The Effects of Dried Beer Extract in the Making of Bread Josh Beedle and Tanya Racke FN 453
Abstract: Dried Beer Extract is used in food production to create a unique and palatable flavor. This experiment analyzed the use of dried beer extract in bread and its effect on the taste, flavor and color of the resulting products. The aspiration was to create a bread of superior appearance, texture and flavor which would satisfy consumer demands for cost effective natural bread, while fulfilling corporate needs for a new innovative bread product and support the US bread production industry. Four variations of bread were tested which included a control containing no dried beer extract, a loaf containing 1.5 percent of dried beer extract, another with 3 percent dried beer extract and a forth variation containing the highest concentration of dried beer extract at 4.5 percent 1. Methods of analysis include use of a Texture Analyzer, Hunter Colorimeter and sensory evaluation performed by untrained panelists. Texture analysis revealed an increase in values regarding the hardness of bread up to the 1.5% percent concentration of dried beer extract loaf. The values then decline from the 3% loaf and there on. Overall results performed by the Hunter Colorimeter indicate a mostly positive correlation of dried beer extract content and darkening of color. Data collected from a total of 140 panelists shows a preference for the control bread in regard to appearance. The 4.5 % percent dried beer extract bread was marginally most preferred for texture over the control. The 4.5% was graded highest for flavor. At this time the findings suggest the addition of dried beer extract in a 4.5% concentration would increases flavor and darken bread color, attributes found preferable by panelists. Data collected in the experiment does not currently indicate any significant findings in regards to dried beer extract content and texture. Introduction: Bread has long been a staple of the consumer diet and for our economy. Estimated production of bread and bakery foods by the US for 2008 is over 3500 million pounds, creating over 7 billion dollars in wages 2. Market trends indicate that growth will continue to stem from consumer lifestyle changes and increased demand for products viewed to be healthier. Manufactures are striving to meet consumer demands for cost effective foods that also contain natural ingredients, such as colorings and flavors 3, through competitive product innovation. The development of dried beer extract bread could satisfy the demand of consumers for a more natural product the dried beer extract is considered a natural flavor, kosher, and would add a natural color to the bread and a need of companies for a new innovative product. Our hypothesis was that the use of dried beer extract could create bread superior in flavor, more pleasing in color and more palatable texture. It was also speculated that there would be a positive, linear correlation between the concentration of dried beer extract used and the desired characteristics. The goal of this study was to explore the effects of Dried Beer Extract in the making of bread and test the hypothesis of possible positive attributes created by the addition of Dried Beer Extract. 1 The percent of Dried Beer Extract is based off of the total weight of bread ingredients. 2 IBISWorld Industry Report. Bread Production in the US:31181. June 2008. 3 Anthony, Mark. Demand for natural colors. Food Processing. March 2008. ( v69).
Methods: The recipe used for bread development is an adaptation of a recipe from Family Favorites from Across the Country by Harley Snyder Company 4. In addition to their recipe, dry active yeast was added, as well as water. The entire list of ingredients and baking times per trial may be found in Table 1.0 at the end of this report. Four total types of bread were created, including a control bread (with no dried beer extract), a bread with a concentration of 1.5% dried beer extract, a third variation containing 3.0% dried beer extract and a bread with a 4.5% concentration of dried beer extract. To ensure consistency, the same location, oven, pans, timer, and scale were used in the production of the bread. Methods of analysis included a subjective panelist rating and two objective methods which included a Texture Analyzer and Hunter Colorimeter. In the subjective testing method, the four loaf variations were evaluated by random, untrained panelists. A total of 140 people were surveyed. The participants ranged from under 18 years of age to over 69 years, as depicted in Figure 2.0. Ninety of the people completed the taste test survey in the Sensory Evaluation Lab located in the Food Science Building. Each person was given three similarly sized pieces of bread on a plate that had been divided into three similar areas. The samples were identified