ISSN 2087-3336 (Prin.
| 2721-4729 (Onlin.
TEKNOSAINS: Jurnal Sains.
Teknologi dan Informatika Vol.
No.
2, 2025, page.
http://jurnal.
id/index.
php/tekno https://doi.
org/10.
37373/tekno.
A completely randomized experimental design with 3 factors on tempe Afifah Dilla Indrawati Haryono.
Atika Nurani Ambarwati* Institut Teknologi Statistika dan Bisnis Muhammadiyah Semarang.
Jawa Tengah Ae Indonesia, 50185 * Corresponding Author: atika.
nurani@itesa.
Submitted: 25/02/2025 Revised: 01/05/2025 Accepted: 13/05/2025 ABSTRACT Tempe produced in Indonesia generally uses soybeans fermented with microorganisms as the main raw material.
This fermented food is liked by many people from various levels of society because of its taste and protein and carbohydrate content.
However, tempe can only last for 1-2 days.
The average per capita consumption of tempe per week in 2023 increased by 0.
143 kg from the previous year.
The Central Statistics Agency revealed that the supply of around 335 thousand tons of local soybeans from 2.
7 million tons of national soybean needs.
The need for soybeans for tempe reaches 70% of the total national need, further strengthening the fact that the amount of local soybeans is decreasing, so that the country must import to meet the supply of soybean needs.
In the city of Semarang, tempe is made from imported soybeans.
Course, other types of local beans are substitutes.
The demand for tempe production continues to increase, but some producers have not been able to maintain the quality and durability of tempe.
The 3-factor design helps to determine the treatment factors that affect the durability of tempe.
Based on the experiment, the durability of tempe was influenced by factors of beans, inoculum, fermentation, bean-inoculum interaction, beans-fermentation, and beans-inoculum-fermentation.
Cowpea tempe with yeast inoculum fermented for 24 hours had the best durability.
Keywords: Experimental design.
variation: tempe INTRODUCTION Tempe originated from Java Island over 300 years ago or the 16th century.
The origin of the word tempe is thought to be Old Javanese, "tumpi" refers to the white color of food made from sago flour .
In addition to the protein content in soybeans, which is equivalent to animal products, tempe is rich in important minerals, high in fiber, and low in fat, which is produced during the fermentation process .
Therefore, tempe has become a very popular dish not only in Indonesia but also abroad, especially among vegetarians.
Tempe was even included in the list of Intangible Cultural Heritage of Indonesia in 2017 .
Indonesia, as a major tempe producer, has approximately 160,000 tempe craftsmen from the Indonesian Tempe and Tofu Cooperative Association in 2022 .
This amount is expected to balance the amount of tempe consumption every day.
The average per capita consumption of tempe per week in 2023 increased by 0.
143 kg from the previous year, which was 0.
140 kg .
The main raw material for tempe, which is generally soybeans, is an important indicator in the production process to meet the needs of the tempe consumer demand.
The need for soybeans for tempe each year is 70% of the national soybean requirement.
However, the Ministry of Trade .
revealed that only around 335 thousand tons of local soybeans from 2.
7 million tons of national soybean requirements .
The scarcity that occurred was caused by the increase in the price of local soybeans necessary to import from America and other countries.
A study on tempe craftsmen in Semarang City indicated that many used imported soybeans with high expansion power as raw materials .
There is research that produces tempe from various types of beans, even local ones, with better quality.
Meanwhile, from observations on the market, the quality of tempe from some market TEKNOSAINS: Jurnal Sains.
Teknologi & Informatika is licensed under a Creative Commons Attribution-NonCommercial 4.
0 International License.
ISSN 2087-3336
(Prin.
| 2721-4729 (Onlin.
236 Atika Nurani Ambarwati.
Afifah Dilla Indrawati Haryono Completely randomized experimental design with 3 factors on tempe resistance traders is often uncertain, causing tempe to have difficulty lasting long.
In general, the shelf life of tempe is only 1-2 days at room temperature, but the lower the storage temperature, the longer the tempe will last .
Good quality tempe has a distinctive sour smell, taste, and a white color that is even on the surface.
Tempe producers say that the instability of tempe quality is due to the price of raw materials, especially soybeans, often going up and down .
