Figure 1 shows the color scores of roast beef treatments. Although the added nitrite levels were different, a typical red cured meat beef color was achieved and scored in the range of “like slightly” to “like moderately”. This could be due to that only small quantities of nitrite are required to produce the cured meat color: theoretically 2 - 14 mg/kg is sufficient to convert half of the myoglobin present in fresh meat, but more nitrite is required to provide for the competing reactions, and 25 mg/kg are required to give a stable cured color throughout the extended shelf life [29] [30] . Moreover, vacuum packaging could be considerably effective to increase the desirable meat color during storage [31] .

Color scores are usually affected by nitrite level, but in our case, the insignificant difference in color could be explained by the fact that the magnitude of differences in color between samples was very little and the panelists were unable to detect it. Color is one of the main food appearance attributes and has very significant role in the quality evaluation of meat and meat products as well as purchase decision [32] [33] . Therefore, it is beneficial to enhance other quality characteristics without affecting this significant parameter.

The values of flavor scores for the different treatments as shown in Figure 2 were numerically close and not statistically different (P > 0.05). There were no significant preferences between treatments regarding flavor and were scored as “like moderately”. The effect of nitrite was much greater at the higher levels than at the lower levels of nitrite, in that the flavor was not significantly different between the different levels. These results are in agreement with those found by [34] , so we can conclude that the amount of nitrite present 0.018% and 0.026% did not influence the flavor of roast beef. It was also indicated that nitrite has no effect in cured flavor development in all cured meat [12] [35] . On the other hand, [36] suggested that cured flavor may be a result of combination of complex nitrite-related flavor and aroma, and suppression of rancidity formation, since nitrite has antioxidant effect.

As shown from Figure 3, the juiciness scores of the roast beef treatments were also not significantly different. There were no significant preferences between treatments which were scored as “like slightly to like moderately”. The juiciness scores of all treatments were acceptable. Higher juiciness score was found in treatment A which contained 0.006% nitrite. The sensory scores indicated a decrease in juiciness in treatment D with 0.026% nitrite, but it was statistically insignificant (P > 0.05). [37] reported that juiciness is related to the type of meat used in the formulations, rather than the chemical composition when he evaluated the sensory properties of five canned luncheon meat formulations.

The texture of the roast beef that ranged from “like slightly” to “like moderately” (Figure 4) indicated that roast beef samples with 0.018% added nitrite had a softer texture than roast beef with 0.006% added nitrite. The inclusion of nitrite, at low level (0.006%), seems to make the texture of roast beef more acceptable. Increasing the

amount of nitrite from 0.006% to 0.018% had no additional effect on texture scores; it seems likely that nitrite curing affected the texture of the meat. That is in disagreement with [4] who reported that reactions between nitrite and myofibrillar proteins positively affected the texture of the cured meats.

[38] reported many factors that can affect the texture of meat products such as product formulation, fat and protein content and characteristics, and other factors. In our results tenderness as well as juiciness were not significantly (P > 0.05) different, this could be due to the using of the same formula for all treatments except of the nitrite concentration. This was demonstrated by the proximate analyses that were similar to the all treatments with no significant differences, which could be reflected on the textural characteristics of the treatments. [12] found that variations in nitrite concentration had no additional effect on the mortadella texture.

Overall acceptability is the summation effect of the all sensory parameters, therefore, treatments that received high evaluation scores regarding their color, flavor, juiciness, and texture will be scored high as overall acceptability. The hedonic evaluation scores for overall acceptability corresponding to “like moderately” show that no significant (P > 0.05) difference between all treatments (Figure 5), which might indicate that different nitrite levels do not affect the overall acceptability [34] .

It seems that the panelists were unable to differentiate between treatments of the roast beef, and so, the results showed that high nitrite level of 0.026% did not cause any sensory differences. Meat is a perishable food, therefore, it is necessary to control meat spoilage in order to increase its shelf life, maintain its nutritional value and other quality and sensory parameters [39] .