Creating several dummy-coded variables from a single categorical variable

[tjohnson]

Hello all

 

I am working with SPSS to recode some data before undertaking statistical analysis.  My problem is knowing the best/simplest way to recode my data.

 

The current data is categorical, with multiple response options, as shown below.  I would like to recode the data to create a new dummy-coded variable for each response option. 

 

Q816A1 (Q14) 1. How often do you use each of the following products?
		Frequency	Percent	Valid Percent	Cumulative Percent
Valid	1. Several times a day	119	5.3	11.7	11.7
	2. Once a day	126	5.6	12.4	24.1
	3. Several times a week	189	8.4	18.6	42.7
	4. Once a week	99	4.4	9.7	52.5
	5. Every 2-3 weeks	95	4.2	9.4	61.8
	6. Once a month	69	3.1	6.8	68.6
	7. A few times a year	265	11.8	26.1	94.7
	8. Less than a few times a year	54	2.4	5.3	100.0
	Total	1016	45.4	100.0
Missing	-77 System	1221	54.6
Total		2237	100.0

 

I know that I can do this manually, by specifying an IF statement for each option, but I have many variables to recode.  I am therefore wondering if anyone could help me with my syntax.

 

I have used Do Repeat before, and have been happy with the results, but I am not sure that this can be used to create several new variables from each existing variable.

 

I am sure that this can be easily done via a macro, but I have not yet mastered the art of writing SPSS macros myself!  Any help/pointers would therefore be greatly appreciated.

 

Cheers

Tim

 

Timothy Johnson | Marketing Sciences Methodologist | Harris Interactive
[phone] | [hidden email]

[Bruce Weaver]

Johnson, Timothy wrote

Hello all

I am working with SPSS to recode some data before undertaking
statistical analysis.  My problem is knowing the best/simplest way to
recode my data.

The current data is categorical, with multiple response options, as
shown below.  I would like to recode the data to create a new
dummy-coded variable for each response option.  

--- snip frequency table ---

I know that I can do this manually, by specifying an IF statement for
each option, but I have many variables to recode.  I am therefore
wondering if anyone could help me with my syntax.

 

I have used Do Repeat before, and have been happy with the results, but
I am not sure that this can be used to create several new variables from
each existing variable.

 

I am sure that this can be easily done via a macro, but I have not yet
mastered the art of writing SPSS macros myself!  Any help/pointers would
therefore be greatly appreciated.

 

Cheers

Tim

 

Timothy Johnson | Marketing Sciences Methodologist | Harris Interactive
[phone] | [e-mail]
<mailto:[e-mail]> 

You could use DO-REPEAT.  E.g.,

numeric newvar1 to newvar8 (f1.0).
do repeat v = newvar1 to newvar8 / #i = 1 to 8 .
- compute v = (oldvar EQ #i).
end repeat.
exe.


Alternatively, you could use VECTOR & LOOP:

numeric newvar1 to newvar8 (f1.0).
vector v = newvar1 to newvar8 .
loop #i = 1 to 8.
- compute v(#i) = (oldvar EQ #i).
end loop.
exe.

[Ruben Geert van den Berg]

Hi Timothy,

The basic codes for (un)dummifying are

*Testdata, single 5-point Likert.

data list free/v1.

begin data

1 2 3 4 5

end data.

*Dummify.

do rep a=v_1 to v_5/b=1 to 5.

comp a=v1=b.

end rep.

exe.

*Undummify.

do rep a=v_1 to v_5/b=1 to 5.

if a=1 v2=b.

end rep.

exe.

HTH!

Ruben van den Berg

Methodologist

TNS NIPO

E: [hidden email]

P: [phone]

I: www.tns-nipo.com

---

Date: Mon, 18 Jan 2010 17:32:25 +0000
From: [hidden email]
Subject: Creating several dummy-coded variables from a single categorical variable
To: [hidden email]

Hello all

 

I am working with SPSS to recode some data before undertaking statistical analysis.  My problem is knowing the best/simplest way to recode my data.

 

The current data is categorical, with multiple response options, as shown below.  I would like to recode the data to create a new dummy-coded variable for each response option. 

