Analysis of Consumer Preferences for Instant Mangosteen ( Garcinia mangostana L.) Powder Drink in Bogor Regency and City

The peel of mangosteen ( Garcinia mangostana L.) constitutes a larger proportion of the fruit in comparison to its flesh. The peel extract product manufactured by nutraceutical industries does not require using the fruit flesh. T o increase the shelf life and value of the product, the development of a mangosteen powder drink can be carried out. Therefore, this research aims to determine the characterization of initial raw materials and factors influencing consumer preferences for instant mangosteen powder drink to improve the acceptance of the product. The data collected were analyzed through descriptive statistics and the chi-square test. Furthermore, the chi-square test showed that demographic regions affected mangosteen fruit price, the level of preferences for the powder drink, consumption time, as well as feelings and interests of potential customers. The results of the Total Plate Count (TPC), Total Yeast and Mold Count (TYMC), and coliform microbiological test conducted on mangosteen purée raw materials met the quality requirements established by National Standardization Body (NSB) for fruit purées.


INTRODUCTION
Mangosteen (Garcinia mangostana L.) is an exotic fruit known as the "Queen of the Tropical Fruit" due to its soft texture, white flesh, and unique combination of sweet and sour flavors. According to (Iswari et al., 2006), it consists of the peel, flesh, and seed, making 70%-75%, 10%-15%, and 15% to 20% of the total fruit composition. The peel contains the highest antioxidants, specifically xanthone, which contains 107.75 mg per 100 g of fruit peel. This is used in traditional Chinese medicine to treat inflammation, diarrhea (Jamal et al., 2001), and cancer (Madury et al., 2013).
In Indonesia, the commercialization of mangosteen which is often consumed fresh, has recently focused on using the fruit peel as an herbal medicine. The fruit flesh is not optimally processed by industries and is considered a waste product. However, its processing into an instant powder drink can lead to increased value, extended shelf life, and simplified product storage.
Before processing, a raw material test on mangosteen fruit flesh must be conducted to ensure safety, particularly with respect to microbiology. Previous research developed F. Nur'azizah et al. / agriTECH, 43 (2) 2023, 141-148 142 The raw materials were prepa to obtain fresh fruit flesh as present wrapped in polyethylene plastic packa Figure 1. Fresh mangosteen flesh a mangosteen beverage product (Nur'Azizah, 2020), but consumer preferences for instant mangosteen powder drink in Bogor have never been identified. Bogor is a promising region for instant powder drink development due to the variety of instant powder drink brands available on the market and the growing popularity of such products (Habib, 2008). Therefore, this research aims to determine the microbiological characteristics of the raw materials and analyze consumer preferences for instant mangosteen powder drink in Bogor Regency and City to identify the factors influencing preferences.

Characterization of Initial Raw Materials
This research was conducted at the Food Microbiology Laboratory of the Department of Food Science and Technology, IPB University, and the Pilot Plant of South East Asian Food and Agriculture Science and Technology (SEAFAST) Center. The materials used were fruit flesh obtained from the mangosteen peel extract industry. Aquadest, Plate Count Agar (PCA) (OXOID), Potato Dextrose Agar (PDA) (OXOID), Lauryl Sulphate Tryptose Broth (LSTB) media, peptone buffer, chloramphenicol, and 70% alcohol were used for Total Plate Count (TPC), Total Yeast and Mold Count (TYMC), and coliform test.

Consumer Preferences Test
To evaluate consumer preferences primarily in Bogor Regency and City, an online questionnaire was administered using Google Forms between December 2020 and May 2021. Respondents accessed and completed the survey using a range of electronic devices such as smartphones, computers, laptops, and tablets. Statistical analysis of the data was carried out using IBM Statistics SPSS 22 and Microsoft Excel 2019.

Equipment
The equipment employed included a pulper cum finisher (Armfield FT53-A), a water bath (Memmert WNE45L4), a petri dish, a glass beaker, a measuring glass beaker, an Erlenmeyer flask, test tubes and lids, a Bunsen burner, measuring pipettes, a bulb, an incubator (Yamato model IC-102), and an autoclave (Gesellschaft für 1008).

