total for tables-two diff categories

[Rodrigo Briceño]

Dear SPSS users. I would like your gently help about the following issue.
I'm preparing a table in PASW18 that combines by sex two different questions of a survey.
Question 1 asks if the person has ever attended school and question 2 asks if the person who answered yes is Currently in the school system. The thing is that I must combine answers from the two questions in the same table, so my output will read: Never attended, currently attending, ever attended but not currently attending.

Although presentation of these results is quite easy, my question to the list is about the possibility of getting a total of the three categories. I have tried used the custom tables GUI, but it seems like I only can put a total at the bottom of each variable (not to sum the three categories of interest).

Do anybody know if this is possible to be defined?
Thanks in advance

[Melissa Ives]

Rodrigo,

 

Seems to me the simplest thing would be to create a new variable (in syntax).

 

The following assumes that yes is coded 1 and No is coded 0 for Q1 (ever attended) and Q2 (currently attending).

(I'm going to assume that you have already confirmed that there are none who said they never attended but are currently attending.)

 

if (Q1=0) schstat=0.

if (Q1=1 and Q2=1) schstat=1.

if (Q1=1 and Q2=0) schstat=2.

var labels schstat 'School status'.

value labels  

  0 'Never attended'

  1 'Currently Attending'

  2 'Past Attendance (Ever but not current)'.

 

Then use this variable in your analysis.

 

Melissa

---

From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of Rodrigo Briceño
Sent: Thursday, April 15, 2010 4:57 PM
To: [hidden email]
Subject: [SPSSX-L] total for tables-two diff categories

Dear SPSS users. I would like your gently help about the following issue.
I'm preparing a table in PASW18 that combines by sex two different questions of a survey.
Question 1 asks if the person has ever attended school and question 2 asks if the person who answered yes is Currently in the school system. The thing is that I must combine answers from the two questions in the same table, so my output will read: Never attended, currently attending, ever attended but not currently attending.

Although presentation of these results is quite easy, my question to the list is about the possibility of getting a total of the three categories. I have tried used the custom tables GUI, but it seems like I only can put a total at the bottom of each variable (not to sum the three categories of interest).

Do anybody know if this is possible to be defined?
Thanks in advance

---

PRIVILEGED AND CONFIDENTIAL INFORMATION
This transmittal and any attachments may contain PRIVILEGED AND
CONFIDENTIAL information and is intended only for the use of the
addressee. If you are not the designated recipient, or an employee
or agent authorized to deliver such transmittals to the designated
recipient, you are hereby notified that any dissemination,
copying or publication of this transmittal is strictly prohibited. If
you have received this transmittal in error, please notify us
immediately by replying to the sender and delete this copy from your
system. You may also call us at [phone] for assistance.

[John F Hall]

Assuming:

 

q1 coded 1 = yes, 2 = no

q2 coded  1 = yes, 2 = no

 

you need something like (untested, and it's very late here)

 

compute newvar = q1 * 10 + q2.

recode newvar (20 =1)(11=1)(12 = 3)

val lab newvar 1 'Never attended'

                            2 'Currently attending'

                            3 'Not attending' 

                        

be careful of missing values!

----- Original Message -----

From: [hidden email]

To: [hidden email]

Sent: Thursday, April 15, 2010 11:56 PM

Subject: total for tables-two diff categories

Dear SPSS users. I would like your gently help about the following issue.
I'm preparing a table in PASW18 that combines by sex two different questions of a survey.
Question 1 asks if the person has ever attended school and question 2 asks if the person who answered yes is Currently in the school system. The thing is that I must combine answers from the two questions in the same table, so my output will read: Never attended, currently attending, ever attended but not currently attending.

Although presentation of these results is quite easy, my question to the list is about the possibility of getting a total of the three categories. I have tried used the custom tables GUI, but it seems like I only can put a total at the bottom of each variable (not to sum the three categories of interest).

Do anybody know if this is possible to be defined?
Thanks in advance

[John F Hall]



Rodriguo

 

Have you checked your data first, to make sure there are no illogical combinations?