The instability of raw material prices does not affect the selling price, producers are forced to reduce raw materials, production volumes, and product sizes.
As a result, the quality of tempe decreases and varies, especially in terms of The demand for tempe production continues to increase, but some producers have not been able to maintain the quality and durability of tempe.
In the study of the development of tempe mold with variations in substrate and incubation time, it turned out that the type of substrate did not have a significant effect on the total mold in tempe The best tempe inoculation material used a mixture of rice flour with an incubation period of 96 hours .
The results of Sapitri's research show that the quality of tempe with tunggal yeast is as good as that of substrate yeast with concentrations of 0.
5%, 1%, and 1.
5% .
The quality of tempe made from sword beans wrapped in banana leaves looks solid, has a distinctive tempe aroma, and is evenly white .
Research on the quality of tempe with variations in fermentation time resulted in the best quality tempe being fermented for 48 hours rather than 36 hours, or 72 hours .
The organoleptic test experiment of soybean (Glycine ma.
and red bean (Phaseolus vulgari.
tempe concluded that from 100% soybean, 100% red bean, and 80% red soybean 20%, on average, untrained analysts liked red bean tempe from the sensory test .
The experiment will focus on measuring the resistance of tempe with raw materials of soybeans, cowpeas, and sword beans after considering the quality of previous research.
Another consideration is that cowpeas and sword beans are local beans cultivated in Indonesia, so they are expected to be an alternative to the scarcity and instability of soybeans.
The inoculation material is also varied by adding wheat flour to determine the effect on each bean variation.
Variations in fermentation time will also be studied for their effect on measuring the resistance of tempe.
The use of teak leaves is useful for maintaining temperature circulation during the fermentation process due to the presence of natural spores on the surface of the leaves .
The factorial completely randomized design (CRD) method will be applied in the study because there are two different treatment factors/variations in the manufacturing process to produce quality tempe.
Factorial CRD is an experimental method to simultaneously test several factors in research.
It is expected that a valuable result from the research will be used in the development and innovation in the production of high-quality tempe in the future.
The objectives of the study include testing the effect of bean variations, inoculation materials, and fermentation time, as well as the possibility of interaction between variation factors on the durability of tempe, and knowing the differences in each variation factor that has an effect.
METHOD
Data source The study began by collecting data from a tempe-making experiment in December 2024, when the temperature in Semarang City ranged from 21oC to 33oC.
The methods for making tempe are divided into 7 types, the study used method IV.
The resulting tempe will be observed for its physical changes, especially changes in odor produced after the incubation/fermentation period at room temperature during the rainy season.
Tahap Bean seeds Peeling Washing Boiling Drain Cooling Inoculation Packaging Bean seeds Boiling Immersion Peeling Washing Drainage Inoculation Packaging Table 1.
How to make tempe Metode Bean seeds Immersion Peeling Washing Boiling Drainage Cooling Inoculation Bean seeds Boiling Immersion Peeling Washing Boiling Drainage Cooling Bean seeds Boiling Cooling Peeling Washing Immersion Boiling Drainage Bean seeds Immersion Boiling Washing Boiling Cooling Peeling Washing Bean seeds Immersion Boiling Cooling Peeling Washing Immersion Boiling ISSN 2087-3336 (Prin.
| 2721-4729 (Onlin.
DOI 10.
37373/tekno.
Tahap Incubation Tempe Metode Incubation Tempe Packaging Incubation Tempe Inoculation Packaging Incubation Tempe Cooling Inokulasi Packaging Incubation Tempe Drainage Inokulasi Packaging Inkubasi Tempe Drainage Cooling Inoculation Packaging Incubation Tempe Table 1 contains various methods for making tempe, which were introduced by Saono in 1986.
The technique used must be adjusted to the conditions and weather at the time of producing tempe .
In this study, the fourth method of manufacturing technique was used.
On average, producers in Semarang City most often use the fourth method, especially during the rainy season.
Research variables The experiment was conducted with 3-factor variables consisting of bean variations, inoculation material variations, and fermentation duration variations to measure the dependent variable.
Bean factor Regarding the raw materials themselves, it turns out that a lot of research has been carried out that along with the development of food technology, tempe can be produced using various types of beans other than soybeans that are cultivated in Indonesia, such as koro benguk, koro pedang, lamtoro, green beans, peanuts, red beans, and cowpeas.