only by three digit codes created by the authors. The participants tried each, evaluating the appearance texture and flavor and manually filling out the sensory score card they were given, listed in Figure 1.0. Two bread variations were evaluated by 50 participants outside of the lab, due to constraints on the availability of the Sensory Kitchen. The same Sensory Score Sheet and coding method (the unknown three digit labels) were used with these evaluators. The objective tests examined color and density. The Hunter Colorimeter was used to evaluate the different breads on the color of their crust and of the inner loaf. Each of the 4 loaves tested were made the day of testing and the loaves were tested together, using the Colorimeter found on the second floor of the Food Science Building. A similar sized cross section of each was sliced, allowing the color to be assessed of the inner and outer loaf and the L, a, and b values were documented. Three trials were done of each loaf. The texture of the bread variations was assessed using a Texture Analyzer (located in the same Food Science building). A total of five trials were performed on each bread variation on the Texture Analyzer. The probe used for the Texture Analyzer was a 1 acrylic cylinder probe (p/1) and used at a test speed of 1.7mm/s to measure force in compression. The loaves tested on the texture analyzer were the same loaves tested by the Hunter Colorimeter, although due to complications the texture data was gathered the day after the loaves had been baked. Results: The findings for the Texture Analysis of each bread variation are listed in Table 3.0. Five trials per concentration were done, with the average of the trial given at the bottom of Table 3.0. The control bread with 0% concentration of dried beer extract averaged 394 grams of pressure, the 1.5% bread loaf had an average of 550.56g, the 3% bread loaf s average was 536.92g, and the 4.5% bread variation had an average of 453.02g. Figure 3.0 graphically depicts these values for each bread loaf variation trial and their standard deviation. For the evaluation of color, three trials each were run per loaf for the L, 4 Burns, Shirley A., ed. Family Favorites From Across the Country. Vol. 2. Breakthru Marketing Services, Inc. 26-26.
a, and b values on the Hunter Colorimeter. Table 4.0 through Table 7.0 list the L, a and b values per trial for each bread variation. Figures 4.0 through 7.0 represent the L, a, b experimental findings, average values and standard deviations for each dried beer extract bread variation. The L value data shows that the control loaf had highest value (i.e. was the lightest in color) and the 4.5% had the lowest value indicating it was the darkest. The a values increased as the extract concentration increased. The b color values increased up to the 3.0% concentration, decreasing slightly at the 4.5% concentration. To assess consumer preference and attitude towards the bread variations, 140 total people taste tested and evaluated the breads. The distribution of the age of panelists is depicted in Figure 2.0, with the majority of participants (48%) falling between the ages of 19 to 28 years old. Table 2.0 displays the gender and age breakdown of all panelists. Each person was given their samples along with the Sensory Score Card which used a 9 point Hedonic Scale, shown in Figure 1.0. Overall results ranked the control loaf as having the best appearance with an average rating falling in between like very much and like moderately. Least favored by panelists for appearance was the 3% concentration sample, with an average rating falling in between like moderately and like slightly. Complete data of all values is depicted in Table 5.0. In regard to texture, the 4.5 %extract bread was the highest rated sample for texture, just beating the control loaf. Both loaves were ultimately given the rating of like moderately, but the 4.5% extract bread variation received more Like extremely evaluations and no Dislike very much which the control bread received 8 comments of as illustrated in Table5.1. The flavor of the 4.5% sample was rated the highest by evaluators, while the control variation was least liked. Discussion: Figure 3.0 illustrates the findings for the texture analysis trials of each of the bread variations. From this graph and Table 3.0 the density of each loaf is depicted. The higher the value the harder in texture each bread variation was. The 1.5% extract loaf s average value of 550.56 grams indicates this loaf is the hardest to bite into, verse the control sample bread with an average value of 394 grams which would be the softest to bite. While the original hypothesis speculated an increase of density proportional to extract concentration, the latter is shown in these results. The bread become tenderer as the dried beer extract concentration