 

cros q1 by q2 .

 

Melissa's solution using if uses much more computer resources than recode so I've suggested an amendment to her syntax.  Instead of: 

 

if (Q1=0) schstat=0.

if (Q1=1 and Q2=1) schstat=1.

if (Q1=1 and Q2=0) schstat=2.

 

you can use

 

compute schstat = q1 + q2 * 10.

recode schstat (0 = 1)(11 = 2)(10 = 3).

var labels schstat 'School status'.

value labels  

  1 'Never attended'

  2 'Currently Attending'

  3 'Past Attendance (Ever but not current)'.

 

I'm never happy about using 0 as a code, so would always use 1 = yes, 2 = no, but my solution should produce the same result.

 

compute newvar = q1 * 10 + q2 .

recode newvar (20 =1)(11=1)(12 = 3) .

val lab newvar 1 'Never attended'

                            2 'Currently attending'

                            3 'Not attending' .

 

Here's real life example I dug out this morning.

 

London Health for All

 

Secondary analysis of data on diet and health from survey of Health and Life-styles by Bloomsbury and  Islington  Health      Authority  (1993)  See: Steve Griffiths with John Hall  The Lie of the Land for the Health of the Nation: an analysis of  the  Bloomsbury  and Islington Health  &  Lifestyle  survey.       Report to Bloomsbury and Islington AHA, May 1993

 

The data set included two variables, v670 (Responsibility for children) and v715 (Marital Status)  The client wanted a single variable combining these into 4 categories:

 

Single, no responsibilty for children

Couple, no responsibility for children

Single with responsibilty for children

Couple with responsibilty for children

 

Step 1:

 

freq v670 v715.
cro v715 by v670.

 

RESPONSIBLE FOR CHILDREN
		Frequency	Percent	Valid Percent	Cumulative Percent
Valid	Yes	432	30.6	30.6	30.6
	No	953	67.5	67.5	98.2
	3	26	1.8	1.8	100.0
	Total	1411	100.0	100.0

 

 

 

MARITAL STATUS
		Frequency	Percent	Valid Percent	Cumulative Percent
Valid	SINGLE	526	37.3	37.5	37.5
	MARRIED	467	33.1	33.3	70.8
	COHABITING	73	5.2	5.2	76.0
	SEPARATED	50	3.5	3.6	79.5
	DIVORCED	122	8.6	8.7	88.2
	WIDOWED	160	11.3	11.4	99.6
	OTHER	5	.4	.4	100.0
	Total	1403	99.4	100.0
Missing	System	8	.6
Total		1411	100.0

 

 

MARITAL STATUS * RESPONSIBLE FOR CHILDREN Crosstabulation
Count
		RESPONSIBLE FOR CHILDREN			Total
		Yes	No	3
MARITAL STATUS	SINGLE	88	422	16	526
	MARRIED	234	228	5	467
	COHABITING	30	43	0	73
	SEPARATED	23	27	0	50
	DIVORCED	46	75	1	122
	WIDOWED	10	147	3	160
	OTHER	0	5	0	5
Total		431	947	25	1403

 

Step 2:

 

The original syntax was run with SPSSX-4 on a Vax mainframe in 1993, but it still works!  It wasn't my data, and there was no indication of what code 3 was in v670 (probably missing). 

 

compute famstat=v670*10+v715.
recode famstat (21 24 25 26=1)(22 23=2)(11 14 15 16=3)(12 13=2)(else=sysmis).
value labels famstat
  1 'Single: no kids' 2 'Couple: no kids' 3 'Single: + kids' 4 'Couple: + kids'.
print formats famstat (f1.0).

freq famstat .

 

I always use abbreviated syntax, but for clients I used the full syntax, except in the last line here to produce the following table.