Table 2.
Bean variation factors Types of Bean Soya bean Legumes with low production levels.
Soybeans for quality tempe raw materials have a characteristic texture with an overall yellow color but slightly brownish and round .
Cowpea The Center for Post-Harvest Agricultural Research and Development in 2006 stated that cowpeas are useful as a raw material for making tempe without having to be mixed with soybeans, so that local cowpeas have the potential to be a raw material for tempe .
Koro Sword The protein content is quite high at 30.
36%, so it is considered capable of replacing soybeans in meeting the nutritional standards of Unfortunately, the hygroscopic nature of protein concentrate causes a short shelf life and is easily damaged .
Types of Beans Usage Proportion Glycine max (L.
) Merr 1 kg Vigna 1 kg Canavalia 1 kg Table 2 explains about the beans that can be used as the main raw material for tempe.
The most frequently used soybeans will be compared in their durability with the cowpeas and sword bean, which have been used several times in experiments in making tempe.
Inoculation factors The best substrate for the growth of tempe mold is rice, which contains high water and carbohydrates as a source of nutrition.
The very high-water content of rice causes damage to the yeast more quickly, so in making yeast, it must be mixed with other substrates with low water content.
Sale argues that in addition to rice, some substrate media that can be used to obtain spores include wheat, rice, corn, or tubers .
Using too much inoculum will cause imperfect fermentation, but using too little will cause the growth of destructive bacteria .
Using too much inoculum will cause imperfect fermentation, but using too little will cause the growth of destructive bacteria.
Fermentation factors Affected by temperature, time, relative humidity (RH), and the availability of oxygen needed for yeast to grow.
The key to tempe quality is, of course, how well the inoculum spreads evenly and then 238 Atika Nurani Ambarwati.
Afifah Dilla Indrawati Haryono Completely randomized experimental design with 3 factors on tempe resistance forms mycelium on the raw material during the fermentation process.
The longer the fermentation takes place, the faster the decay phase, which reduces the quality of tempe .
Phases in the fermentation process .
at this stage, namely the rapid growth period of around 0-30 hours, the formation of mycelium on the surface of the seeds will be seen to be increasingly dense, the transition period of around 30-50 hours is the most optimal time to be marketed or consumed, and the decay period of around 50-90 hours where the growth of microorganisms stops but protein degradation continues to form ammonia.
Therefore, tempe fermentation will be tested for 24 hours and 36 hours.
Dependent variable Steps to determine the durability of tempe can be known by conducting a sensory test.
The purpose of sensory testing is to ensure more in-depth the quality of tempe, including the texture, color, and aroma of tempe.
The Indonesian National Standard stipulates that the quality of tempe in Indonesia must be characterized by a clean white that is evenly distributed on its surface, has a homogeneous and compact structure, does not fall off easily, and the taste, shape, and aroma of tempe are free from ammonia odor .
If the tempe does not have these characteristics, then the tempe is said to have rotted.
Research method Factorial completely randomized design Experimental design has three basic principles so that an experiment can be said to be valid .
Three basic principles in designing experiments include: repetition, randomization, and environmental Repetition will be done 2 times with 3 factors, so the experimental design is 3 y 2 y 2 y 2 .
otal sample .
Table 3 explains the experimental design in the factorial completely randomized design method.
Each factor goes through repetition, randomization, and combination, which is denoted by Y as resistance with a unit of time in hours.
Table 3.
Experimental design
Fermentation Bean
Inoculum
Hours Hours Koro
Yeast
Y111
Y112
Sword
Yeast Flour
Y121
Y122
Yeast
Y211
Y212
Cowpea
Yeast Flour
Y221
Y222
Soya
Yeast
Y311
Y312
Yeast Flour
Y321
Y322
Total Yij1 Yij2 Average Total Y11k Y12k Y21k Y22k Y31k Y32k Yijk In the factorial design guidebook, the procedures written include normality and homogeneity tests, variance analysis, and advanced/mean difference tests.
Normality Find out the truth of the distribution of research sample data from a population that is evenly The normalization test uses the Lilliefors method.
The hypothesis is used.
