Do People Believe in Refilled Water?

. Currently, Jambi residents still believe in refilling drinking water products made by themselves and sold freely in Jambi. Still, there is no data on the factors that cause / influence residents' high confidence in these products. This research aims to identify the factors that influence residents' confidence in repeated refilling drinking water in the city of Jambi. The research results show that the factors that affect people's confidence in refilled drinking water are; aspects of fairness, experience, empathy, ability in science, trust, attention, received energy, fulfillment, reliability in managing the refilled drinking water business. This research uses an aspect analysis procedure, with 434 respondents who live in the city of Jambi. Result in this study showed that the factors that influence people's trust in refilling drinking water in Jambi are; Fairness, Experience Four, Ability to Knowledge, Faith, Attention, Acceptance, Fulfillment, and Reliability. Meanwhile, the most substantial factor dominated by forming people's trust in refilled drinking water in the city of Jambi is the Fairness Factor.


INTRODUCTION
Water suitable for drinking is clean and healthy water, namely water that must be free from microorganisms that cause illness and chemicals that can endanger the health of human beings and other life.This situation is a business opportunity for some residents to open a smallscale refill drinking water business.Based on the regulation of the Minister of Health in Indonesia: 1405/MENKES/SK/XI/2002, the definition of clean water is water that is used for daily needs and the quality meets the health requirements of hygienic water according to applicable regulations and can be drunk when cooked (Yuki, 2013).Several sources of water are used for raw water for drinking water; 1. Seawater.2. Rainwater.3. Earth's surface water (swamp/lake water, river water, and groundwater) (Kencanawati, 2017).
In Jambi, repeated refilling of drinking water is currently being bustling again.Residents' attention to using refilled gallon water has brought business opportunities for drinking water depot entrepreneurs.Especially at this time, the business of filling drinking water depots is increasingly mushrooming in Jambi.At this time, in Jambi, there are at least 72 refill drinking water businesses that are managed by the community (Parastuti, 2018).The business activities carried out by these residents are thought to be the result of the case of drinking water which has repeatedly become a phenomenon in the city of Jambi, especially during the dry season.Fulfilling the need for hygienic water for the residents of Jambi City is still a matter to this day.Four Mayors of Jambi in the last 20 years have been unable to overcome the problem of hygienic water in this city (Saragih, n.d.) and (Saputra, 2014).With the limited availability of drinking water from the Jambi city water supply company, it is not easy.Hence, a refill drinking water business helps residents meet their water needs.
To ensure that if there are no problems in the security and safety and the confidence of the residents who consume refilled drinking water, the government has issued Regulation of the Ministry of Trade No. 651 the Year 2004 Regarding special requirements for drinking and commercial water, it has been regulated.Regarding business provisions, first, the depot must have registration with the health agency and test whether or not drinking water is appropriate for health laboratories or laboratories that have been accredited (Andika, 2014a).In addition, the depot must periodically test to ensure the quality of the water, sometimes every three months, to try its microstructure.And it is also required to test every six months regarding the analysis in the water (Andika, 2014b).Several factors that can affect the quality of water products produced are raw materials, handling of buyer containers, operator cleanliness, and depot conditions (Wandrivel et al., 2012).
The people of Jambi still have high confidence in consuming refilled drinking water.But no research reveals what factors shape people's trust in filled drinking water (Effiyaldi, 2018).Public trust like this is not easy to recognize by other parties but must be developed from the start and proven (Rofiq, 2007).Trust is a person's willingness to be sensitive to the actions of other individuals based on that other people will do something to those who believe, not binding on the ability to be vigilant (Mayer et al., 1995).Trust means the willingness to make someone responsive to the actions taken by trusted people based on trust and responsibility (Gefen & Straub, 2004).Trust is an evaluation of the relationship between one person and another who will make certain transactions based on trust in an uncertain condition (Pavlou & Gefen, 2002) (Pavlou & Gefen, 2002).
Trust is needed from honesty and cooperative behavior to share the same customs and values (Doney et al., 1998).Trust is the degree to which a person who believes conveys good behavior towards good dreams and the reliability of trusted others in uncertain and risky situations.(Das, 1998).A person's trust is claimed to be a catalyst in various transactions between sellers and buyers so that consumers get satisfaction or receive their rights as needed (Made, 2016).The exchange of goods is possible based on the confidence of personnel who have a long reach and trust (Arumsari, 2012).The dimensions of faith or elements that encourage the formation of individual belief in other parties and or individuals are divided into three components: capable, kind, and integrity (Mayer et al., 1995).In the event that a person chooses to drink tap water, bottled water, or water obtained directly from natural sources, the choice of water consumption is very central in importance.In situations where dependence places people in an uncertain situation that exposes them to potential risks, trust plays an important role in helping them make decisions in their own best interest.If people trust another person or institution in a situation that puts them at potential risk, they use a trust strategy.If they feel that the person or institution cannot be trusted, they can rely on a strategy of distrust, that is, a strategy that aims to reduce an individual's vulnerability to risk (Heimer, 2021).Knowing who one can trust, what other options are available, and when to act are important skills in society (Cook, 2001).However, it was generally found that public trust in tap drinking water was higher.certain groups are more concerned with water quality and health effects than others (Brouwer et al., 2020).
Study (Putra, 2013), CV.Satria Putra Jaya, who is engaged in the bottled drinking water business, said that based on internal factor analysis, it was revealed that the company's management is essential to pay more attention to improving the ability and quality of good human resources.Because the quality of excellent human capabilities will increase public trust for the company.A study (Aubert & Kelsey, 2000) shows an effect of trust between groups of students at two universities in Canada on the performance disparity between these groups.Studies from (Tung, Lai, 2001) and (Walczuch, 2001) intend to reveal the participation of its group members in electronic trading activities in terms of the extraordinary impact in understanding the risks in electronic trading and its advantages.Furthermore, research (Ridings et al., 2002); examines the level of trust each citizen of the virtual group has on their willingness to share and receive up-to-date information.Trust is influenced by the variable perception of responsiveness, other people's personal information, and disposition to trust.(Gefen & Straub, 2004) focuses on research on buyer trust in terms of service quality variables described by variables; tangibleness, empathy, reliability, responsiveness, and assurance.In this case, he examines the direct and indirect impact of consumer loyalty on the best college students who shop for books online in the United States on the amazon.comsite.

