Average 4 variables, but take into account missings

[torvon]

Hey there.

I want to average 4 variables that have some missings. I don't want to lose all variables with only 1
missing value. How do I tell SPSS this?

IF 4 values are present: then average 4 values (standard)
IF only 3 values are present: then average only 3.
IF less than 3 values present: code as missing.


Thanks
T

=====================
To manage your subscription to SPSSX-L, send a message to
[hidden email] (not to SPSSX-L), with no body text except the
command. To leave the list, send the command
SIGNOFF SPSSX-L
For a list of commands to manage subscriptions, send the command
INFO REFCARD

[Poes, Matthew Joseph]

You want the compute mean function.  It looks like this:
COMPUTE testv=MEAN(var1,var2).

If you want to make sure it computes the mean only when at least X number of variables are present (since this would literally give you the mean of 1 variable if all others are missing), include a .# after the mean.

Compute testv=mean.3(var1, var2, var3, var4)
In the above example you get the mean of the 4 variables, and the mean is based on the valid non-missing.  However, if a case is missing more than 1 of the variable, it will make it a sysmis.


Matthew J Poes
Research Data Specialist
Center for Prevention Research and Development
University of Illinois
510 Devonshire Dr.
Champaign, IL 61820
Phone: [phone]
email: [hidden email]


-----Original Message-----
From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of Torvon
Sent: Friday, October 05, 2012 1:29 PM
To: [hidden email]
Subject: Average 4 variables, but take into account missings

Hey there.

I want to average 4 variables that have some missings. I don't want to lose all variables with only 1 missing value. How do I tell SPSS this?

IF 4 values are present: then average 4 values (standard) IF only 3 values are present: then average only 3.
IF less than 3 values present: code as missing.


Thanks
T

=====================
To manage your subscription to SPSSX-L, send a message to [hidden email] (not to SPSSX-L), with no body text except the command. To leave the list, send the command SIGNOFF SPSSX-L For a list of commands to manage subscriptions, send the command INFO REFCARD

=====================
To manage your subscription to SPSSX-L, send a message to
[hidden email] (not to SPSSX-L), with no body text except the
command. To leave the list, send the command
SIGNOFF SPSSX-L
For a list of commands to manage subscriptions, send the command
INFO REFCARD

[Art Kendall]

something like this untested syntax should work.

compute avg_x = mean.3 (x1 to x4).
recode avg_x (sysmis = -1).
missing values avg_x (-1).
value labels avg_x
 -1 'too few values to get valid average'.

On 10/5/2012 2:29 PM, Torvon wrote:

Hey there.

I want to average 4 variables that have some missings. I don't want to lose all variables with only 1
missing value. How do I tell SPSS this?

IF 4 values are present: then average 4 values (standard)
IF only 3 values are present: then average only 3.
IF less than 3 values present: code as missing.


Thanks
T

=====================
To manage your subscription to SPSSX-L, send a message to
[hidden email] (not to SPSSX-L), with no body text except the
command. To leave the list, send the command
SIGNOFF SPSSX-L
For a list of commands to manage subscriptions, send the command
INFO REFCARD

===================== To manage your subscription to SPSSX-L, send a message to [hidden email] (not to SPSSX-L), with no body text except the command. To leave the list, send the command SIGNOFF SPSSX-L For a list of commands to manage subscriptions, send the command INFO REFCARD

[JKRockStomper]

untested

COMPUTE AVG = (V1 + V2+ V3 + V4) / 4.

IF V1 = SYSMIS(V1) AVG = (V2 + V3 + V4) / 3.
IF V2 = SYSMIS(V2) AVG = (V1 + V3 + V4) / 3.
IF V3 = SYSMIS(V3) AVG = (V2 + V1 + V4) / 3.
IF V4 = SYSMIS(V4) AVG = (V2 + V3 + V1) / 3.

IF V1 = SYSMIS(V1) AND V2 = SYSMIS(V2) AVG = SYSMIS(AVG).
IF V1 = SYSMIS(V1) AND V3 = SYSMIS(V2) AVG = SYSMIS(AVG).
IF V1 = SYSMIS(V1) AND V4 = SYSMIS(V2) AVG = SYSMIS(AVG).

