Indonesia, which has 18,108 islands, is the largest archipelago in the world [1]. More than 250 million people live on these islands, supported by the existence of 1,200 ethnic groups [2]. With this fact, the existence of ethnic, cultural, and linguistic diversity is inevitable [3]. Indonesia accepts that and embodies it in a concept, namely Bhinneka Tunggal Ika. It is not just a slogan but a tool to unify the nation.

Bhinneka Tunggal Ika (BTI) is related to multiculturalism, which describes Indonesia as a nation with a diversity of cultures, ethnicities, races, religions, and groups. BTI is the basic concept of unity, and multiculture becomes a condition of diversity [4]. BTI must be adhered to by its citizens with an attitude of tolerance, mutual respect, justice, and mutual assistance. Although BTI is a guideline for the Indonesian nation in addressing diversity, there are still a number of conflicts that arise due to the problem of rejection of diversity. One of the biggest cases is in Sambas, West Kalimantan, between ethnic Dayaks and Madurese, which originated from socio-cultural differences that gave birth to negative attitudes and behaviours, hatred, and antipathy [5]. The casualties of the ethnic conflict were 1,189 people killed, 168 people seriously injured, 34 people slightly injured, and 3,833 houses burned and damaged. Then, 29,823 ethnic Madurese fled Kalimantan.

Diversity in society can potentially lead to conflict, such as mutual distrust and dissatisfaction with social, political, and economic issues that lead to violence. Such issues have the potential to foster hostility rooted in inequality, which increases vulnerability to civil war [6]. Therefore, the idea of unity in diversity here is not an ideological principle but a shared appreciation of togetherness and plurality. Hess [7] said that individuals need to have a sense of tolerance and mutual respect for one another because respect and acceptance of differences are important to keep diversity running harmoniously.

In the last decade, there has been a rapid increase in interest in studying attitudes towards the diversity of societies. Most of these themes are included in the study of attitudes towards multiculturalism [8]. As we know, a multicultural attitude can reduce the potential for conflict to occur by emphasizing understanding and acceptance of differences in culture, ethnicity, religion, and language [9], and especially the possibility of violence stemming from an inability to accept diversity [10, 11].

The first measurement tool that systematically measures attitudes towards multiculturalism was developed by Berry and his colleague, the Multicultural Ideology Scale, to describe Canada's policies regarding its cultural diversity [12-14]. A measuring scale that discusses various aspects of multiculturalism, such as whether diversity is good for society and whether minorities must be assimilated, is reported to have a reliability score of 0.80 [13] in a sample of Canadian citizens. When applied to Dutch citizens, it is reported that the Multicultural Ideology Scale has a reliability score of 0.82 [15], and a = 0.90 [16]. With this reliability score, the Multicultural Ideology Scale provides a reliable measurement tool for multiculturalism. Apart from the Multicultural Ideology Scale, another scale that measures attitudes towards multiculturalism is the Multicultural Attitude Scale [17]. With 28 Likert-type items and adapted [18], into 24 items with two dimensions, obtained a reliability score of a = 0.95 with a sample of Dutch citizens.

In Indonesia, it’s important to learn about the attitudes of its citizens towards diversity. However, researchers have not seen reports on measurements of attitudes towards multiculturalism, especially when making measuring scales regarding this phenomenon. In fact, as explained earlier, measuring the attitude of citizens in accepting differences is a priority. BTI as a concept can be used as a reference for the development of measuring scales.

At least there are three reasons for developing an attitude measurement tool toward cultural diversity with the concept of BTI. First, BTI is a unique concept of unity passed down by the predecessors of the Indonesian nation as a guideline for the nation and state. This concept can be used as a reference for measuring acceptance of diversity, which can not only be applied in Indonesia, but universally throughout the world. Second, with the birth of this measurement tool, policymakers, academics, researchers, activists and social observers can get an overview of people's attitudes toward diversity. Third, measuring scales that are validated and meet psychometric adequacy can later be used for data collection related to policy-making for the government or for further research related to the impact of diversity, as well as providing valuable insights for future studies.

