1. On June 14, the first native application of blockchain - "Du universe" app was launched online. This is another important action of network in the field of blockchain
it is reported that Du universe is currently recruiting content procers. The earlier you join, the greater the profit. In the future, "Du universe" will also open the third-party developer platform, introce more applications and play methods, complete value transfer and interaction through their own passes, further improve the ecological construction, and better serve users
since 2018, the network has been acting frequently in the field of blockchain. In April, it released the totem of blockchain original image service platform, which adopts the blockchain right registration network independently developed by the network to provide one-stop services for image rights confirmation, monitoring and protection. Then in May, the blockchain network operating system super chain was released. The operating system is compatible with the developer ecology of bitcoin and Ethereum. It can not only plug in and pull out the consensus mechanism to solve the current energy consumption problem, but also support 100000 concurrent transactions in a single chain
2. The relationship between Y and X is described by regression coefficient a + cm, which is a linear function of M, and C measures the moderating effect, When the independent variable is a category variable and the regulatory variable is also a category variable, the analysis of variance of two factor interaction effect is used, and the interaction effect is the regulatory effect; When the regulatory variable is continuous, the independent variable uses pseudo variable, centralizes the independent variable and regulatory variable, and does the hierarchical regression analysis of y = ax + BM + CXM + e: 1. Do the regression of y to X and m to get the measurement coefficient r 12.2, do the regression of y to x, m and XM to get r 22, if R 22 is significantly higher than R 12, then the regulatory effect is significant; When the independent variable is a continuous variable, the regulating variable is a category variable. Grouping regression: group according to the value of M, and do the regression of y to X. if the difference of regression coefficient is significant, then the regulating effect is significant. When the regulating variable is a continuous variable, do the hierarchical regression analysis of y = ax + BM + CXM + E, Independent variable is latent variable; When the moderator is a category variable, the group structural equation analysis is done. The method is to limit the regression coefficients of the two groups of structural equations to be equal, and get one χ Then we remove this restriction, re estimate the model and get another one χ 2 and the corresponding degrees of freedom χ 2 minus the following χ 2 get a new one χ 2, the degree of freedom is the difference between the degrees of freedom of the two models χ If the test result is statistically significant, the moderating effect is significant; There are many different analysis methods when both the regulatory variable and the independent variable are latent variables. The most convenient one is the unconstrained model proposed by marsh, Wen and Hau. 3. The definition of the intermediate variable is the influence of the independent variable x on the dependent variable y. if x affects y by influencing the variable m, then M is called the intermediate variable. Y = CX + E1, M = ax + E2, y = C ′ x + BM + E3, When there is only one mediating variable, there is C = C ′ + AB between the effects, and the size of the mediating effect is measured by C-C ′ = ab. 4. Mediating effect analysis method mediating effect is indirect effect, no matter whether the variables involve latent variables or not, we can use structural equation model to analyze the mediating effect, If C is not significant and y has no significant correlation with X, the mediating effect analysis will be stopped. If C is significant, the second step will be carried out; The second step is to test a and B once. If they are all significant, then test C ′, C ′, and the mediating effect is significant. If C ′ is not significant, then the complete mediating effect is significant; If at least one of a and B is not significant, do Sobel test, significant mediating effect is significant, not significant mediating effect is not significant, ^5. The comparison between moderator and mediator moderator m moderator m research purpose when does x affect y or when does x have a greater impact? How does x affect the moderating effect, interaction effect, mediating effect of y-related concepts When the influence of X on y is considered, the influence of strong x on y is strong and stable. The typical model is y = am + BM + CXM + e, M = ax + E2, y = C ′ x + BM + E3. The position of m in the model is x, M is in front of Y, M can be in front of X, the function of m after X and before y affects the direction (positive or negative) and strength of the relationship between Y and X. x influences the relationship between Y, m and X, X and X through it The correlation between M and X, y can be significant or not (the latter is ideal) the correlation between M and X, y is significant effect regression coefficient C regression coefficient proct AB effect estimate ^ C ^ A ^ B effect test whether C is equal to zero, AB is equal to zero test strategy do hierarchical regression analysis, test the significance of partial regression coefficient C (t test); 6. SPSS operation method of mediating effect and moderating effect. First, descriptive statistics, including M SD and internal consistency reliability (a) are used. Second, all variables are correlated, including statistical variables and hypothetical x, y, y, The independent variable and m should be centralized first, that is, subtracting their respective mean 1. Now, input m (regulatory variable or intermediary variable), y dependent variable, and demographic variable related to any of the independent variable, dependent variable, and M regulatory variable into independent 2 Then press next to input the independent variable of X (the intermediate variable ends here). 3. To do the analysis of the adjustment variable, we should also input the opportunity of X and m in next for further regression. The test mainly depends on whether f is significant
3. 如何用stata做中介效应检验
How to perform Sobel-Goodman mediation tests in Stata?The purpose of the Sobel-Goodman...