Research Stages
This research was divided into three stages: (1) characteristics of the raw material, (2) raw material preparation, and (3) data collection on consumer preferences.

Characteristics of Initial Raw Materials
TPC, TMYC, and coliform tests were performed in accordance with Bacteriological Analytical Manual (BAM) (BAM, 2001) to determine the microbiological quality and safety of the raw materials (mangosteen purée after 12 months of storage in a freezer). The colony growth was calculated and recorded descriptively based on Standard Plate Count Agar (CFU/g or mL) and MPN table (in coliform test).

Preparation of Raw Materials
The raw materials were prepared by separating mangosteen fruit flesh and seed using a pulper to obtain fresh fruit flesh as presented in Figure 1. Up to 2 kg of mangosteen fruit flesh purée was wrapped in polyethylene plastic packaging and stored in a freezer for 12 months.

Consumer Preferences
Consumer preferences data were obtained through a consumer survey method, which was carried out virtually using the Google Forms application. The survey comprised four steps: 1) Questionnaire creation, 2) Questionnaire testing, 3) Sample size determination, and 4) Data collection. Then, the data collected were analyzed with the SPSS software.

Questionnaire Creation
The questionnaire was created with clear purposes aimed at the respondents. It was accessible from December 2020 to January 2021 and included questions related to consumer profiles and characteristics of mangosteen powder drink which affected consumer preferences. In this research, the descriptive and chi-22 software. The correlation analysis aimed to ob preferences and determine the factors affecting c analyzed with G*Power 3.1.9.7 to measure the s

Characterization of Initial Raw Materials
TPC, TYMC, and coliform tests were perfo in a freezer for 12 months. Meanwhile, the physi remained normal and in accordance with NSB appearance of frozen mangosteen purée is show  The questionnaire was subjected to both validity and reliability testing. The validity test was performed on 50 respondents, distributed according to the normal curve (Pujihastuti, 2010). This was performed by analyzing the correlations between answers to each question and the total score using SPSS software at a 5% significance level. The questions were considered valid once the r-table appeared lower than the r-count, indicating that they were correlated with the total score at a 5% significance level. In cases where the r-table was higher than the r-count, the questions were considered invalid (Alfian & Putra, 2017;Sulistiawan et al., 2018).
The reliability test was conducted using Cronbach's alpha to measure the consistency level of the research instrument. This determined the extent to which the instrument can be trusted to produce consistent scores when tested under varying conditions (Sugiyono, 2005). The questionnaire was considered reliable once its Cronbach's alpha value exceeded 0.60 (Candradewi et al., 2020;Aryanti et al., 2013;Sulistiawan et al., 2018).

Sample Size Determination
Sample size determination was accomplished through non-probabilistic sampling with the purposive sampling method. The purposive sampling method involves the determination of specific criteria that align with the objective of this research (Etikan et al., 2016). The population data were obtained from Statistics Indonesia (SI), with the population of Bogor City and Regency being 1,081,009 (BPS, 2018) and 5,715,009 people (BPS, 2017), respectively.
The required sample size for this research in Bogor City and Regency was calculated using Slovin's formula as follows: F. Nur'azizah et al. / agriTECH 43 (2) 2023, xxx-xxx es ferences data were obtained through a consumer survey method, which was sing the Google Forms application. The survey comprised four steps: 1) , 2) Questionnaire testing, 3) Sample size determination, and 4) Data collection. d were analyzed with the SPSS software.
ion ire was created with clear purposes aimed at the respondents. It was accessible to January 2021 and included questions related to consumer profiles and osteen powder drink which affected consumer preferences. g aire was subjected to both validity and reliability testing. The validity test was ndents, distributed according to the normal curve (Pujihastuti, 2010). This was g the correlations between answers to each question and the total score using significance level. The questions were considered valid once the r-table appeared , indicating that they were correlated with the total score at a 5% significance the r-table was higher than the r-count, the questions were considered invalid Sulistiawan et al., 2018). test was conducted using Cronbach's alpha to measure the consistency level of t. This determined the extent to which the instrument can be trusted to produce n tested under varying conditions (Sugiyono, 2005). The questionnaire was e its Cronbach's alpha value exceeded 0.60 (Candradewi et al., 2020;Aryanti et t al., 2018).
ination etermination was accomplished through non-probabilistic sampling with the ethod. The purposive sampling method involves the determination of specific the objective of this research (Etikan et al., 2016). The population data were s Indonesia (SI), with the population of Bogor City and Regency being 1,081,009 ,009 people (BPS, 2017), respectively. ample size for this research in Bogor City and Regency was calculated using ws: 1,081,009 + 5,715,009 1 + 6,796,018 (0.1) 2 = 99.99 ≈ 100 mple size pulation size or margin (10 %) in's formula calculation with a 10% error margin, the minimum sample size for egency was 100 respondents. The purpose of 10% error margin utilization was eriod more efficient.
ction was performed virtually for approximately one month, by distributing the various social media platforms to the respondents with suitable criteria. After m 100 respondents, the data were analyzed. For security reasons, ethical permit SM-IPB/SK/2021 was received from the Commission on Research Ethics Involving University.
Remarks: n: sample size N: population size e: error margin (10 %) Based on Slovin's formula calculation with a 10% error margin, the minimum sample size for both Bogor City and Regency was 100 respondents. The purpose of 10% error margin utilization was to make the research period more efficient.