For that reason, the researcher intends to develop attitude measurement tools for cultural diversity in Indonesia. The first step is to develop it into an Indonesian version by using the values embedded in Bhinneka Tunggal Ika. Additionally, this scale's validity and reliability were tested after its development in compliance with psychometric standards.

The developing scale will go through a number of procedures in order to produce a quality measuring scale and comply with psychometric rule standards. Following the creation of the items, the author first established a discussion group forum before moving on to the judges' evaluation of the items. The internal structure of the factors is examined in the second, and the reliability of the Binneka Attitude Scale is examined in the third.

In the judge evaluation, we get a score on items has a range of 0.72 - 0.89. The test should produce a minimum score of 0.70 [41] so that the item can be considered good or represents a factor. This shows that the items on the Bhinneka Attidute scale have fulfilled the content validity requirements as well.

After going through judge evaluation, the model analysis is continued with confirmatory factor analysis (CFA). But, it is necessary to determine whether the data that is owned meets the normality of the data. If the data meets the assumption of normality, then the linear CFA method with the maximum likelihood estimator will be used. Conversely, suppose the normality of the data is not met. In that case, the linear CFA method cannot be applied because it is not based on the assumption that the observed variables are measured on a continuum scale. Another technique that must be applied to perform the analysis is item factor analysis [42]. To find out the normality of the ordinal data that is owned by the Likert scale score, namely by looking at the skewness and kurtosis of descriptive data reports. Skewness is a measure of symmetry or the lack of symmetry of the normal distribution if the score is large. Meanwhile, kurtosis is the peak size of the distribution. The distribution is considered normal if the skewness and kurtosis are in the range between +1 and -1 [43].

However, in Table 2, the assumption of normality for almost all items is not fulfilled, or only 2 items meet the normality criteria, namely item 11 and item 4, so it is not possible to carry out an analysis using the linear CFA method. Forcing the analysis using the linear CFA method while the data obtained is not normally distributed will result in biassed estimates [44], We, therefore, need an analysis method that does not prioritize normality assumptions on categorical data, namely item factor analysis as a special case of CFA [42].

The author obtains the inter-item correlation and confirms that the average has a score of 0.50, which is unquestionably acceptable [45], which is part of confirmatory factor analysis testing. The lowest and greatest values of the inter-item correlation are 0.25 and 0.74, respectively, and discovered three factors that are theorised to be correlated with each other. The index fit model was obtained with a Chi-Square of 302.341 (149), an RMSEA of 0.53, and a probability RMSEA of 0.251. Furthermore, SRMR was obtained at 0.44, CFI at 0.96, and TLI at 0.96. The fit model above was generated after removing a number of items that made the model unfit, namely items 2, 3, 11, and 13. The correlation matrix between items forms the fit model that has been accepted and can be continued by analyzing items to see their validity.

In Table 3, it is done by looking at the amount of the item's contribution to the factor, as a item validity. Later, readers can find out which items have the biggest contribution and can be considered good by measuring attitudes towards diversity. The reference for whether an item is good or not is the t-value statistic of items that are outside the range of -1.96 to 1.96. In Table 3, the standard factor loading has a range of 0.44 - 0.85. Analysis of the items in this study showed that all items had good validity or none were in the range of -1.96 to 1.96. That is, all items are valid for measuring diversity attitudes, although each item certainly has a standard error, which is meant to measure something other than diversity attitudes. The item with the smallest factor loading is item 5, and the item with the largest factor loading is item 22.

Furthermore, in this study, the Bhinneka Attitude Scale has a reliability score of 0.94, which is the ratio between the standard factor loading and the standard factor loading plus the error score. The reliability score of 0.94 on the Bhinneka Attitude Scale is considered sufficient to say that it is a scale that has a good reliability score because it is above 0.70 [46].

