Manzoni (1989) in Schoute & Wiersma (2011) has shown long-term strategic plans into short-term plans, namely the budget. When the budget has a strong influence on the company's strategic plan, it will strengthen the impression of subordinates that the strategy that must be pursued is a long-term strategic plans into short-term plans, namely the To clarify and facilitate understanding of research, it is Table |. Descriptive Statistics of Research Variables The purpose of budget use has a theoretical average value of 27 while the value of the actual range is 38.07 and the standard deviation is 5.884. The value of the actual range of at least 19 and a maximum of 45 indicates that the respondent's answer to the purpose of using the budget is in the range of 4 and 5, which means that the respondents rate high and very high about the purpose of the budget they have compiled. The actual average value above the theoretical average value indicates that the answers of the average respondent rate highly on the purpose of the budget they have compiled. The standard deviation value of 5.884 is smaller than the actual average value of 38.07 so that it does not indicate the variation of research data on the purpose of budget use. indicates that there is no variation in the research data on the variable budget participation. The purpose of budget use has a theoretical average IV. RESULT AND DISCUSSION The budget emphasis variable has a theoretical average value of 12 while the actual range value is 16.77 and the standard deviation is 2.335. The value of the actual range of it least 12 and a maximum of 20 indicates that the ‘espondent's answer to the budget suppression variable in he range 4 and 5 means that the respondent judges agree ind strongly agree on the emphasis on the budget they have compiled. The actual average value above the theoretical iverage value indicates that the answers of the average ‘espondent rate high on the budget emphasis they have compiled. The standard deviation value of 2.335 is smaller han the actual average value of 16.77 so that it indicates no variation in research data on the budget suppression variable. Figure 2. PLSAlgorithm Results Before Elimination Figure 3. PLSAlgorithm Results After Elimination 1 must be eliminated in the following figure 3: (Cross Loadings) Table 2. Discriminant Validity Figure 4. PLSAlgorithm Results After Elimination 2 eliminated and the model is_ re-estimated age Based on Figure 4 it can be concluded that the construct Next is to look at the value of AVE (Average Variance Extracted) and the root value of AVE for each variable with a correlation between variables with other variables in the model. The AVE value for each variable must be greater than 0.5. The following table 3 shows the output of the AVE value of the model. From Table 2 it can be seen that the loading value for each indicator of each latent variable has the greatest value compared to the loading value if it is associated with other latent variables. This means that every latent variable has good discriminant validity where some latent variables have values that are low correlated with other constructs. The Effect of Purpose of Budget Use on Budgetary Slack with Budget Participation and Budget Emphasis As Mediating Variable all latent variables have met the reliability test criteria so that each variable can be reliable. Tabel 5. Cronbach’s alpha dan Composite Reliabilty above, it can be seen that all variables have Cronbach's that all variables meet the criteria of Discrimant Validity. Tabel 3. Average Variance Extracted Figure 5. PLS Bootstrapping results The Effect of Purpose of Budget Use on Budgetary Slack with Budget Participation and Budget Emphasis As Mediating Variable Testing the influence of mediation uses a procedure developed by Baron and Kenny (1986) in Gozhali and Latan (9015) throuch three models. Testing of the structural model is done by looking at the Figure 7. Second Model The Effect of Purpose of Budget Use on Budgetary Slack with Budget Participation and Budget Emphasis As Mediating Variabl. In Figure 7 it can be seen that the effect of the purpose of using the budget on budget participation has a value of 4,099 which is more than 1,6682. Thus the value of the PBU t-statistics on BP is 4,099, which means that the inter- variable influence is significant with a P value of 0,000. Likewise, the effect of the purpose of using the budget on budget emphasis has a value of 9.512, which is significant because it has a value of more than 1,668. In Figure 7 it can be seen that the effect of the purpose must be below 0.05 with a 5 percent error rate or 95% confidence level. It can be seen that the results of both tests have a positive effect, so that H1 in this study was rejected. In figure 6, with direct testing between the purpose of using the budget for budgetary slack, it has a significant effect that is equal to 4,268 because it is greater than 1,668. But in table 4.10 shows that the purpose of using the budget (PBU) does not affect the budgetary slack (BS) with a t- Statistics value of 1.596. The size of the significance of the support of hypotheses can be used with a comparison of the value of t-statistics must be greater than 1,668, then the hypothesis is accepted and also seen from the P value values Table 6. Path Coefficient Third Model (Mean, STDEV, t-Values)