increases. Due to the positive nature of this unexpected result, further investigation of dried beer extract and its relationship to bread texture is suggested. Testing on the Colorimeter showed that the control bread had the highest L value. The L value represents the lightness of the bread, indicating the control was the lightest colored bread out of all the variations. The 1.5% extract concentration followed as the second lightest loaf, while the loaf containing 3% dried beer extract was the third lightest and the high extract concentration variation of 4.5% was the darkest. These results are understandable and follow suit with the hypothesis that as the extract amount increases, the color of the bread should as well. From the control sample to the 4.5% concentration sample, the L values slowly decrease. This can be seen by comparing values listed in Table 4.0 and Table 7.0. The results for the a color values denote a linear relationship with extract concentration as well; as the extract amount increased in each sample so does the a value. The a value is indicative of the amount
of green or red in a sample. For the beer bread loaves, each sample increased in the amount of (red) color as the extract concentration increased. The control bread, with zero percent extract in it, was assessed to have the lowest a value, or be the lightest in red. The 4.5% sample had the highest a value, or possessed the most red. The final color value tested by the Hunter colorimeter was the presence of blue and yellow, or the b value. The more positive a b value, the more yellow in color the sample is. When looking at Table 4.0, it can be seen that the control received the lowest b value of 23.56, compared to the 3.0% extract bread (shown in Table 6.0) which had a value of 25.37. The b value continued to increase with increase in extract until after 3.0%. Following the 3% concentration bread the b value decreased. The 4.5% extract concentration actually presented b values lower than the 3% loaf. These observed trends, an increase in red and yellow color as well as over all darkness, support the original hypothesis. It was suspected that the color would darken as the dried beer extract content increased, and this is confirmed by the L, a, and b experimental findings. Evaluating the appearance, texture and flavor of the bread using a sensory panel, required the creation of a sensory score card. The 9 point Hedonic Scale was chosen for a couple of reasons. With using the 9 points or categories there is no in between room for the panelists, a decision has to be made. A line scale could have been used, however, one panelist s rating of 5 could be another panelist s rating of 8. With the 9 point Hedonic Scale each panelists has the same categories to choose from, thus leaving less room for personal discrepancies. The options for rating of each sample were Like Extremely, like very much, like moderately, like neither slightly, like nor dislike, dislike slightly, dislike moderately, dislike very much or dislike extremely. This format provided ease when surveying, but not necessarily for compiling data. To quantify the appearance, texture and flavor results, number values were given for each of the categories starting with 1 for like extremely and increasing in value up to 9 for dislike extremely. These numbers were multiplied by the number of responses with in each category. Next, each category was added up for each sample, such as all the categories for the control are added up to get the total sum. This number was then divided by the total number of panelists that sampled that variation of bread. This can be seen in Tables 5.0, 5.1 and 5.2. The goal of the samples was to have the lowest average values, such as in golf, where the lowest score wins. Regarding the appearance of the samples, the control had the best average rating from the panelists with a value of 2.86, which narrowly beat out the 1.5% extract sample which had a value of 2.89. These values would put the responses in between like very much and like moderately. These can be seen in Table 5.0. The sample with 4.5% extract was rated third in likeness for appearance ahead of the 3% sample. From the results, the likeness of appearance awards went to the control sample, which was white bread, however it was very closely followed by the 1.5% extract sample. As the amount of extract increased, the likeness slightly decreased as the bread became more concentrated with extract and less airy. This can been seen by the 3 percent sample which had the most votes for dislike moderately and the 4.5 percent sample had the most votes for dislike slightly (however this sample also had the second most votes for like very much).the panelists preference for lighter colored bread contradicts the original hypotheses of darker bread color being more highly favored.