 

famstat
		Frequency	Percent	Valid Percent	Cumulative Percent
Valid	Single: no kids	671	47.6	48.9	48.9
	Couple: no kids	535	37.9	39.0	87.8
	Single: + kids	167	11.8	12.2	100.0
	Total	1373	97.3	100.0
Missing	System	38	2.7
Total		1411	100.0

 

Not sure if the list will show the tables: if not I'll attach them in a Word document to you, but not the list..

 

John Hall

----- Original Message -----

From: [hidden email]

To: [hidden email]

Sent: Thursday, April 15, 2010 11:56 PM

Subject: total for tables-two diff categories

Dear SPSS users. I would like your gently help about the following issue.
I'm preparing a table in PASW18 that combines by sex two different questions of a survey.
Question 1 asks if the person has ever attended school and question 2 asks if the person who answered yes is Currently in the school system. The thing is that I must combine answers from the two questions in the same table, so my output will read: Never attended, currently attending, ever attended but not currently attending.

Although presentation of these results is quite easy, my question to the list is about the possibility of getting a total of the three categories. I have tried used the custom tables GUI, but it seems like I only can put a total at the bottom of each variable (not to sum the three categories of interest).

Do anybody know if this is possible to be defined?
Thanks in advance

[Thien Hop]



Hi all,

 

May I know if there is a way to add the significant test result inside the Ctable output. For now, when I add the Test Statistics to a Ctable, the results are indicated in a separate output from the normal tab for that same question. To add the sigtest result to the main output table, I have to export both tables to Excel & manually combine them together. This practice is quite time-consuming & easy to go wrong as there are so many column & rows.

 

Thank you so much for any advice in advance,

 

Hop

[Jon K Peck]

In Version 18 there is an option in CTABLES to put the significance results in the main table, although the representation is different from what is shown in the separate table.

There is also an extension command, SPSSINC MERGE TABLES that can be used to merge the signifcance results into the main table.  This works for Version 17 and later.

The command is available from SPSS Developer Central and requires the Python programmability plugin.

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

From:	Thien Hop <[hidden email]>
To:	[hidden email]
Date:	04/18/2010 11:20 PM
Subject:	[SPSSX-L] Incorporate Sigtest in SAME Ctable output
Sent by:	"SPSSX(r) Discussion" <[hidden email]>

---

Hi all,
 
May I know if there is a way to add the significant test result inside the Ctable output. For now, when I add the Test Statistics to a Ctable, the results are indicated in a separate output from the normal tab for that same question. To add the sigtest result to the main output table, I have to export both tables to Excel & manually combine them together. This practice is quite time-consuming & easy to go wrong as there are so many column & rows.
 
Thank you so much for any advice in advance,
 
Hop

[Rodrigo Briceño]

Thanks to all of you that provided me help with the syntax. Certainly the best way to do it was to create a new variable.

R.

El 4/16/2010 12:21 AM, John F Hall escribió:

Rodriguo

�

Have you checked your data first, to make sure there are no illogical combinations?

�

cros q1 by q2 .

�

Melissa's solution using if uses much more computer resources than recode so I've suggested an amendment to her syntax.� Instead of:�

�

if (Q1=0) schstat=0.

if (Q1=1 and Q2=1) schstat=1.

if (Q1=1 and Q2=0) schstat=2.

�

you can use

�

compute schstat = q1 + q2 * 10.

recode schstat (0 = 1)(11 = 2)(10 = 3).

var labels schstat 'School status'.

value labels� �

� � 1 'Never attended'

� � 2 'Currently Attending'

� � 3 'Past Attendance (Ever but not current)'.

�

I'm never happy about using� 0 as a code, so would always use� 1 = yes,� 2 = no, but my solution should produce the same result.

�

compute newvar = q1 * 10 + q2 .

recode newvar (20 =1)(11=1)(12 = 3) .

val lab newvar 1 'Never attended'

� � � � � � � � � � � � � � � � � � � � � 2 'Currently attending'

� � � � � � � � � � � � � � � � � � � � � 3 'Not attending'� .

�

Here's real life example I dug out this morning.