Data Collection
The data collection was performed virtually for approximately one month, by distributing the questionnaires through various social media platforms to the respondents with suitable criteria. After obtaining responses from 100 respondents, the data were analyzed. For security reasons, ethical permit number 383/IT3.KEPMSM-IPB/SK/2021 was received from the Commission on Research Ethics Involving Human Subjects at IPB University.

Data Analysis
In this research, the descriptive and chi-square correlation analysis was conducted using SPSS 22 software. The correlation analysis aimed to observe the influence of respondent profiles on consumer preferences and determine the factors affecting consumer preferences. The survey results were further analyzed with G*Power 3.1.9.7 to measure the strength of the obtained data.

Characterization of Initial Raw Materials
TPC, TYMC, and coliform tests were performed on mangosteen fruit purée that had been stored in a freezer for 12 months. Meanwhile, the physical quality scores consisting of aroma, taste, and color remained normal and in accordance with NSB (2013) quality requirements for fruit purées. The appearance of frozen mangosteen purée is shown in Figure 9.
The limit of microbiological contaminations according to NSB (2013) for TPC, TYMC, and coliform count in fruit purées were 1 x 10 4 colony/g, 1 x 10 2 Figure 9. Frozen mangosteen purée after 12 months of storage

7.
What color preferences do you think are ideal for steeping instant mangosteen powder drink? 0.698 (g) 8.
What are the ideal taste preferences for the above-mentioned instant mangosteen powder drink product? 0.625 (h) 9.
How do you feel after consuming an instant mangosteen powder drink? 0.336 (i) 10. In your opinion, when is the best time to consume an instant mangosteen powder drink? 0.359 (j) Notes: Numbers in the same column followed by different letters are significantly different at a 5% significance level from the Pearson correlation test (valid questions) colony/g, and 20 MPN/g, respectively. The results of TPC, TYMC, and coliform count presented in Tables  4 and 5 showed that both sample repetitions fulfilled the required qualities for mangosteen purée. This implied that based on the quality scores obtained, mangosteen purée stored for 12 months in the freezer was considered acceptable and safe to be used as raw materials for instant mangosteen powder drink production.

Questionnaire Validity and Reliability Test Results
The validity of the 10-question unit was tested with a total of 50 respondents, and the r-count was found to be higher than the r-table value of 0.2732 as presented in Table 1. Also, the reliability test conducted with Cronbach's alpha showed a value higher than 0.60, namely 0.753, indicating that the used questionnaires produced reliable and consistent results. Based on the validity and reliability test results, the questionnaires were concluded as a suitable measurement instrument for this research.

Respondent Profiles
The consumer survey aimed to generally identify the respondent profiles and preferences for the instant mangosteen powder drink. Furthermore, it was conducted virtually to collect data from 100 respondents who met the specific criteria of previous mangosteen fruit consumption, namely 90 from Bogor Regency and 10 from Bogor City. The number of respondents was selected based on the population proportions of 84% from Bogor Regency and 16% from Bogor City. Demographic characteristics were the most fundamental factors in segmenting consumer groups based on the population map and simplifying the measurement of consumer preferences. These included gender, age, occupation, education level, income per month, and frequency of fruit consumption. Table 2 shows the respondent profiles in terms of demographic factors.

