A questionnaire was used to collect data related to the attributes affecting consumers’ meat purchase decisions and consumers’ cooking endpoint temperature preferences. The sample size (n = 384) was determined with following formula [11–16]:

where the sample size was represented by n, the significance level is represented by t (assumed to be 95% with the table value of 1.96), the stage of sampling is represented by b (which is equal to 1), the probability of the examined situation occurring is represented by p (p = 0.5 was used for the absence of preliminary information related to the criteria affecting consumers’ meat purchase decisions), the probability of the examined situation not occurring is represented by q (q = 1–p), and the accepted error is represented by e (assumed to be 5%). When b equal to 1, the Eq. (1) was converted to the following equation:

The survey was conducted outside the selected supermarkets selling beef products. The questionnaire was prepared as an intercept survey designed to be completed in a few minutes by participating customers [16]. The questions were designed to determine (a) consumers’ purchasing preferences related to the marbling degrees by showing them the pictures of marbling illustrations of North American Meat Processors Association [17] and their general knowledge related to relationship between marbling and meat quality (b) consumers’ perception of beef quality related to some intrinsic and extrinsic quality cues during the meat purchase [8,18] and (c) consumers’ consumption preferences related to the degree of doneness by showing them the pictures of degree of doneness illustrations of North American Meat Processors Association [17].

In the second phase, the aim of the study was to investigate the consumers’ preferences related to the degree of doneness using two whole pieces of boneless ribeye (high quality and standard quality defined by the manufacturer) that were available in the retail market. In addition, it was also investigated that whether any differences could be detected by the consumers between two whole pieces of boneless ribeye that were cooked to same degree of doneness (71°C, 77°C, and 82°C). The whole boneless ribeye samples were transported to the laboratory after obtained from the supplier within 45 minutes in an insulated container and stored at refrigerated conditions. The samples were firstly subjected to proximate analyses to determine moisture, ash, fat, and protein content. Then, the boneless ribeye samples were cut into 2.54 cm thick steaks.

Panelists were recruited from three different districts of Adana, Turkey. A total of 103 untrained consumer panelists participated in a total of three sessions completed in same day with 40 participants from Seyhan district, 31 participants from Yuregir district and 32 participants from Cukurova district. The samples were cooked on electrical grill (Grill Comfort; Tefal, Rumilly, France). The internal temperature of each sample was continuously monitored using a handheld thermometer (Type K Thermometer; Verth, Taipei, Taiwan) during the cooking. The samples were cooked to internal temperatures of 71°C (medium), 77°C (well done) and 82°C (very well done) [17,19].

The cooked samples were subjected to a simple ranking test to determine the most preferred sample [20]. 1 cm²×2.5 cm pieces were cut from cooked steaks and served for consumers. Both high quality and standard quality boneless ribeye were cooked to three degrees of doneness (medium, 71°C; well done, 77°C; very well done, 82°C) and served to panelists in random order separately. The panelists were asked to rank the samples from the most preferred to the least preferred (1 to 3). In addition, a paired comparison (two-sided) test was used to determine consumers’ preference for the quality grade after cooking samples to three different degrees of doneness (medium, 71°C, well done, 77°C, very well done, 82°C). Cooked boneless ribeye samples were served to panelists in pairs. The panelists were asked to choose the preferred sample between standard quality and high quality.

The oven (Memmert; Universal Oven Tech., Schwabach, Germany) drying method was used to determine moisture contents of the samples at 100°C for 18 h. All the organic matter was ashed in a muffle furnace (PLF 130/45; Protherm, Ankara, Turkey) at 600°C to determine ash contents of the samples. While Soxhlet extraction method was used to determine fat content, the Kjeldahl method was used to determine crude protein content of the samples [21,22].

In the first phase of the study, the data collected from the survey were subjected to Chi-square test, binomial test or Kruskal Wallis test using SPSS software version 20 (IBM SPSS Statistics, Armonk, NY, USA). A simple ranking test was used to determine consumers’ preference for the degree of doneness in sensory evaluation. The data were analyzed with calculating rank sums for each sample. Then a Friedman-type rank test was performed. The nonparametric analogue to Fisher’s least significant difference for rank sums was calculated to further evaluate the samples differed significantly. In addition, a paired comparison (two-sided) test was used to determine consumers’ preference for the quality grade at three different degrees of doneness [20].

