5.3. Instrument

The study employed three instruments to measure the variables. For measuring the variable of national identity, the Indonesian national identity scale was used [59]. This scale consists of three aspects. The national pride aspect has four items (e.g., when I am listening to the national anthem of Indonesia, i.e., Indonesia Raya, a sense of pride would well up; Pancasila is a unique and amazing national guideline, I feel unique and proud of Indonesia), the aspect of exclusive national belongingness consisting of seven items (e.g., I will accept any conditions happening in Indonesia and still be proud of it; I feel belongingness that stems from profound love to Indonesia makes one not want to change one's nationality), and the aspect of national devotion consisting of six items (e.g., I am participating in daily communal activities and at the same time preserving the spirit of gotong royong; I am ready to become the volunteer in defending my country). The measuring instruments employed the Likert scale that consisted of five responses (highly suitable = 5, suitable = 4, somewhat suitable = 3, less suitable = 2, and not suitable = 1).

The Indonesian national identity scale has been tested by confirmatory factor analysis (CFA). The information regarding the model fitness can be evaluated using The Root Mean Square Error of Approximation (RMSEA), which had a cut-off value of 0.8 [60], the Comparative Fit Index (CFI), and the Tucker-Lewis Index (TLI), whose cut off value is 0.95 [61]. The result of the analysis showed that RMSEA = 0.06 (<0.08), CFI = 0.98 (>0.95), and TLI = 0.97 (>0.95) which indicates the measurement model fits the data. After testing the fit model, it can be interpreted as the contribution of factor loading from all items to each of the factors, and three factors are referred to as the standardized loading value.

As shown in Table 2, the 18 items dan the three factors had a significant standardized loading factor because the t-value is above 1.96. It can be concluded that all items and the three factors (National pride, Exclusive national belonging, and National devotion) have a significant contribution to measuring national identity. In addition, based on the value of the standardized loading factor, construct reliability (CR) and variance extracted (VE) were also analyzed by the Fornel and Lacker formula [62]:

Scores for reliability, validity, and CFA Standardized Loading Factor (SLF of Indonesian national identity scale).

Where e_j is the measurement error for each item or indicator [63]. From the results of the analysis it was found that CR = 0.96 and variance extracted (VE) = 0.59 which proves that The Indonesian national identity scale is adequate because it has CR > 0.7 and VE > 0.5 [63].

The study employed the perception of threat scales as the measurement instrument to measure the national threat perceptions, which the researcher designed. This instrument is named the Indonesian perception of threat scale. The instrument consisted of 10 items that can be divided into three aspects, i.e., the perception of drugs aspect, which consisted of three items, the perception of consumerism, and the perception of terrorism that each comprised four items, respectively. The perception of drugs aspect has three items (e.g., narcotics cases that are rife with worry for the nation's young generation; Drug trafficking poses a big threat to youth in Indonesia), and the aspect of the perception of consumerism consisting of four items (e.g., a life orientation that is only to obtain luxury destroys the noble values of the Indonesian nation; admiration for people who have houses, cars, and expensive clothes is not a reflection of the noble values of the nation), and the aspect of the perception of terrorism consists of three items (e.g., in my opinion, the Indonesian nation needs to continue to be vigilant against terrorists because this is a form of other countries' efforts to destroy Indonesia; acts of terrorism threaten the security and integrity of the Indonesian nation). Similar to the previous measurement tool, the perception of threat scale also used a Likert scale. It consisted of five responses (highly appropriate, appropriate, somewhat appropriate, less appropriate, and not appropriate).

Indonesian perception of threat scale has also been analyzed with CFA and found that the model fits the data because RMSEA = 0.045 (<0.08), CFI = 0.99 (>0.95), and TLI = 0.98 (>0.95). Information regarding the contribution of each item to the Indonesian perception of threat scale can be found in Table 3 below.

Standardized Loading Factor (SLF), Standard Error (SE), and t-value of Indonesians’ perception of threat scale.

Based on the information summarized in Table 3, 10 items and three factors (perception of drugs, perception of consumerism, and perception of terrorism) have a significant standardized loading factor, because the t-value is above 1.96. It can be concluded that, all items and the three factors have a significant contribution in measuring Indonesian perception of threat. In addition, it was found that construct reliability = 0.96 (>07) and variance extracted = 0.59 (>0.5) which indicates that the Indonesian perception of threat is adequate.

In contrast, the collective self-esteem variable is measured by the collective self-esteem scale compiled by [49] and its scale has been adapted to the Indonesian context. The scale consists of four aspects, namely the membership self-esteem, the private collective self-esteem, the public collective self-esteem, and their importance to an identity. The membership self-esteem aspect has four items (e.g., I am an important part of the Indonesian nation; I feel I have no contribution to the Indonesian nation), and the aspect of private collective self-esteem consists of four items (e.g., I feel proud of Indonesia as my homeland; I feel that Indonesia as my country is not useful for me personally), the public collective self-esteem aspect has four items (in general, people rate my country positively; many people think that my country is not functioning well compared to other countries), the aspect of importance to identity consisting of four items (eg. In general, identity as an Indonesian is an important part of my self-image; my identity as an Indonesian citizen has little effect on my assessment of my personality).

Similar to the two previous measurement instruments, the collective self-esteem scale has also been analyzed by CFA and it was found that the model fits the data RMSEA = 0.065 (<0.08), CFI = 0.98 (>0.95), and TLI = 0.97 (>0.95). Furthermore, it can be tested the contribution of each item to its factors and the contribution of the four factors (membership self-esteem, private collective self-esteem, public collective self-esteem, and importance to identity) on collective self-esteem. Interpretation is based on those in the following table.

Based on the information summarized in Table 4, 16 items and four factors of the collective self-esteem scale have a significant standardized loading factor, because the t-value is above 1.96. It can be concluded that all items and the four factors (perception of narcotics, perception of consumerism, and perception of terrorism) had a significant contribution to measuring Indonesian perception of threat. In addition, it was found that construct reliability (CR) = 0.97 (>07) and variance extracted (VE) = 0.67 (>0.5) which indicates that the collective self-esteem scale is adequate.

Standardized Loading Factor (SLF), Standard Error (SE), and t-value of collective self-esteem scale.

Based on the loading factor information, testing of discriminant validity can also be performed. The results of the test can be presented in Table 5 as follows.

Discriminant validity.

Notes: The square root of the AVE is shown in italics on the diagonal of the matrix. The value below the diagonal is the correlation coefficient between variables.

Looking at the discriminant validity matrix in Table 5, the square root of AVE on Natid is 0.71, perception of threat is 0.73, and collective self-esteem is 0.82. They all had greater value than the value of the correlation between latent variables, which indicates that all research variables had met the discriminant validity measure.