METHOD
This research uses a factor analysis tool.Using quantitative methods, the researchers used the people of Jambi as respondents.The population is residents who live in Jambi using population data in 2010 of 529,118 people (BPS kota Jambi, 2010).The sample size calculation was taken randomly for the residents/consumers of Jambi batik users.Based on the proportional method, the number of samples was 434 people.A questionnaire test device was used to obtain data from the origin of the research variables.Questionnaires are used to obtain data to obtain primary data and secondary data.The respondents filled out the questionnaire in the real problem to get data that describes the actual situation and conditions.The questionnaire is designed in such a way, adapted to what is experienced by each respondent.Each variable is distributed in indicators and statements according to the factors.The scale used to measure the variables is a Likert scale with five answer choices.
Openly validity is the use of tools to measure something that should be measured (Ahiri, 2016).Validity refers to meaningful, correct, valuable aspects, and according to test scores (Ahiri, 2016).Validity is a concept related to the ability of an instrument to measure something that is being measured.(Supranata, 2004).A good test instrument or device is a device with a high validity value.The test is carried out when the items selected through a mechanism item analysis have been compiled as one or use another name.The reliability test is applied to item instruments that are trusted valid only.For reliability, calculations are only used for items that are considered valid.To see whether an instrument is valid or not is to look at the results of the calculation product moments.