: The sample comes from a normally distributed population : The sample does not come from a normally distributed population If the sig value.
> 5%,
failed to reject the intention of the sample from a normally distributed Homogeneity Find out the truth of the research sample comes from a population that has the same variance.
The homogeneity test uses the Bartlett method with the Chi-Square test statistic.
The hypothesis is used.
Samples come from populations with the same variance O .
Samples do not come from populations with the same variance ISSN 2087-3336 (Prin.
| 2721-4729 (Onlin.
DOI 10.
37373/tekno.
If the sig value.
> 0,05 or < ( from populations with the same variance.
) , then failed to reject with the intention of sampling Analysis of Variance Correction factor calculation Oc = y!" .
Where:
Oc = Total amount of observation data # = Many repetitions a = Total of treatments on the bean factor b = Total of treatments on the inoculation factor Calculation of the total sum of squares $ % = Oc .
Oe Where:
Oc o = the number of each observation in each repetition FK = Correction factor
Calculation of the sum of squares of treatment
$ %=
Oc A Oe y* .
Where:
Oc o = the amount of each variation in the observed data Calculation of the sum of squares of interactions .
xample: interaction between beans and $ (,-) = Oc .
Oe Oe $ %(,) Oe $ %(-) .
Where:
Oc 12Oo = the number of each bean factor and inoculation in the observation data Calculation of the sum of squares of the interaction of 3 factors $ (,-.
= Oc .
Oe Oe $ %(,) Oe $ %(-) Oe $ %.
Where:
Oc 12 = the number of each row and column in the observation data Calculation of the sum of squares of errors $ 4 = $ 5 Oe $ %(,) Oe $ %(-) Oe $ %.
Oe $ (,-) Oe $ (-.
Oe $ (,.
Oe $ (,-.
Where:
JKT = Total sum of squares The calculation results are continued using Table 4, namely the table of sources of diversity/variance, which helps the analysis process with the F test.
The F test can also be done more practically with SPSS.
Table 4.
Analysis of variance Source of Diversity Fcount
JKP(A) JKP(B) JKP(C) JKI(AB) KTA
KTB
KTC
KT(AB)
KTA/KTG
KTB/KTG
KTB/KTG
KT(AB)/KTG
Ftabel 6.
" .
7" .
" :.
7" .
" .
7" .
" :.
7" .
240 Atika Nurani Ambarwati.
Afifah Dilla Indrawati Haryono
Completely randomized experimental design with 3 factors on tempe resistance Source of Diversity ABC
Error
Total
Fcount
JKI(BC) JKI(AC) JKI(ABC) JKG
JKT
KT(BC)
KT(BC)/KTG
KT(AC)
KT(AC)/KTG
KT(ABC) KT(ABC)/KTG
KTG
Ftabel 6.
" : .
7" .
7" .
" : .
7" .
The hypothesis tested is:
< O >? = >@ There was no real influence of variations in treatment factors ? O >? O >@ There is a real influence of variations in treatment factors Further test/mean difference The Coefficient of Variance (CC) is a coefficient that measures the degree of precision or accuracy, and reliability of the results of an experiment .
OoBC9 y 100% .
Where:
12 = average observation data When the KK value produced is high, it means that the degree of accuracy, and reliability is high, and the validity of the conclusion of an experiment is high.
In Table 5, the criteria for further test methods that need to be carried out are written based on the value of the coefficient of diversity Table 5.
Further test criteria
Data/KK Small
Homogen
Heterogen
< 5%
< 10 %
Medium
Large 5% Ae 10% > 10% 10% Ae 20% > 20% BNT
BNJ
Duncan
LSD
Tukey HSD
RESULTS AND DISCUSSION
The initial procedure was carried out with a normality test on the data before analyzing it by Table 6 is the result of the normality test on SPSS on the distribution of the studied tempe Table 6.
Normality test (SPSS) Tempe Resistance Mean Std.
Deviation Most Extreme Differences Absolute Positive Negative Test Statistic Asymp.
Sig.
-taile.
Monte Carlo Sig.
-taile.
Sig.
99% Confidence IntervalLower Bound Upper Bound Normal Parameters a,b The significance value of the Liliefors test results is 0.
231, which is more than 0.
05, so the data is assumed to be normally distributed for each treatment factor.