This study aims to develop an attitude measurement tool for multicultural conditions in Indonesia. Bhinneka Tunggal Ika, which is one of the basic concepts of Indonesian society, is used as a reference for researchers to develop and validate attitudes towards multiculturalism. Indonesia is a multicultural country with a diversity of cultures and languages spread across thousands of islands.

The development of measuring scale begins with determining theoretical concepts, followed by making items and using two-step testing, namely, content validity and construct validity, with different samples. We believe the content and construct validity testing approaches can add quality to the results. Thus, there is less anxiety about using this scale in the next period. In the process of determining the construct validity-tested scale using confirmatory factor analysis, the researcher found a number of items that did not fit into the diversity scale. The first is item 2, “I feel uncomfortable when other people pray in ways that are different from my beliefs,” is an unfavorable item. Place item 2 adjacent to item 1, which reads, “I allow friends of other religions to pray in their own way when they visit my house”.

The word “berdoa” or pray in item 1 and continued in item 2, which is unfavorable, may make the sample's response unstable, which can affect the response pattern of the sample's answers. In addition, it is also possible for samples to experience confusion when the word “berdoa” in item 1, which is favorable, is followed by the word “pray” in item 2, which is unfavorable. Next, item 3, “I easily make friends with people of religions or faiths that are different from mine,” has the same pattern as item 4 i.e. “I feel comfortable when discussing with others who have a different understanding” that is, related to friendship relations with other people. Samples may be more comfortable and understand more about the intent of item 4 than item 3.

Likewise, in item 11, “I prioritize the interests/rights of myself and my group over the interests/rights of other individuals or groups,” with item 12, “Majority and minority groups have equal rights and status in the eyes of the law,” and item 13 “Majority groups may have more authority over minority groups” which has a case.

In this case, items that have the same intent should not be close together because it will affect the stability and understanding of the sample's answer. After removing items that indicated the creation of an unfit model, the researcher tested the fit model again and found a fit model. The model is considered appropriate to the context of the theory that the researcher traces, that is, the three factors are allowed to correlate with each other. These factors are highly correlated between gotong royong and justice 0.89, Gotong Royong with tolerance 0.92 and justice and tolerance with a correlation value of 0.95. Supposedly, this model can be applied to the high-order model, but researchers did not get the fit model from the high order model. Researchers have concluded that these three factors can indeed stand alone, which can describe the attitude of diversity.

A fit model indicates that the theoretical concept can be accepted by the data [47]. So the three factors of gotong royong, justice, and tolerance can describe the attitude of diversity. Then, the items from the three factors, which are manifestations of the theory have fulfilled sufficient statistical values to be used as items that can describe the factors to be tested.

The last, Bhinneka Tunggal Ika includes attitudes needed by individuals, such as tolerance, fairness, and mutual assistance, even if the individual is not in a plural and diverse social condition. The attitudes summarised in Bhinneka Tunggal Ika can be the basis for anticipation of religious radicalism, narrow nationalism, chauvinism, racism, and any violations based on the rejection of diversity. Indonesia, as a country that guarantees its citizens rights such as a sense of security, has campaigned for this attitude to become a motto since hundreds of years ago [48, 49]. However, what is the reference for the government and the state to determine the level of tolerance, fairness, and mutual assistance of its citizens? So that if the government, as a policymaker, sees that its citizens have a low diversity attitude, they can make policies to create programmes that guarantee their citizens respect differences again. We don't want the tragedy of the September 11, 2001, attacks on America, the Bali bombings of October 12, 2002, or the conflicts between the Madurese and Kalimantan tribes to repeat or intolerant attitudes that could destroy a country. Measurements of tolerance and justice are needed not only in Indonesia but in any country. And, in Bhinneka Tunggal Ika, there is an addition, namely gotong royong, where this attitude has a tendency to reject narrow individualism, and it becomes an obligation for its citizens to practice it [50]. Gotong-royong as an Indonesian national identity [51] can be a solution to reduce hatred because humans can have space to help each other. Gotong-royong spawned musyarawah to avoid conflict, where deliberation is part of Pancasila, which is called democracy.