IF V1 = SYSMIS(V2) AND V3 = SYSMIS(V2) AVG = SYSMIS(AVG).
IF V1 = SYSMIS(V2) AND V4 = SYSMIS(V2) AVG = SYSMIS(AVG).

IF V1 = SYSMIS(V3) AND V4 = SYSMIS(V2) AVG = SYSMIS(AVG).

[Jon K Peck]

Use the mean function and specify the minimum number of nonmissing values, .e.g.
compute avg = mean.3(v1 to v4).

(or list the variables explicitly.

Jon Peck (no "h") aka Kim
Senior Software Engineer, IBM
[hidden email]
new phone: [phone]




From:        Torvon <[hidden email]>
To:        [hidden email]
Date:        10/05/2012 12:32 PM
Subject:        [SPSSX-L] Average 4 variables, but take into account missings
Sent by:        "SPSSX(r) Discussion" <[hidden email]>

---

Hey there.

I want to average 4 variables that have some missings. I don't want to lose all variables with only 1
missing value. How do I tell SPSS this?

IF 4 values are present: then average 4 values (standard)
IF only 3 values are present: then average only 3.
IF less than 3 values present: code as missing.


Thanks
T

=====================
To manage your subscription to SPSSX-L, send a message to
[hidden email] (not to SPSSX-L), with no body text except the
command. To leave the list, send the command
SIGNOFF SPSSX-L
For a list of commands to manage subscriptions, send the command
INFO REFCARD

[David Marso]

"something like this tested syntax will work."
data list free / v1 to v4.
begin data
1 2 3 5
1 2 3 .
1 2 . .
1 . . .
. 1 2 3
. . 1 2
. . . 1
end data.
COMPUTE AVG3=MAX(-1,MEAN.3(v1 TO v4)).
MISSING VALUES AVG3 (-1).

LIST.

      V1       V2       V3       V4     AVG3

1.000000 2.000000 3.000000 5.000000 2.750000
1.000000 2.000000 3.000000  .       2.000000
1.000000 2.000000  .        .       -1.00000
1.000000  .        .        .       -1.00000
 .       1.000000 2.000000 3.000000 2.000000
 .        .       1.000000 2.000000 -1.00000
 .        .        .       1.000000 -1.00000


Number of cases read:  7    Number of cases listed:  7

Art Kendall wrote

something like this
        untested syntax should work.
       
        compute avg_x = mean.3 (x1 to x4).
        recode avg_x (sysmis = -1).
        missing values avg_x (-1).
        value labels avg_x
         -1 'too few values to get valid average'.
       
     
      Art Kendall
Social Research Consultants
      On 10/5/2012 2:29 PM, Torvon wrote:
   
   
      Hey there.

I want to average 4 variables that have some missings. I don't want to lose all variables with only 1
missing value. How do I tell SPSS this?

IF 4 values are present: then average 4 values (standard)
IF only 3 values are present: then average only 3.
IF less than 3 values present: code as missing.


Thanks
T

=====================
To manage your subscription to SPSSX-L, send a message to
[hidden email]  (not to SPSSX-L), with no body text except the
command. To leave the list, send the command
SIGNOFF SPSSX-L
For a list of commands to manage subscriptions, send the command
INFO REFCARD


   
   
 


=====================
To manage your subscription to SPSSX-L, send a message to
[hidden email] (not to SPSSX-L), with no body text except the
command. To leave the list, send the command
SIGNOFF SPSSX-L
For a list of commands to manage subscriptions, send the command
INFO REFCARD

[Bruce Weaver]

I like Art's use of a user-defined missing value here, because the reason for the "missingness" is known.  But bear in mind that depending on the range of possible values for v1 to v4, -1 might not be a good choice of that user-defined value.  The general idea is to use some clearly out of range value.  Also, 9 or some variation on it (e.g., -9, 99, 999 etc) is often used to indicate missing.  

HTH.

Art Kendall wrote

something like this
        untested syntax should work.
       
        compute avg_x = mean.3 (x1 to x4).
        recode avg_x (sysmis = -1).
        missing values avg_x (-1).
        value labels avg_x
           -1 'too few values to get valid average'.
       
     
      Art Kendall
Social Research Consultants
      On 10/5/2012 2:29 PM, Torvon wrote:
   
   
      Hey there.

I want to average 4 variables that have some missings. I don't want to lose all variables with only 1
missing value. How do I tell SPSS this?