The data was collected by showing pictures of marbling illustrations (moderately abundant, slightly abundant, moderate, modest, small, and slight) to the consumers and asking them to indicate their purchasing preference (Figure 1). In the current study, 30.7% of the participants indicated that slight was their purchasing preference while 25% of the participants indicated that small was their purchasing preference. In addition, 16.9% and 13.3% of the participants indicated that modest and moderate were their purchasing preferences, respectively. Furthermore only 6.5% and 7.6% of the total consumers surveyed in the current study reported that slightly abundant and moderately abundant were their purchasing preferences, respectively. There were significant differences (p<0.05; χ² = 107.125; df = 5) among the consumers’ preferences for marbling degree. Consumers use some intrinsic and extrinsic quality cues including marbling, leanness, color, packaging and price etc. to make a buying decision at the point of purchase [25]. Historically, tenderness and juiciness of meat were associated with a pleasant consumption experience [26]. Although health-conscious consumers often preferred low-fat meat to the palatability, consumers perceived the meat juicier when the amount of fat increased. Some of the fat present was released in the first bite or during chewing, thereby stimulating the salivary glands. Thus the meat with higher fat content was perceived as juicier [27]. Savell and Cross [26] developed the “window of acceptability” to show the overall relationship between increased intramuscular fat (marbling) and palatability. They stated that meat with intramuscular fat content of 3% to 7.3% was acceptable. In addition, quality grades of beef carcasses are used to determine eating characteristics including tenderness and palatability of the product. Assignment of a specific quality grade to a carcass requires evaluation of sex characteristics, maturity, the quality of the lean muscle and the degree of marbling [17]. Furthermore, marbling affects the juiciness, tenderness, flavor and appearance of the meat [27]. In the current study, the total preference rate of visuals with slight and small marbling was determined as 55.7% while the total preference rate of visuals with slightly abundant and moderately abundant marbling was determined as 14.1%. These results indicated that the higher degrees of marbling might be considered too fatty for purchasing by the consumers.

Furthermore, consumers’ perceptions were also questioned regarding the relationship between marbling and beef quality (Figure 2). Firstly, the participants were asked whether they consider marbling or leanness as a quality attribute for beef. There was a significant difference between consumers’ perception for the leanness and marbling as a quality attribute (p<0.05; χ² = 9.375; df = 1). While 57.8% of the participants indicated that leanness was a quality attribute, 42.2% of the participants indicated that marbling was a quality attribute. In addition, the consumers were also asked how the quality of beef would change when the degree of marbling increased. There was a significant difference between consumers’ interpretations on how the degree of marbling would affect the quality of beef (p<0.05; χ² = 10.010; df = 1). Although 58.1% of the participants indicated that the quality of beef would decrease as the degree of marbling increase, 41.9% of the participants indicated that the quality of beef would increase. While 42.2% of consumers considered marbling as a quality attribute, 41.9% of them stated that higher degree of marbling would affect the quality of beef positively. Nevertheless, the overall preference rate of the pictures of marbling illustrations with moderately abundant (7.6%) and slightly abundant (6.5%) was 14.1%. In addition, 58.1% of the consumers indicated a negative correlation between degree of marbling and beef quality. Even though some participants stated that they cared about intramuscular fat as a quality criterion, they might not have a clear understanding of what marbling was. In fact, the percentage of those who considered leanness as a quality attribute was 57.8%. This agreed with the overall preference rate of the pictures of marbling illustrations with small and slight (55.7%). In general, the results indicated that the higher degrees of marbling might be perceived as too fatty by the consumers and might not be acknowledged as a cue for a higher quality meat.

A questionnaire was designed using a Likert Scale of 1 to 5 (1 “definitely not important”, 2 “not important”, 3 “neutral”, 4 “important”, 5 “definitely important”) to collect data of importance of some attributes (color, smell, marbling, storage conditions, the presence of inspector, the presence of legal stamp, fat content, fat color, use-by date, the temperature of fresh meat aisle and the compliance with quality standards) related to intrinsic and extrinsic quality cues for the consumers during meat purchase (Figure 3).