RESULT
Referring to the output of the validity test and instrument reliability test, it can be seen that all question items have a higher (positive) rvalue (0.480, 0.600, 0.262, 0.340), r table is 0.1650, R Cronbach's Alpha means 0.636, then r 0.1650 table.It can be concluded that the questionnaire was valid and reliable.As well as the Kindness variable, the calculated R-value was obtained (0.415, 0.500, 0.336, 0.352).Cronbach's Alpha R is 0.618, while the table-r is 0.1650.It can be concluded; The questionnaire was valid and reliable.Then, for Integrity, obtained R count (0.530, 0.490, 0.589, 0.626, 0.467, 0.570).R Cronbach's Alpha 0.928, r table is still 0.786.concluded; the questionnaire is valid and reliable.,000 Source: processed data According to Hair, Anderson, Tatham, and Black (1995) in Yamin and Kurniawan; 2009), the clarification of the KMO value is as follows;<0.9:marvelous; 0,8-0,9: meritorious; 0.7-0.8:middling; 0.6-0.7:media core; 0.5-0.6:miserable; >0.5: unacceptable.This factor analysis can generally be applied if the KMO index value exceeds 0.5.The Barttlet test is applied to test whether the correlation matrix between variables is a matrix of characteristics.This test is used to test the appropriateness of the correlation between variables in the identity matrix, which contains a diagonal matrix with a value of 1, while the other means 0.Then, the results are the hypothesis; H0: correlation matrix = feature matrix, H1: relationship matrix feature matrix.The KMO and Bartlett's Test (a) table above shows the KMO value = 0,913652 (>0.5) and the p-value of Bartlett's Test = 0.00 (<0.05), then the factor model formed is feasible to use.The table above shows the message that all MSA values > 0.5, namely, so that factor analysis can continue.

Anti Image Matrices
By looking at the results of the analysis on the outputs that have a mark in the Anti-Image Correlation column, it shows, from the twenty variables that were immediately analyzed, all of the variables had MSA values above (> 0.5), namely; .899a,0.891(a), 0.923(a), 0.929(a), 0.904(a), 0.920(a), 0.931(a), 0.916(a), 0.819(a), 0.895(a), 0.897(a), 0.924(a), 0.939(a), 0.942(a).Thus, the variables/attributes partially deserve to be analyzed.The Communalities table explains that the percentage of factors that can be formed that come from factor analysis can prove the variance of these variables.Source: processed data Derived from all the quantities in the communal table, obtained 7 (seven) variables whose communal values exceed (> 0.55).This means that only the seven attributes used have a significant correlation with the formation of the factor.From attribute two (experience), the extracted communal value is 0.589; it means that 58.9% of the formation of the variance sign factor comes from attribute variable 2 (experience).56.1% of the factors formed can show the variance of the origin of attribute 4 (ability in science), and so on.Total Variance Explained describes the % value of the variance explained by the number of factors formed.The eigenvalue gives an overview of the importance of each element in calculating the variance of the 14 variables/attributes analyzed.