 

Q816A1 (Q14) 1. How often do you use each of the following products?
		Frequency	Percent	Valid Percent	Cumulative Percent
Valid	1. Several times a day	119	5.3	11.7	11.7
	2. Once a day	126	5.6	12.4	24.1
	3. Several times a week	189	8.4	18.6	42.7
	4. Once a week	99	4.4	9.7	52.5
	5. Every 2-3 weeks	95	4.2	9.4	61.8
	6. Once a month	69	3.1	6.8	68.6
	7. A few times a year	265	11.8	26.1	94.7
	8. Less than a few times a year	54	2.4	5.3	100.0
	Total	1016	45.4	100.0
Missing	-77 System	1221	54.6
Total		2237	100.0

 

I know that I can do this manually, by specifying an IF statement for each option, but I have many variables to recode.  I am therefore wondering if anyone could help me with my syntax.

 

I have used Do Repeat before, and have been happy with the results, but I am not sure that this can be used to create several new variables from each existing variable.

 

I am sure that this can be easily done via a macro, but I have not yet mastered the art of writing SPSS macros myself!  Any help/pointers would therefore be greatly appreciated.

 

Cheers

Tim

 

Timothy Johnson | Marketing Sciences Methodologist | Harris Interactive
[phone] | [hidden email]

---

Express yourself instantly with MSN Messenger! MSN Messenger

[Jon K Peck]

The recoding part of this problem could also be solved in a more general way with the SPSSINC CREATE DUMMIES extension command available from SPSS Developer Central (www.spss.com/devcentral).  That command, which requires the programmability plugin, discovers the values for a (categorical) variable and generates a set of dummy variables for them using a user-specified rootname.

Regards,

Jon Peck
SPSS, an IBM Company
[hidden email]
312-651-3435

From:	Ruben van den Berg <[hidden email]>
To:	[hidden email]
Date:	01/18/2010 11:17 AM
Subject:	Re: [SPSSX-L] Creating several dummy-coded variables from a single              categorical              variable
Sent by:	"SPSSX(r) Discussion" <[hidden email]>

---

Hi Timothy,

The basic codes for (un)dummifying are

*Testdata, single 5-point Likert.

data list free/v1.
begin data
1 2 3 4 5
end data.

*Dummify.

do rep a=v_1 to v_5/b=1 to 5.
comp a=v1=b.
end rep.
exe.

*Undummify.

do rep a=v_1 to v_5/b=1 to 5.
if a=1 v2=b.
end rep.
exe.

HTH!

Ruben van den Berg
Methodologist
TNS NIPO
E: ruben_van_den_berg@...
P: [phone]
I: www.tns-nipo.com

---

Date: Mon, 18 Jan 2010 17:32:25 +0000
From: [hidden email]
Subject: Creating several dummy-coded variables from a single categorical variable
To: [hidden email]

Hello all
 
I am working with SPSS to recode some data before undertaking statistical analysis.  My problem is knowing the best/simplest way to recode my data.
 
The current data is categorical, with multiple response options, as shown below.  I would like to recode the data to create a new dummy-coded variable for each response option.  
 

Q816A1 (Q14) 1. How often do you use each of the following products?
		Frequency	Percent	Valid Percent	Cumulative Percent
Valid	1. Several times a day	119	5.3	11.7	11.7
	2. Once a day	126	5.6	12.4	24.1
	3. Several times a week	189	8.4	18.6	42.7
	4. Once a week	99	4.4	9.7	52.5
	5. Every 2-3 weeks	95	4.2	9.4	61.8
	6. Once a month	69	3.1	6.8	68.6
	7. A few times a year	265	11.8	26.1	94.7
	8. Less than a few times a year	54	2.4	5.3	100.0
	Total	1016	45.4	100.0
Missing	-77 System	1221	54.6
Total		2237	100.0

 
I know that I can do this manually, by specifying an IF statement for each option, but I have many variables to recode.  I am therefore wondering if anyone could help me with my syntax.
 
I have used Do Repeat before, and have been happy with the results, but I am not sure that this can be used to create several new variables from each existing variable.
 
I am sure that this can be easily done via a macro, but I have not yet mastered the art of writing SPSS macros myself!  Any help/pointers would therefore be greatly appreciated.
 
Cheers
Tim
 
Timothy Johnson | Marketing Sciences Methodologist | Harris Interactive
[phone] | tjohnson@...

---

Express yourself instantly with MSN Messenger! MSN Messenger