Color was indicated as definitely important by 80.5% of the participants while 17.7% and 1.8% of the participants indicated that color was important and neutral, respectively. One of the most important quality attribute of meat was color which also used as an indication of eating quality and safety of meat by consumers [8]. The bright color of meat was associated with the freshness of meat for predicting the quality [5]. Furthermore, the bright red color was defined by most consumers as a desirable meat color and an important factor for purchasing decision due to probable perception of the discoloration as a spoilage indicator [28]. In addition, Droval et al [29] reported that the main quality attributes of meat include appearance, texture, juiciness and flavor and the initial selection of meat is mostly related to the appearance during purchase. Likewise, Arenas de Moreno et al [28] reported that the intrinsic attributes of meat including color, tenderness, juiciness, smell, flavor, and freshness were considered important for the purchasing decision for the most of Venezuelan consumers. Similarly, 98.2% of the consumers indicated that color was definitely important or important during meat purchase in the current study.

Smell was indicated as definitely important by 77.1% of the participants while 20.3%, 2.3%, and 0.3% of the participants indicated that smell was important, neutral, and not important, respectively. Smell was among the experience attributes that mostly related to the safety [8]. In Turkey, meat can be displayed unpackaged in supermarkets. Any spoilage odors generating from the meat might be detected by consumers and could be discouraging during meat purchasing. Becker et al [30] reported that smell was considered either very important or quite important by over 90% of the consumers in Germany. Likewise, 97.4% of the consumers indicated that smell was definitely important or important in the current study. Conversely, Seko et al [5] indicated that consumers were guided by extrinsic quality cues (the salesperson’s expertise, price and quality of the service) during purchasing dibiterie meat while intrinsic quality cues including color and smell of fresh meat had lower concern in Dakar, Senegal.

Marbling was indicated as definitely important by 38% of the participants while 28.4%, 18.5%, 8.1%, and 7.0% of the participants indicated that marbling was important, neutral, not important, and definitely not important, respectively. Marbling was associated with increased beef eating quality [31]. Savell and Cross [26] stated that meat with intramuscular fat content of 3% to 7.3% was acceptable (window of acceptability) for most consumers. However, they also indicated that the palatability decreased significantly with decreases in the fat content if the amount of intramuscular fat was less than 3%. When the fat content exceeded 7.3%, health-conscious consumers defined those meats as too fatty due to the apparent fat content. In addition, Verbeke et al [32] reported that most of European consumers considered lean beef as the healthiest meat. Furthermore, Australian consumers preferred learner meat and avoided visible fat due to health reasons while Asian consumers distinguished intramuscular fat (marbling) from the subcutaneous fat and intramuscular fat was associated with a premium eating experience [33]. Although a total of 66.4% of consumers indicated that marbling definitely important and important during meat purchase, only 14.1% of the consumers preferred the visuals with slightly abundant and moderately abundant marbling in the present study. Similarly, 58.1% of the consumers indicated a negative correlation between degree of marbling and beef quality. These results confirmed that majority of the Turkish consumers might not have a clear understanding of the marbling and they perceived the meat with higher degrees of marbling as too fatty for purchasing. These results were also supported by the importance of fat content and fat color (Figure 3) for consumers during purchasing. Fat content and fat color were indicated by 86.2% and 86.5% of the consumers, respectively as definitely important or important which showed that consumers were very concern about fat and fat related indicators.