Table 3 Total Variance Explained
Source: Processed data Eigenvalue values can be seen in this table in the initial Eigenvalue column (total column).The eigenvalue for attribute 1 is 5,556; the eigenvalue for a factor of two is 1.335; the Eigenvalue for a factor of three means 1.956; and so on.If we add up, the 14 Eigenvalues will be worth 14 (the same as the number of attributes).The amount of variance explained by the new factor formed if we only take one aspect is 39.689% (Cumulative Column).The amount of variance that can be explained by the new attribute/factor that is formed if we take two factors (factor 1 and factor 2), namely 49.225% (Cumulative Column) and so on.In general, the number of factors that must be taken is based on the suitability of the Eigenvalue>1, so that in this problem, we take two attributes, namely; attribute 1 (5,556).Attribute/Factor 1 shows the diversity of the data using the most significant proportion, which is 39.689% of the extraction method using factor analysis (before rotation) and using factor analysis (after rotation).then make the second factor to indicate the diversity of the initial data using a proportion of 9.537 % of the extraction technique using factor analysis (before rotation) and using factor analysis (after rotation).The variation of the previous data can be explained as much as 18.945%.The scree plot shows the correlation of each number of factors formed using the Eigenvalues in graphs.Referring to the plot image, it can be seen that when one part has been formed, the position of the curve still looks steep when it is in the 1.5 th position.Finally, through the 1.5 th position, the curve line has slowly sloped, showing the slope.And the conclusion is that in this dal two components have been formed.Source: processed data The two factors that have been formed earlier form the correlation coefficient of each variable to all of these factors.If it is seen that all the variables that correlate to each factor, it is proven that the calculated factor loading value has not been able to explain the meaning it should have.Each factor should have a significant correlation of only one factor in this condition.(it does not have a correlation < 0.5 in both factors).In conditions like this, it can be seen from variable/attribute 1 (Competence) that the magnitude of the relationship between the variable and factor one is 0.654.The relationship to the second factor is -0.190 (the minus sign only gives the direction of correlation).Consequently, there is difficulty in placing a variable into factor one or factor two.Each factor cannot be interpreted, so rotation is needed using the varimax technique.Varimax rotation is an orthogonal rotation that results in a maximum variance loading factor for each factor.In other words, the variable will only have a significant correlation and level of significance with the main factor (the relationship is close to 1), and it is clear that the relationship is weak to other factors (the relationship is close to 0).And this is what has not been achieved in the matrix component table above.In other words, the variable will have a significant correlation and level of significance with the main factor only (the relationship is close to 1), and it is a weak relationship with other factors (the relationship is close to 0).And this is what has not been achieved in the matrix component table above.In other words, the variable will have a significant correlation and level of significance with the main factor only (the relationship is close to 1), and it is clearly a weak relationship with other factors (the relationship is close to 0).And this is what has not been achieved in the matrix component table above.Referring to Table 5 the component rotation matrix is the load factor value of each variable.The load factor is the strength of the relationship between each element and variable value..3 ,623 ,189 Attribute.10 ,153 ,624 Attribute.4 ,713 ,229 Attribute.11 ,301 ,613 Attribute.5 ,702 ,129 Attribute.12 ,604 ,372 Attribute.6 ,729 ,200 Attribute.13 ,408 ,422 Attribute.7 ,708 ,270 Attribute.14 ,334 ,524 Source: processed data By excluding the factor loading code (+/-), it is also seen that each variable is only strongly related to one part of the factor (no variable is correlated < 0.5 in both aspects).Then, by referring to the table and description above, it can be explained that sequentially the factors that influence people's trust in refilled drinking water in Jambi are; Fairness, Experience, Empathy, Ability in Knowledge, Confidence, Attention, Acceptance, Fulfillment, Reliability.In Table 5, of the component rotation matrix, it is known that the attribute that has the most substantial influence on public trust in refilled drinking water in the city of Jambi is attributed 2 (Experience) with a value of = 0.755.Furthermore, to find out the factors that most influence people's trust in refilled drinking water, look at the relationship values in the primary diagonal are above 0.5, namely 0.829; -0, 559.This shows that the two factors formed are correct, causing a high correlation on the main diagonals.Concerning Table 5.18.From the Component Transformation Matrix above, it can be concluded that the most commonly controlled factor in shaping public confidence in refilled drinking water in the city of Jambi is attribute 9 (fairness) = 0.829 Source: processed data

Figure 2 Component Plot In Rotated Space
Referring to the loading factor, a plot of four dimensions is formed that appears in the chart component plot as above, where the eleven attributes are closely related to each other and are grouped in adjacent positions.On the other hand, the other three attributes move away from the eleven attributes.Factor one consists of; Competence, Experience, Institutional Approval, Ability in knowledge, Attention, empathy, and Confidence.Factor two consists of origin; acceptability, fairness, fulfillment, reliability.

CONCLUSION
Whereas the factors that influence people's trust in refilling drinking water in Jambi are; Fairness, Experience Four, Ability to Knowledge, Faith, Attention, Acceptance, Fulfillment, and Reliability.Meanwhile, the most substantial factor dominated by forming people's trust in refilled drinking water in the city of Jambi is the Fairness Factor.

Table 6
Component Transformation Matrix below;