参考地址：http://www.360doc.com/content/16/0308/16/31392768_540527838.shtml
4. Stata is a complete and integrated statistical software that provides users with data analysis, data management and professional chart drawing. It provides many functions, including linear mixed model, balanced repetitive model and polynomial proby model. 1. Sort instruction is the sort instruction of Stata database maintenance. Figure 2. The tsset instruction is an estimation command for time series data. How to create a section data file? First transfer the data to Stata, and then use the tsset command. Tssettime, yearly (or weekly, monthly, quarterly) at this time, make sure that the name of the column of data (that is, the year) representing the time is time. The regression of time series data mainly needs to pay attention to the following points: multicollinearity (when the sample size is small, such as less than 100) and sequence correlation. Moreover, t statistic, R2 (Adj-R2), F statistic and D.W. value should be considered. First, use reg command to regress, for example: regyx1x2x3x4x5, and investigate the D.W. value (use the estatdwatson command). If the D.W. value is far away from 2 seriously, then adjust it (such as yellow shading) until it is near 2. Then, examine whether the regression result conforms to the economic meaning. If not, the, Then we should pay attention to whether it is affected by multicollinearity (judged by correlation coefficient and Vif value). When dealing with multicollinearity, we can use the method similar to dealing with cross-section data (eliminating variable method), and also look at the D.W. value. In addition, the difference method can be used to deal with multicollinearity (this method is rarely used). The command to test the DW value: estatdwatson uses the generalized difference method to consider the sequence correlation (i.e. the command to adjust the DW value): regyx1x2x4x5l. Y (l.yl2. Y can be used later); the command to consider the sequence correlation using the sequence correlation robust standard error method (i.e. the command to adjust the DW value): regyx1x2x3x4x5, In addition to the method used in the above cross-section data, the difference method can also be used to consider multicollinearity, and then the Vif value can be seen. regD.yD.x1D.x2D.x3D.x4D.x5
5. The bootstrapping mediating effect test method (5000 iterations) proposed by preacher and Hayes (2008) is used. This method provides 95% confidence interval estimation of mediating effect. If the interval estimation contains 0, it means that the mediating effect is not significant. If the interval estimation does not contain 0, it means that the mediating effect is significant< In addition, the calculation results of mediating effect show that the confidence intervals of the four effects do not include 0, so resilience has a significant mediating effect between self-esteem and coping style
extended data:
notes:
bootstrap cannot be used for many statistics, such as common nonparametric kernel regression, and some estimators whose objective function is not very smooth, such as quantitative regression, maximum score estimators, etc
In addition to the simplest sampling method of
bootstrap, there are various other sampling methods, including parametric, nonparametric, bolck and resial based. If these methods are expanded, it will be a bit complicated. If the test is to be done, different sampling methods will lead to different size and power
6. 如何用stata做中介效应检验
How to perform Sobel-Goodman mediation tests in Stata?The purpose of the Sobel-Goodman..
7. The bootstrap method of SPSS can only be carried out by stages, and regression analysis of distribution is needed. The structural equation model Amos and others can be very convenient for mediating effect Analysis on the mediating effect of bootstrap in nanxinwang