Storage conditions were indicated as definitely important by 79% of the participants while 21% of the participants indicated that storage conditions were important. The presence of an inspector was indicated as definitely important by 50% of the participants while 33%, 10%, 6%, and 1% of the participants indicated that the presence of an inspector was important, neutral, not important and definitely not important, respectively. The presence of the legal stamp on carcasses was indicated as definitely important by 39% of the participants while 37%, 15%, 6%, and 3% of the participants indicated that the presence of the legal stamp on carcasses was important, neutral, not important and definitely not important, respectively. Fat content of beef was indicated as definitely important by 43% of the participants while 43%, 8%, 5%, and 1% of the participants indicated that fat content of beef was important, neutral, not important and definitely not important, respectively. Fat color was indicated as definitely important by 41% of the participants while 46%, 7%, 5%, and 1% of the participants indicated that fat color was important, neutral, not important and definitely not important, respectively. Use-by date was indicated as definitely important by 51% of the participants while 35%, 5%, 5% and 4% of the participants indicated that use-by date was important, neutral, not important and definitely not important, respectively. The temperature of fresh meat aisle was indicated as definitely important by 65% of the participants while 30%, 4%, and 1% of the participants indicated that the temperature of fresh meat aisle was important, neutral and not important, respectively. The compliance with quality standards was indicated as definitely important by 51% of the participants while 39%, 4%, 4% and 2% of the participants indicated that the compliance with quality standards was important, neutral, not important and definitely not important, respectively. Becker et al [30] reported that extrinsic cues may play an important role for quality selection at the point of purchase and some of the extrinsic and intrinsic cues could be used by consumers to assess the safety of meat. Similarly, intrinsic cues (color, smell, tenderness, flavor, freshness, and juiciness) and extrinsic cues (aging, hygiene, origin, breed, and animal feeding information) were important for Venezuelan beef consumers in making a buying decision [28]. In contrary, Senegalese consumers were reported to be guided by extrinsic quality cues (the salesperson’s expertise, price, and quality of the service) rather than intrinsic quality cues (color and smell of fresh meat) during purchasing dibiterie meat [5]. In the current study, most of the consumers considered storage conditions, inspector, legal stamp, use-by date, the temperature of fresh meat aisle and the compliance with quality standards definitely important or important during meat purchase. These results indicated that consumers were looking for the cues that indicated not only quality of meat but also safety of the meat at the point of purchase.

The participants were asked to indicate their preference for degree of doneness (very well done, 82°C; well done, 77°C; medium, 71°C; medium rare, 63°C; rare, 60°C; and very rare, 55°C) of beef steak (Figure 4). In this study, 26% and 30% of the participants indicated that very well done and well done were their degrees of doneness preferences, respectively while 24% of the participants indicated that medium was their degree of doneness preference for the consumption of beef. Only, 11%, 5% and 4% of the participants indicated that their degree of doneness preferences were medium rare, rare and very rare, respectively. There were significant differences (p<0.05; χ² = 151.063; df = 5) among consumers’ degree of doneness preferences. The total preference rate of visuals with very well done and well done was 56% while the total preference rate of visuals with medium rare, rare and very rare was 20%. Lorenzen et al [10] reported that steaks cooked to lower temperatures were tenderer than those cooked to higher temperature due to having higher scores for liking of tenderness, higher scores for liking of juiciness and lower Warner-Bratzler shear values. However, they found no differences in overall liking or liking of flavor among the steaks cooked to different end-point temperatures (very rare, 55°C; rare, 60°C; medium rare, 63°C; medium, 71°C; well done, 77°C; and very well done, 82°C). Reicks et al [34] conducted a study in three different cities including Phoenix, AZ; Baltimore, MD/Washington DC; Lubbock, TX to represent the United States consumers. Among the consumers, 3.4% preferred rare (cool-red center), 26.6% preferred medium rare (warm-red center), 30% preferred medium (pink throughout), 30% preferred medium well (thin line of pink) and 9.9% preferred well done for beef degree of doneness. Conversely, 56% of Turkish consumers preferred well done or very well done while only 9.9% of the U.S. consumer indicated that they preferred well done.

In the second phase of the study, whole boneless ribeye available in the retail markets were used to investigate the consumers’ preferences related to the degree of doneness. These pieces of whole boneless ribeye represented meat samples that could be purchased by the consumers without any additional effort and classified by the manufacturer as “standard quality” and “high quality”. In Turkish markets, meat is offered to customers without any information related the quality grade of the meat during purchase. For most customers, the quality grade is not one of the determining factors for the meat prices since the meat prices are not determined in relation to the quality grade. However, the manufacturers are grading the carcasses to sell higher quality meat to their special customers for higher prices. Thus, we simply asked a local manufacturer for a high quality and a standard quality whole boneless ribeye. The manufacturer supplied us with two whole boneless ribeye samples obtained from healthy 16 to 18 months old male Charolais steers.

