Mplus设置starting alue
仿照老外的Mplus多层结构方程模型程序，运行后结果输出显示“CHANGE YOUR MODEL AND/OR STARTING VALUES.”，请教高人知道怎样设置这个starting value，谢谢！<br /> 原来用过EQS6.1，当出现迭代不能收敛时，也需要设定初始值，在EQS中有专门的语句可以用蒙特卡罗方法把初始值估计出来，使用者只要原样复制就OK了，我检查了哪些初始值，有正数，负数，小数，我想人工去尝试设置的话，只怕是不可能的事，就想问下有哪位高人可以指点（哪怕是指点相关的参考资料也可以）再次感谢！ <p class="reward_button"><button name="answer" value="http://bbs.pinggu.org/ture" onclick="floatwin('open_reply', 'post.php?action=reply&fid=135&tid=997117', 600, 410, '600,0');return false;">我来回答</button></p><br /> 有知道的吗？ 看看对你是否有用：<br /> If a parameter is not free by default, when the parameter is mentioned in<br /> the MODEL command, it is free at the default starting value unless<br /> another starting value is specified using the asterisk (*) followed by a<br /> number or the parameter is fixed using the @ symbol followed by a<br /> number. The exception to this is that variances and residual variances<br /> for latent response variables corresponding to categorical observed<br /> dependent variables cannot be free in the Delta parameterization. They<br /> can be free in the Theta parameterization. In the Theta parameterization,<br /> scale factors for latent response variables corresponding to categorical<br /> observed dependent variables cannot be free. They can be free in the<br /> Delta parameterization.<br /> GENERAL DEFAULTS<br /> Following are the default starting values:<br /> Means/intercepts of continuous and 0 or sample mean<br /> censored observed variables depending on the<br /> analysis<br /> Means/intercepts of count observed variables 0<br /> Thresholds of categorical observed variables 0 or determined by the<br /> sample proportions<br /> depending on the<br /> analysis<br /> Variances/residual variances of .05 or 1 depending on<br /> continuous latent variables the analysis<br /> Variances/residual variances of .5 of the sample<br /> continuous and censored observed variables variance<br /> Variances/residual variances of 1<br /> latent response variables for categorical<br /> observed variables<br /> Scale factors 1<br /> Loadings for indicators of continuous 1<br /> latent variables<br /> All other parameters 0<br /> For situations where starting values depend on the analysis, the starting<br /> values can be found using the TECH1 option of the OUTPUT command. 非常感谢你的帮助，最后一句试试看有没有用。 For situations where starting values depend on the analysis, the starting<br /> values can be found using the TECH1 option of the OUTPUT command.<br /> 就是这句看看有用吗？再次感谢你的帮助！ 请版主先把我的20个论坛币拨给sunwq2007 ，谢谢！ 好像还是不行啊！页:
[1]