Factors Influencing Consumer Preferences for Instant Mangosteen Powder Drink
The chi-square test was conducted to analyze the correlation of demographic factors to consumer preferences in purchasing instant mangosteen powder drink. According to Table 3, the results showing a significance value below 0.05 indicated a significant correlation. To strengthen the research samples, GPower 3.1.9.7 was employed to test their data, with an effect size value of 0.5 and a total of 100 respondents. The G*Power test produced a Power (1-β err prob) of 0.99 (> 0.80), meaning that the samples were robust and could be trusted.

Level of Preferences for Mangosteen Fruit
Based on the chi-square test, income per month significantly affected the level of preferences for mangosteen fruit. Figure 2 shows that respondents with a monthly income < IDR 1,000,000 had preferences for mangosteen fruit either like or like very much. This was in accordance with the statement by Santosa et al. (2018) that during the peak of the harvest season, the mangosteen price became cheaper, making it seen as an affordable fruit for those with income < IDR 1,000,000.

Mangosteen Fruit Price Preferences
According to Figure 3, the chi-square analysis showed that mangosteen fruit price was influenced by the frequency of fruit consumption. Respondents who consumed fruit 1-2 times per week purchased mangosteen fruit of IDR 10,000 -IDR 15,000 and IDR 16,000 -IDR 20,000 per 1 kg (i.e. 18 and 21% of total respondents). Those who consumed 3-4 times spent up to IDR 16,000 -IDR 20,000 and IDR 21,000 -IDR 25,000 per 1 kg (10% and 13% respectively). These differences were influenced by the variations in fruit consumption frequency. The more fruit eaten, the higher the price paid compared to cases of rare consumptions.  Level of preferences for mangosteen fruit < IDR 1,000,000 IDR 1,000,000 to IDR 2,000,000 IDR 2,000,000 to IDR 3,000,000 IDR 3,000,000 to IDR 5,000,000

The Interest of Potential Customers in Instant Mangosteen Powder Drink
Based on the chi-square test results, the interest of potential customers in instant mangosteen powder drink was strongly influenced by employment status and income per month.

Employment status
The influence of professions on the interest of potential customers can be seen in Figure 4. Both employed and unemployed respondents showed interest and high interest in instant mangosteen powder drink. The unemployed were dominated by students and college students aged 21 to 25 years old. Figure 5 shows that income per month significantly influenced the interest of potential customers. Overall, respondents with a monthly income of < IDR 1,000,000 to > IDR 5,000,000 were interested in instant mangosteen powder drink, with the highest interest

The Interest of Potential Customers in Instant Mangosteen Powder Drink
Based on the chi-square test results, the interest of potential customers in instant mangosteen powder drink was strongly influenced by employment status and income per month.

Employment status
The influence of professions on the interest of potential customers can be seen in Figure 4. Both employed and unemployed respondents showed interest and high interest in instant mangosteen powder drink. The unemployed were dominated by students and college students aged 21 to 25 years old. Income per month Figure 5 shows that income per month significantly influenced the interest of potential customers. Overall, respondents with a monthly income of < Rp1.000.000 to > Rp5.000.000 were interested in instant mangosteen powder drink, with the highest interest percentage coming from those earning approximately < Rp1.000.000.
The lack of an existing instant mangosteen powder drink product in the market might have contributed to the interest. This was supported by the research of Nur'Azizah (2020) on mangosteen fruit drink formulation, as all formulations were well-liked by the panelists, with score ratings of 5.36 -5.74 out of a point scale ranging from 1 (disliked extremely) to 7 (liked extremely).

Interests in instant mangosteen powder drink
< IDR 1,000,000 IDR 1,000,000 to IDR 2,000,000 IDR 2,000,000 to IDR 3,000,000 IDR 3,000,000 to IDR 5,000,000 > IDR 5,000,000 to the tropical climate in Indonesia, which is relatively hot to very hot during the day. Therefore, daytime was considered the most appropriate for consumption in order to obtain a subsequent fresh feeling.