ISSN 2087-3336 (Prin.
| 2721-4729 (Onlin.
DOI 10.
37373/tekno.
Table 7.
Results of analysis of variance (SPSS) Source of Diversity Correct Model
Intercept
ABC
Error Total Fhitung Sig.
Decision 11 79871,208 7261,019 2403,648 ,000 Good Model 1 208507,042 208507,042 69023,021 ,000 Signifikan 79386,396 39693,198 13139,817 ,000 H0 rejected 51,042 51,042 16,897 ,001 H0 rejected 273,375 273,375 90,497 ,000 H0 rejected 39,396 19,698 6,521 ,012 H0 rejected 9,375 9,375 3,103 ,104 H0 accepted 84,938 42,469 14,059 ,001 H0 rejected 26,688 13,344 4,417 ,037 H0 rejected 36,250 3,021 R Square 1,000 24 288414,500 Adj.
R Square 0,999 Table 7 indicates that both variations in the type of beans, inoculum, and fermentation time effect the durability of tempe because they have a significance value below 0.
Other factors that have an effect include the interaction between variations in the type of beans and inoculum, the interaction between the type of beans and fermentation time, and the interaction between the type of beans, inoculum, and fermentation time.
Meanwhile, the interaction factor of inoculum variation and fermentation time does not affect the durability of tempe.
The GLM model that was formed was good and fit with R Square 1 and Adj.
R Square 0.
This means that the factors in the analysis to explain the dependent variable in this case, the durability of tempe, by 99.
2,413
Table 8.
Homogeneity test (SPSS) Sig.
,135
1,86%
The significance value in the homogeneity test in Table 8 proves that the variance in each treatment factor is the same or not heterogeneous.
The resulting coefficient of diversity is also worth 86% or below 5%.
The average difference test will use the Duncan test method.
Table 9.
Duncan test of bean factors Subset Bean Cowpea (A.
155,875 Soya bean (A.
Koro Sword (A.
17,125 Table 9 shows that the treatment .
of the bean factor produces tempe resistance, divided into 3 subsets.
The influence given by each treatment of factor A is significantly different from each Cowpeas were tested to be able to survive the longest among the other 3 beans for an average of 156 hours.
The resistance of soybeans is different from cowpeas and sword beans, which lasted for 107 hours.
The lowest resistance is sword bean tempe, which only lasted for 17 hours.
Table 10.
Duncan test of inoculum factors Inoculum Subset Yeast (B.
94,75 Yeast wheat (B.
91,92 Table 10 states that the effect of variation of factor B is not significantly different.
However, yeast inoculum lasted 3 hours longer than yeast flour inoculum.
Table 11.
Duncan test fermentation factors Subset Fermentation 24 Hours (C.
96,67 36 Hours (C.
242 Atika Nurani Ambarwati.
Afifah Dilla Indrawati Haryono Completely randomized experimental design with 3 factors on tempe resistance Table 11 states that the effect of variation of factor C is significantly different.
Tempe fermented for 24 hours lasts 7 hours longer than tempe fermented for 36 hours.
Table 12.
Duncan test of AB factors Subset
Interaction n A2B1
A2B2
4 154,75
A3B1
A3B2
A1B1
20,25
A1B2
Table 12 shows that the effect of the interaction of AB factors is divided into 4 subsets.
The first subset consists of cowpea tempe with yeast wheat inoculum, which lasted 3.
75 hours longer than cowpea tempe with yeast inoculum.
The second subset contains soybean tempe with yeast inoculum, which lasted 0.
5 hours longer than soybean tempe with yeast wheat inoculum.
The third subset contains sword bean tempe with yeast inoculum, and the fourth subset contains sword bean tempe with yeast wheat inoculum.
Table 13.
Duncan test of AC factors Subset
Interaction n A2C1
A2C2
A3C1
A3C2
A1C1
18,75
A1C2
Table 13 shows that the effect of the AC factor interaction is divided into 5 subsets.
Each subset indicates a significant difference from each group containing the AC interaction factor treatment.
The first subset consists of 24-hour fermented cowpea tempe.
The second subset contains 36-hour fermented cowpea tempe.
The third subset contains 24-hour fermented soybean tempe.
