internation- al Agriculture Economy, Konya, Turkey, September 5-7. Topcu Y. 2012. Integrated marketing approach as a rural development tool. Ch. 13. In “Rural Development-Con- temporary Issue and practices”, R.S. Adisa (Ed.), p. 257. InTech-Open Access Publisher, ISBN: 978-953-307-942- 4, Rijeka, Croatia. Topcu Y. 2015. Turkish consumer decisions affecting ice cream consumption. Ital. J. Food Sci. 27 (1): 29-39. Troy D.J. and Kerry J.P. 2010. Consumer perception and the role of science in the meat industry. Meat Sci. 86: 214-226. Verbeke W. and Vackier I. 2004. Profile and effects of con- sumer involvement in fresh meat. Meat Sci. 67: 159-168. Yavuz F., Bilgic A., Terin M. and Guler, I.O., 2013. Policy im- plications of trends in Turkey’s meat sector with respect to 2023 vision. Meat Sci. 95(4): 798-804. Yaylak E., Taskın T., Koyubenbe N. and Konca Y. 2010. A study on determination of red meat consumption be- haviors in Odemiş, Izmir. J. Animal Prod. 51(1): 21-30. Paper Received February 14, 2014 Accepted August 6, 2014