IF 4 values are present: then average 4 values (standard)
IF only 3 values are present: then average only 3.
IF less than 3 values present: code as missing.


Thanks
T

=====================
To manage your subscription to SPSSX-L, send a message to
[hidden email]  (not to SPSSX-L), with no body text except the
command. To leave the list, send the command
SIGNOFF SPSSX-L
For a list of commands to manage subscriptions, send the command
INFO REFCARD


   
   
 


=====================
To manage your subscription to SPSSX-L, send a message to
[hidden email] (not to SPSSX-L), with no body text except the
command. To leave the list, send the command
SIGNOFF SPSSX-L
For a list of commands to manage subscriptions, send the command
INFO REFCARD

[David Marso]

"untested"???
Clearly not tested!!!  Please review the functions provided in SPSS .
MEAN, SUM ... MEAN.n etc...
---

      V1       V2       V3       V4      AVG

1.000000 2.000000 3.000000 5.000000 2.750000
1.000000 2.000000 3.000000  .        .
1.000000 2.000000  .        .        .
1.000000  .        .        .        .
 .       1.000000 2.000000 3.000000  .
 .        .       1.000000 2.000000  .
 .        .        .       1.000000  .


Number of cases read:  7    Number of cases listed:  7

JKRockStomper wrote

untested

COMPUTE AVG = (V1 + V2+ V3 + V4) / 4.

IF V1 = SYSMIS(V1) AVG = (V2 + V3 + V4) / 3.
IF V2 = SYSMIS(V2) AVG = (V1 + V3 + V4) / 3.
IF V3 = SYSMIS(V3) AVG = (V2 + V1 + V4) / 3.
IF V4 = SYSMIS(V4) AVG = (V2 + V3 + V1) / 3.

IF V1 = SYSMIS(V1) AND V2 = SYSMIS(V2) AVG = SYSMIS(AVG).
IF V1 = SYSMIS(V1) AND V3 = SYSMIS(V2) AVG = SYSMIS(AVG).
IF V1 = SYSMIS(V1) AND V4 = SYSMIS(V2) AVG = SYSMIS(AVG).

IF V1 = SYSMIS(V2) AND V3 = SYSMIS(V2) AVG = SYSMIS(AVG).
IF V1 = SYSMIS(V2) AND V4 = SYSMIS(V2) AVG = SYSMIS(AVG).

IF V1 = SYSMIS(V3) AND V4 = SYSMIS(V2) AVG = SYSMIS(AVG).

[David Marso]

I do as well. I think I will adopt -666 as my choice.
COMPUTE AVG3=MAX(-666,MEAN.3(v1 TO v4)).
MISSING VALUES AVG3 (-666).

Bruce Weaver wrote

I like Art's use of a user-defined missing value here, because the reason for the "missingness" is known.  But bear in mind that depending on the range of possible values for v1 to v4, -1 might not be a good choice of that user-defined value.  The general idea is to use some clearly out of range value.  Also, 9 or some variation on it (e.g., -9, 99, 999 etc) is often used to indicate missing.  

HTH.

Art Kendall wrote

something like this
        untested syntax should work.
       
        compute avg_x = mean.3 (x1 to x4).
        recode avg_x (sysmis = -1).
        missing values avg_x (-1).
        value labels avg_x
           -1 'too few values to get valid average'.
       
     
      Art Kendall
Social Research Consultants
      On 10/5/2012 2:29 PM, Torvon wrote:
   
   
      Hey there.

I want to average 4 variables that have some missings. I don't want to lose all variables with only 1
missing value. How do I tell SPSS this?

IF 4 values are present: then average 4 values (standard)
IF only 3 values are present: then average only 3.
IF less than 3 values present: code as missing.


Thanks
T

=====================
To manage your subscription to SPSSX-L, send a message to
[hidden email]  (not to SPSSX-L), with no body text except the
command. To leave the list, send the command
SIGNOFF SPSSX-L
For a list of commands to manage subscriptions, send the command
INFO REFCARD


   
   
 


=====================
To manage your subscription to SPSSX-L, send a message to
[hidden email] (not to SPSSX-L), with no body text except the
command. To leave the list, send the command
SIGNOFF SPSSX-L
For a list of commands to manage subscriptions, send the command
INFO REFCARD