The 4.5 percent sample was the highest rated sample by the panelists for texture preference. This sample had an average rating of 3.2 and just beat out the control which had a rating of 3.25, both of which fall in the better end of like moderately. These two samples were then followed by the 1.5 percent sample with a rating of 3.33 and the 3 percent sample with a rating of 3.4. The results between the 4.5 percent and the control are interesting because those two samples were fairly close in similarity when tested with the texture analyzer as well. One would think those two samples would have contrasting textures, with so much extract added to the 4.5 percent sample. This set of data indicates that the concentration of dried beer extract has negligible effect on bread texture in grams and taste preference. The evaluation of flavor of the beer bread samples designated the 4.5 percent sample as the highest rated, or having the most flavor, with a score of 3.19. Following was the 3 percent sample with a rating of 3.27, then the 1.5 percent sample with a rating of 3.41, and finally the control sample with an average rating of 3.58. All of these bread sample fell in between the like moderatley and like slightly categories. In Table 5.2 it can be seen in the data that the more extract added to the sample resulted in a higher ranking sample on likeness which is what would be expected. The average similarity in bread rankings differs from the orginal theory of the extract concentration increasin flavor and therefore likability but the data trend of increased extract concetration resulting in higher ratings supports the hypothesis. Problems with panelists: People may have eaten something else before hand, effecting taste judgment, or may have been sick, resulting in less use of their senses. People surveyed were in two separate areas (food lab and everywhere else). Tried to make the ratings consistent with the descriptions like very much verse a 0 to 10 scale [meaning everyone s 7 could be different, but like very much leaves less room for personal discrepancies]. Other issues of concern: Oven was not right on with temperature and had to be adjusted as needed. The time of day that was available for testing left less amounts of people to perform the sensory test with sensory tests done before hand. People had to be surveyed elsewhere due to the lack of participants during scheduled times in the sensory kitchen. Leaves were found in flour during an initial test batch, possibly to help with freshness. This led to the flour being sorted through and sifted before adding to the rest of the mixture. During the test batch the dry ingredients were added to the wet ingredients vs. wet to dry. Adding the wet ingredients to dry ingredients was the process used when making other trials. One day of tasting was pushed back due to conflict with a Hot Pockets sensory test. Conclusion: Our hypothesis was that there would be a direct linear correlation of percent concentration of dried beer extract to color, flavor, and texture. To a certain extend the analysis in this experiment supported that. Texture analysis revealed that the bread becomes more as dried beer extract concentration increases. Results gained from the Hunter Colorimeter indicated an overall positive correlation of dried beer extract content and darkening of color. Unfortunately, the sensory panel evaluations showed a general disliking of the darker colored bread and the dried beer extract
concentration made little to no difference in their preferences of bread texture. From the evaluations, however, a positive relationship between extract concentration and flavor preference was demonstrated. Although the hypothesis was not entirely correct, the response to the bread was still over all positive. It is the ultimate suggestions of the authors that more trials be done with dried beer extract to conclusively determine the relationship between its usage and bread characteristics such as color, texture and flavor. Additional studies including the comparison of dried extract bread to current breads on the market are suggested. Table 1.0 Ingredient and Amounts Used for Loaf Variations and Trials Self Rising Flour Control (0% Dried Beer Extract) 1 2 3 4 5 6 3cups (384g) 3cups (384g) 3cups (384g) 3cups (384g) 3cups (384g) 3cups (384g) Sugar 2T/31g 2T/31g 2T/31g 2T/31g 2T/32g 2T/32g 12 oz 12 oz 12 oz 12 oz 11 oz 11 oz (386g) (386g) (386g) (386g) (378g) (378g) Water Dry Active 2¼ tsp 2¼ tsp 2¼ tsp 2¼ tsp 2¼ tsp 2¼ tsp Yeast (12g) (12g) (12g) (12g) (10g) (10g) Dried Beer Extract 0 0 0 0 0 0 Baking Time 45min 45min 45min 45min 45min 45min Baking Temp 350⁰F 350⁰F 350⁰F 350⁰F 350⁰F 350⁰F 1.5% Dried Beer Extract loaf 1 2 3 4 5 6 Self 384g 384g 384g 384g 384g 384g Rising Flour Sugar 30g 30g 32g 32g 32g 32g Water 376g 376g 381g 381g 378g 378g Dry Active Yeast 10g 10g 10g 10g 10g 10g Dried Beer Extract 12g 12g 12.1g 12.1g 12g 12g