Table 2 presents results of the proximate analyses of boneless ribeye samples including moisture, ash, fat, and protein obtained from high quality and standard quality carcasses. There were no statistical differences between the boneless ribeye samples classified as high quality and standard quality for ash, fat, and protein values. However, the moisture values of the samples were statistically different (p<0.05). In addition, chemical lipid values of boneless ribeye samples in the current study were in the range of Low Choice to Top Choice since Low Choice steaks were reported to contain 4% to 5% of chemical lipid with a small degree of marbling and Top Choice steaks were reported to contain 6% to 7% of chemical lipid with modest and moderate degrees of marbling [26,27]

A simple ranking test was used to determine consumers’ preference for the degree of doneness (Figure 5). Both high quality and standard quality boneless ribeye samples were cooked separately to three different degrees of doneness (medium, 71°C; well done, 77°C; very well done, 82°C) and served to panelists separately in random order. The panelists were asked to rank the samples from the most preferred to the least preferred sample (1 to 3). There were significant differences among the rank sums of standard quality samples cooked to three different degrees of doneness (p<0.05). The rank sums of samples were 249 for medium, 195 for well done and 174 for very well done. Similarly, there were significant differences among the rank sums of high-quality samples cooked to three different degrees of doneness (p< 0.05). The rank sums of samples were 256 for medium, 192 for well done and 170 for very well done. Furthermore, multiple comparison tests for the rank sums indicated that the samples cooked to medium were preferred the least, and the samples cooked to well done and very well done were preferred the most for both standard quality and high-quality samples. The results of the simple ranking test seemed to agree with the results of the survey study that indicated that the total preference rate of visuals with very well done and well done was 56%. Conversely, Lorenzen et al [10] studied the effects of cooking beef steaks to six end point temperatures from very rare (55°C) to up to very well done (82°C). Although cooking steaks to lower end point temperatures were liked by consumers for tenderness and juiciness, the end point temperature had no effect on overall liking. Similarly, beef eating satisfaction was reported that greatly affected by degree of doneness. When cooking end point temperature increased, beef cuts reported to get tougher, less juicy and overall liking decreased [35]. Furthermore, Drey et al [36] reported that increased degree of doneness caused similar negative impact on consumer ratings regardless of marbling level. Contradictory to the literature our results indicated that Turkish consumers might prefer higher levels of degree of doneness even though the tenderness and juiciness were affected negatively. Although marbling level and degree of doneness were important factors for higher consumers’ satisfaction, consumers’ preferences for lower marbling degrees and higher degrees of doneness in some countries would create challenges for the meat industry to satisfy these consumers. For this purpose, a study conducted by Benli and Tokgoz [37] indicated that meat cuts might need to be tenderized to overcome the difficulties due to the raw material quality to satisfy the consumers.

A paired comparison (two-sided) test was used to determine consumers’ preferences for the quality grade after cooking samples to three different degrees of doneness. Both high quality and standard quality boneless ribeye samples were cooked separately to specific degrees of doneness including medium (71°C), well done (77°C), very well done (82°C) and served to panelists in pairs. The panelists were asked to choose the preferred sample between standard quality and high quality samples (Figure 6). There were no significant differences among the paired samples at any degree of doneness. In this study we simple asked from a meat producer for a high quality and standard quality whole boneless ribeye that were normally supplied to their customers. Overall, our results and experience indicated that it was difficult to obtain high quality meat in Turkish market even after paying for it if you were not one of the producers’ privileged customers. It seemed that consumers are still at the mercy of the butchers or suppliers to obtain high quality meat from the market. Consumers must develop some personal relationship (regularly buying meat from the same retailer, being a friend of the butcher etc.) to get better quality meat in regular basis. In addition, these results also indicated that the suppliers might be aware of consumers’ choice of the lower marbling levels due to their concerns about the fatty meat.