请问spss当中system missing跟user defined...

请问spss当中system missing跟user defined missing在计算的过程中有什么差别吖？传说在有些计算过程中二者会不一样，大家知道具体详细的差别吗？

木有详细研究过，只知道一个是系统定义的，一个是人为定义的，一般人为定义的会把缺失值设的大点。缺失值的处理上我觉得应该不会有太多差别吧。
http://www.ats.ucla.edu/stat/spss/modules/missing.htm

System-missing values are values automatically recognized as missing by SPSS. You might notice that some of the reaction times are left blank in the data below. That is the accepted way of indicating system missing data in the data set. For example, for subject 2, the second trial is blank. The only way to read raw data with fields left blank is with fixed field input. The values left blank automatically are treated as system-missing values.

Note: It is possible to hold the missing place with a single dot in the field, but if you do you will get a warning message each time SPSS encounters one of these values. The resulting variable is coded with system-missing values.


DATA LIST FIXED/
  id 1 trial1 3-5 (1) trial2 6-8 (1)
       trial3 11-13 (1).

BEGIN DATA .
1 1.5 1.4 1.6
2 1.5     1.9
3     2.0 1.6
4         2.2
5 2.1 2.3 2.2
6 1.8 2.0 1.9
END DATA .

LIST .  
One reason for missing data might be that the equipment failed for that trial was missing.  The result of the list follows, notice that SPSS marks system-missing values with a dot in the listing. There is a dot everywhere in the listing that there was a blank in the data.


ID TRIAL1 TRIAL2 TRIAL3

1   1.5    1.0    1.6
2   1.5     .     1.9
3    .     2.0    1.6
4    .      .     2.2
5   2.1    2.0    2.2
6   1.8    2.0    1.9

Number of cases read:  6    Number of cases listed:  

User-defined missing values are numeric values that need to be defined as missing for SPSS. You might notice that some of the reaction times are -9 in the data below. You may use any value you choose to stand for a missing value, but be careful that you don't choose a value for missing that already exists for the variable in the data set. For that reason many people choose negative numbers or large numbers to represent missing values. For example, for subject 2, the second trial is -9. You may read raw data with user-missing values either as fixed field input or as free field input. We will read it as free field input in this example. When defined as such on a missing values command these values of -9 are treated as user-missing values.


DATA LIST FREE/
      id trial1 trial2 trial3 .
MISSING VALUES trial1 TO trial3 (-9).
COMPUTE  trialr1=trial1.
COMPUTE  trialr2=trial2.
COMPUTE  trialr3=trial3.
VARIABLE LABELS trial1 "Trial 1 User Miss"
                   trialr1 "Trial 1 Sys Miss".

BEGIN DATA .

1 1.5 1.4 1.6
2 1.5  -9 1.9
3  -9 2.0 1.6
4  -9  -9 2.2
5 2.1 2.3 2.2
6 1.8 2.0 1.9
END DATA .

LIST .
The compute command is used to create the new variables trialr1 through trialr3, which will contain system-missing values where there were user-defined missing values in the original variables. User-defined missing values on the original variable become system-missing values on the new variables.  The result of the list follows, notice that SPSS marks user-missing values with a -9 in the listing. There is a -9 everywhere in the listing that there was a -9 in the data, so the value of the user-defined missing is preserved for the original variables ().




  ID   TRIAL1   TRIAL2   TRIAL3  TRIALR1  TRIALR2  TRIALR3

1.00     1.50     1.40     1.60     1.50     1.40     1.60
2.00     1.50    -9.00     1.90     1.50      .       1.90
3.00    -9.00     2.00     1.60      .       2.00     1.60
4.00    -9.00    -9.00     2.20      .        .       2.20
5.00     2.10     2.30     2.20     2.10     2.30     2.20
6.00     1.80     2.00     1.90     1.80     2.00     1.90

Number of cases read:  6    Number of cases listed:  6