Baking Time 45min 45min 45min 45min 45min 45min Baking Temp 350⁰F 350⁰F 350⁰F 350⁰F 350⁰F 350⁰F Self Rising Flour 3.0% Dried Beer Extract loaf 1 2 3 4 5 6 384g 384g 384g 384g 384g 384g Sugar 32g 32g 32g 32g 32g 32g Water 384g 384g 379g 379g 378g 378g Dry Active Yeast 10g 10g 10g 10g 10g 10g Dried 24g 24g 24g 24g 24g 24g Beer Extract Baking Time 45min 45min 45min 45min 45min 45min Baking Temp 350⁰F 350⁰F 350⁰F 350⁰F 350⁰F 350⁰F 4.5% Dried Beer Extract loaf Self Rising Flour 1 2 3 4 5 6 384g 384g 384g 384g 384g 384g Sugar 32g 32g 32g 32g 32g 32g Water 378g 378g 378g 378g 378g 378g Dry Active Yeast 10g 10g 10g 10g 10g 10g Dried 36g 36g 36g 36g 36g 36g Beer Extract Baking Time 45min 45min 45min 45min 45min 45min Baking Temp 350⁰F 350⁰F 350⁰F 350⁰F 350⁰F 350⁰F
Figure 1.0 Sensory Score Card Please indicate your gender: Beer Bread with Dried Beer Extract Male Female Please indicate your age range: 18 and under 19 28 29 38 39 48 49 58 59 68 69 and older Please rate each sample for appearance Sample No. Sample No. Sample No. Like extremely Like extremely Like extremely Like very much Like very much Like very much Like moderately Like moderately Like moderately Like slightly Like slightly Like slightly Neither like nor dislike Neither like nor dislike Neither like nor dislike Dislike slightly Dislike slightly Dislike slightly Dislike moderately Dislike moderately Dislike moderately Dislike very much Dislike very much Dislike very much Dislike extremely Dislike extremely Dislike extremely Please rate each sample for texture Sample No. Sample No. Sample No. Like extremely Like extremely Like extremely Like very much Like very much Like very much Like moderately Like moderately Like moderately Like slightly Like slightly Like slightly Neither like nor dislike Neither like nor dislike Neither like nor dislike
Dislike slightly Dislike slightly Dislike slightly Dislike moderately Dislike moderately Dislike moderately Dislike very much Dislike very much Dislike very much Dislike extremely Dislike extremely Dislike extremely Please rate each sample for flavor Sample No. Sample No. Sample No. Like extremely Like extremely Like extremely Like very much Like very much Like very much Like moderately Like moderately Like moderately Like slightly Like slightly Like slightly Neither like nor dislike Neither like nor dislike Neither like nor dislike Dislike slightly Dislike slightly Dislike slightly Dislike moderately Dislike moderately Dislike moderately Dislike very much Dislike very much Dislike very much Dislike extremely Dislike extremely Dislike extremely Please write what you like/dislike about Sample No. Please write what you like/dislike about Sample No. Please write what you like/dislike about Sample No.
Figure2.0 Participants Surveyed Table 2.0 Age and Gender Ranges of Participants Surveyed AGE Male Female Total 18 & under 5 2 7 19-28 36 31 67 29-38 15 12 27 39-48 4 11 15 49-58 6 11 17 59-68 3 3 69 & older 2 2 4 Total 71 69 140
Table 3.0 Texture Analyzer Data Control 1.5% Extract 3.0% Extract 4.5 % extract Trial 1 438.5g 417.2g 724.5g 489.2g Trial 2 265.5g 841.6g 477g 409.1g Trial 3 521.6g 541.4g 414.5g 522.7g Trial 4 446.3g 393.3g 433.2g 450.4g Trial 5 298.1g 559.3g 635.4g 393.7g Average 394g 550.56g 536.92g 453.02g Std Dev 108.0583 178.4518 136.2163 53.8957 Figure 3.0 Results of Texture Analysis Trials and Standard Deviations
Table 4.0 Hunter Colorimeter Values and Standard Deviation for Control Bread Control (0% Extract) Trials 1 2 3 Average Std Dev "L" value "a" value "b" value 68.69 66.39 69.58 68.22 1.646117 3.04 1.74 2.21 2.33 0.658255 24.3 22.43 23.95 23.56 0.994133 Figure 4.0 Control Bread Variations and Corresponding L a b Trial values and Standard Deviation Table 5.0 Hunter Colorimeter Values and Standard Deviation for Low Concentration Bread Low Trials 1 2 3 Average Std Dev "L" value 67.45 65.75 64.9 66.03 1.298396 "a" value 2.84 2.9 4.55 3.43 0.970412