The fourth subset contains 36-hour fermented soybean tempe.
The fifth subset contains sword bean tempe with 24-hour and 36-hour fermentation.
Table 14.
Duncan test of ABC factors Subset
Interaction n A2B1C1 2 158,5 A2B2C1 2 158
A2B1C2 2 155,5
A2B2C2 2
A3B1C1 2
A3B2C1 2
A3B2C2 2
A3B1C2 2
A1B1C1 2
A1B1C2 2
A1B2C1 2
A1B2C2 2
Table 14 shows that the effect of the ABC factor interaction consists of 6 subsets.
Each subset indicates a significant difference from each group containing ABC interaction factor treatments.
The ISSN 2087-3336 (Prin.
| 2721-4729 (Onlin.
DOI 10.
37373/tekno.
subsets in question are subset 1 .
empe tunggal inoculated with yeast fermentation for 24 hours, tempe tunggal inoculated with yeast wheat fermentation for 24 hours, and tempe tunggal inoculated with yeast fermentation for 24 hour.
, subset 2 .
empe tunggal inoculated with yeast wheat fermentation for 36 hour.
, subset 3 .
oybean tempe inoculated with yeast wheat fermentation and yeast fermentation for 24 hour.
, subset 4 .
oybean tempe inoculated with yeast wheat fermentation and yeast fermentation for 36 hour.
, subset 5 .
oro Pedang tempe inoculated with yeast fermentation for 24 hour.
, and finally subset 6 .
oybean tempe inoculated with yeast wheat fermentation for 24 hours and 36 hour.
It turns out that cowpea tempe can last for 156 hours or about 6.
5 days.
As seen in Figure 1.
mycelium is quite evenly spread, and the color, smell, and texture also look good.
The addition of wheat flour to wheat flour does not cause a significant effect, but if the fermentation time is longer, the durability of the tempe will decrease.
The characteristics of cowpea are very similar to soybeans, and the protein content is almost the same.
Therefore, the effect of the inoculum will be more visible along with the effect of the fermentation time.
Figure 1.
Cowpea tempe Mixing flour and yeast turned out to have only a moderate effect on sword bean tempe.
Figure 2 shows that although the smell is good, the texture of the sword bean is not dense enough, and the mycelium is not spread to all sides, so that the color of the beans is still visible.
Sword bean tempe is less durable because the protein content in the sword bean is very high, affected by the carbohydrate content in the flour, thus accelerating the development of microorganisms in the yeast.
Figure 2.
Koro Sword tempe Soybean tempe is still the best in terms of texture, color, and aroma, as seen in Figure 3, while soy tempe can generally last for 107 hours or around 4.
5 days.
244 Atika Nurani Ambarwati.
Afifah Dilla Indrawati Haryono Completely randomized experimental design with 3 factors on tempe resistance Figure 3.
Soybean tempe The average durability of tempe from the experiments that have been carried out is 93.
3 hours, or almost 4 hours.
In terms of the bean variation factor, the longest-lasting tempe and the sword bean tempe cannot last long.
When viewed from the length of fermentation, tempe with a fermentation time of 24 hours can last longer than the 36-hour one.
Regarding the length of fermentation, in the first 24 hours, on average, the tempe is ready to eat, but the texture that is formed is not perfect.
In the next 12 hours, the tempe looks perfect in terms of color .
ycelium covers the entire surface of the bean.
, texture, and aroma.
Meanwhile, based on the inoculum, there is no real difference so it is said that the durability of tempe is considered the same.
From its physical form, soybean tempe is indeed better in terms of color, mycelium tightness, odor, and texture than sword bean tempe and sword bean tempe.
However, in terms of taste, the sword bean tempe tastes more-savory, has more protein, and absorbs spices better.
Tempe may have great potential as a substitute for soybeans as the main raw material for CONCLUSION Experiments conducted using 3 treatment factors indicated that variations of beans, inoculum, fermentation, interaction of beans with inoculum, interaction of beans with fermentation, and interaction of 3 factors simultaneously had a significant effect on the durability of the tempe produced.
Tempe that lasted a long time was tempe from cowpeas that was given yeast inoculum and fermented for 24 hours.
It is highly recommended to try to commercialize tempe from cowpeas because it has a long shelf life.
REFERENCE