Ideal Taste and Color Preferences in Instant Mangosteen Powder Drink
According to Figure 8, most of the respondents preferred the sweet-sour taste and white color in instant mangosteen powder drink. Specifically, up to 71 persons selected sweet-sour taste, and 41 selected white color preferences. This choice was due to the unique sweet-sour taste and white-colored fruit flesh of fresh mangosteen. Additionally, preferences for a product could be further explored based on consumption experiences and sensations (Boesveldt et al., 2018).

Ideal Taste and Color Preferences in Instant Mangosteen Powder
According to Figure 8, most of the respondents preferred the swe in instant mangosteen powder drink. Specifically, up to 71 persons select selected white color preferences. This choice was due to the unique sweetfruit flesh of fresh mangosteen. Additionally, preferences for a product cou on consumption experiences and sensations (Boesveldt et al., 2018). to the tropical climate in Indonesia, which is relatively hot to very hot during the day. Therefore, daytime was considered the most appropriate for consumption in order to obtain a subsequent fresh feeling.

Ideal Taste and Color Preferences in Instant Mangosteen Powder Drink
According to Figure 8, most of the respondents preferred the sweet-sour taste and white color in instant mangosteen powder drink. Specifically, up to 71 persons selected sweet-sour taste, and 41 selected white color preferences. This choice was due to the unique sweet-sour taste and white-colored fruit flesh of fresh mangosteen. Additionally, preferences for a product could be further explored based on consumption experiences and sensations (Boesveldt et al., 2018).

Ideal Taste and Color Preferences in Instant Mangosteen Powder Drink
According to Figure 8, most of the respondents preferred the sweet-sour taste and white color in instant mangosteen powder drink. Specifically, up to 71 persons selected sweet-sour taste, and 41 selected white color preferences. This choice was due to the unique sweet-sour taste and white-colored fruit flesh of fresh mangosteen. Additionally, preferences for a product could be further explored based on consumption experiences and sensations (Boesveldt et al., 2018).

Ideal Taste and Color Preferences in Instant Mangosteen Powder Drink
According to Figure 8, most of the respondents preferred the sweet-sour taste and white color in instant mangosteen powder drink. Specifically, up to 71 persons selected sweet-sour taste, and 41 selected white color preferences. This choice was due to the unique sweet-sour taste and white-colored fruit flesh of fresh mangosteen. Additionally, preferences for a product could be further explored based on consumption experiences and sensations (Boesveldt et al., 2018). percentage coming from those earning approximately < IDR 1,000,000. The lack of an existing instant mangosteen powder drink product in the market might have contributed to the interest. This was supported by the research of Nur'Azizah (2020) on mangosteen fruit drink formulation, as all formulations were well-liked by the panelists, with score ratings of 5.36 -5.74 out of a point scale ranging from 1 (disliked extremely) to 7 (liked extremely).

Feelings and Consumption Time Preferences for Instant Mangosteen Powder Drink
Feeling and consumption time preferences for instant mangosteen powder drink were found to be influenced by age as indicated in Figures 6 and 7. Respondents of 21-25 years old preferred a fresh feeling and the daytime as the most suitable period to consume the product. This could be attributed to the tropical climate in Indonesia, which is relatively hot to very hot during the day. Therefore, daytime was considered the most appropriate for consumption in order to obtain a subsequent fresh feeling.

Ideal Taste and Color Preferences in Instant Mangosteen Powder Drink
According to Figure 8, most of the respondents preferred the sweet-sour taste and white color in instant mangosteen powder drink. Specifically, up to 71 persons selected sweet-sour taste, and 41 selected white color preferences. This choice was due to the unique sweet-sour taste and white-colored fruit flesh of fresh mangosteen. Additionally, preferences for a product could be further explored based on consumption experiences and sensations (Boesveldt et al., 2018).

CONCLUSION
The microbiological characteristics (TPC, TYMC, and coliform count) of mangosteen purée stored in a freezer for 12 months met NSB (2013) requirements. Additionally, consumer preferences were found to be influenced by demographic factors, namely mangosteen fruit price, level of preferences, consumption time, feelings towards instant mangosteen powder drink, and interest of potential customers.