"b" value 23.7 24.35 25.76 24.6 1.053107 Figure 5.0 Low Bread Variation and Corresponding L a b Trial values and Standard Deviation Table 6.0 Hunter Colorimeter Values and Standard Deviation for Medium Concentration Bread Medium Trials 1 2 3 Average Std Dev "L" value 63.69 64.44 62.38 63.5 1.042609 "a" value 4.03 3.6 4.76 4.13 0.58643 "b" value 24.29 25.34 26.48 25.37 1.095308
Figure 6.0 Medium Bread Variation and Corresponding L a b Trial values and Standard Deviation Table 6.0 Hunter Colorimeter Values and Standard Deviation for Medium Concentration Bread High Trials 1 2 3 Average Std Dev "L" value 61.8 60.7 62.83 61.78 1.065192 "a" value 4.59 5.53 4.46 4.86 0.583866 "b" value 25.21 25.37 25.01 25.2 0.18037
Figure 6.0 Medium Bread Variation and Corresponding L a b Trial values and Standard Deviation Figure 4.1 Trend lines of Average Colorimeter L a and b values for each loaf Variation Table 5.0: Appearance Data from Sensory Analysis for Dried Beer Extract Bread Variations Corresponding number value ª Sensory Score Card Options Control (0%) 1.5% extract 3.0% Extract 4.5% Extract
Number of Votes Received 1 Like extremely 6 10 6 3 2 Like very much 76 56 52 60 3 Like moderately 93 90 96 81 4 Like slightly 40 68 64 68 5 Neither like/dislike 20 45 30 35 6 Dislike slightly 30 6 36 42 7 Dislike moderately 7 0 28 0 Average Numeric Rating per Panelist 2.863158 2.894737 3.25 3.175824 Equivalent Descriptive Sensory Score Card Ratings Like moderately and in between like very much. Leaning more towards like moderately Like moderately and in between like very much. Leaning more towards like moderately Like moderately and in between like slightly. More towards like moderately Like moderately and in between like slightly. More towards like moderately ªNumbers are designated to each verbal description option by authors and used for statistical analysis * Average Rating equals the number of votes corresponding to the point rank divided by the total number of panelists whom voted.
Figure 5.0 Possible Appearance Ratings vs. Number of Ratings Received Table 5.1 Texture Data from Sensory Score Card Evaluations Corresponding number value ª Sensory Score Card Options Control (0%extract) 1.5% extract 3.0% Extract 4.5% Extract Number of Votes Received 1 Like extremely 2 6 10 6 2 Like very much 66 60 44 58 3 Like moderately 96 60 69 75 4 Like slightly 40 68 64 56 5 Neither like/dislike 30 60 50 35 6 Dislike slightly 36 42 78 30 7 Dislike moderately 28 21 0 28 8 Dislike very much 8 0 8 0 Average Numeric Rating per Panelist 3.255319 3.336842 3.4 3.2 Equivalent Descriptive Sensory Like Like In between Like
Score Card Ratings moderately and in between like slightly. Leaning more towards like moderately moderately and in between like slightly. Leaning more towards like moderately like moderately and like slightly moderately and in between like slightly. Leaning more towards like moderately ªNumbers are designated to each verbal description option by authors and used for statistical analysis * Average Rating equals the number of votes corresponding times the corresponding numeric value, divided by the total number of panelists whom voted. Figure 5.1 Sensory Score Card Rating Options vs. Number of Ratings Received for the Control, 1.5%, 3% and 4.5% Dried Beer Extract Bread Loaves for Texture Table 5.2 Data from Sensory Score Cards on Flavor of Dried Beer Extract Bread Variations Corresponding numeric value ª Sensory Score Card Options Control (0%extract) 1.5% extract 3.0% Extract 4.5% Extract Number of Votes Received 1 Like extremely 1 3 7 4 2 Like very much 42 48 54 54
3 Like moderately 105 108 84 96 4 Like slightly 76 44 56 60 5 Neither like/dislike 30 60 50 25 6 Dislike slightly 36 30 48 24 7 8 Dislike moderately 35 35 7 28 Dislike very much 16 0 8 0 Average Numeric Rating per Panelist 3.589474 3.416667 3.270833 3.197802 Equivalent Descriptive Sensory Score Card Ratings In between like moderately and like slightly. Leaning more towards like slightly In between like moderately and like slightly. Like moderately and in between like slightly. Leaning more towards like moderately Like moderately and in between like slightly. Leaning more towards like moderately ªNumbers are designated to each verbal description option by authors and used for statistical analysis * Average Rating equals the number of votes corresponding to the point rank divided by the total number of panelists whom voted.
Figure 5. Sensory Score Card Rating Options vs. Number of Ratings Received for the Control, 1.5%, 3% and 4.5% Dried Beer Extract Bread Loaves for Flavor