�

London Health for All

�

Secondary analysis of data on diet and health from survey of Health and Life-styles by Bloomsbury and� Islington� Health� � � � � Authority� (1993)� See: Steve Griffiths with John Hall� The Lie of the Land for the Health of the Nation: an analysis of� the� Bloomsbury� and Islington Health� &� Lifestyle� survey.� � � � � � Report to Bloomsbury and Islington AHA, May 1993

�

The data set included two variables, v670 (Responsibility for children) and v715 (Marital Status)� The client wanted a single variable combining these into 4 categories:

�

Single, no responsibilty for children

Couple, no responsibility for children

Single with responsibilty for children

Couple with responsibilty for children

�

Step 1:

�

freq v670 v715.
cro v715 by v670.

�

RESPONSIBLE FOR CHILDREN
		Frequency	Percent	Valid Percent	Cumulative Percent
Valid	Yes	432	30.6	30.6	30.6
	No	953	67.5	67.5	98.2
	3	26	1.8	1.8	100.0
	Total	1411	100.0	100.0

�

�

�

MARITAL STATUS
		Frequency	Percent	Valid Percent	Cumulative Percent
Valid	SINGLE	526	37.3	37.5	37.5
	MARRIED	467	33.1	33.3	70.8
	COHABITING	73	5.2	5.2	76.0
	SEPARATED	50	3.5	3.6	79.5
	DIVORCED	122	8.6	8.7	88.2
	WIDOWED	160	11.3	11.4	99.6
	OTHER	5	.4	.4	100.0
	Total	1403	99.4	100.0
Missing	System	8	.6
Total		1411	100.0

�

�

MARITAL STATUS * RESPONSIBLE FOR CHILDREN Crosstabulation
Count
		RESPONSIBLE FOR CHILDREN			Total
		Yes	No	3
MARITAL STATUS	SINGLE	88	422	16	526
	MARRIED	234	228	5	467
	COHABITING	30	43	0	73
	SEPARATED	23	27	0	50
	DIVORCED	46	75	1	122
	WIDOWED	10	147	3	160
	OTHER	0	5	0	5
Total		431	947	25	1403

�

Step 2:

�

The original syntax� was run with SPSSX-4 on a Vax mainframe in 1993, but it still works!� It wasn't my data, and there was no indication of what code 3 was in v670 (probably missing).�

�

compute famstat=v670*10+v715.
recode famstat (21 24 25 26=1)(22 23=2)(11 14 15 16=3)(12 13=2)(else=sysmis).
value labels famstat
� 1 'Single: no kids' 2 'Couple: no kids' 3 'Single: + kids' 4 'Couple: + kids'.
print formats famstat (f1.0).

freq famstat .

�

I always use abbreviated syntax, but for clients I used the full syntax, except in the last line here to produce the following table.

�

famstat
		Frequency	Percent	Valid Percent	Cumulative Percent
Valid	Single: no kids	671	47.6	48.9	48.9
	Couple: no kids	535	37.9	39.0	87.8
	Single: + kids	167	11.8	12.2	100.0
	Total	1373	97.3	100.0
Missing	System	38	2.7
Total		1411	100.0

�

Not sure if the list will show the tables: if not I'll attach them in a Word document to you, but not the list..

�

John Hall

----- Original Message -----

From: [hidden email]

To: [hidden email]

Sent: Thursday, April 15, 2010 11:56 PM

Subject: total for tables-two diff categories

Dear SPSS users. I would like your gently help about the following issue.
I'm preparing a table in PASW18 that combines by sex two different questions of a survey.
Question 1 asks if the person has ever attended school and question 2 asks if the person who answered yes is Currently in the school system. The thing is that I must combine answers from the two questions in the same table, so my output will read: Never attended, currently attending, ever attended but not currently attending.

Although presentation of these results is quite easy, my question to the list is about the possibility of getting a total of the three categories. I have tried used the custom tables GUI, but it seems like I only can put a total at the bottom of each variable (not to sum the three categories of interest).

Do anybody know if this is possible to be defined?
Thanks in advance