SPSSX Discussion

Ascending-descending frequencies in mult response

Classic

List

Threaded

8 messages Options

John F Hall

Ascending-descending frequencies in mult response

There doesn’t appear to be a facility in MULT RESP to arrange rows or columns in ascending/descending order of frequencies. Be nice if we could.

John F Hall (Mr)

[Retired academic survey researcher]

Email: [hidden email]

Website: www.surveyresearch.weebly.com

SPSS start page: www.surveyresearch.weebly.com/1-survey-analysis-workshop

===================== 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

Jignesh Sutar

Re: Ascending-descending frequencies in mult response

Download the Extension STATS CATEGORY ORDER.

Which is designed to achieve this, and more, such as anchoring "Other" &/or "Don't know" to the bottom.

Jignesh Sutar

Re: Ascending-descending frequencies in mult response

John, You can do this directly in CTABLES, see CATEGORIES VARIABLES sub command. I've also shown how you can use STATS CATEGORY ORDER to anchor Other/Don't know, if required.

dataset close all.
new file.
output close all.

set seed = 10. 
input program. 
loop #i = 1 to 500. 
  compute case = #i. 
  
  vector v(20).
  loop #j=1 to 20.
      compute v(#j)=rnd(rv.uniform(0,1)).
  end loop.
end case. 
end loop. 
end file. 
end input program. 
dataset name sim. 
execute. 

descrip v1 to v20 /stat sum.

mrsets /mdgroup name=$mrsset_origOrderVars categorylabels=varlabels variables = v1 to v20 value=1 /display name=[$mrsset_origOrderVars].

ctables /table $mrsset_origOrderVars /titles title="Table1: Unodered".
ctables /table $mrsset_origOrderVars /titles title="Table2: Ordered" /categories variables=$mrsset_origOrderVars order=d key=count.


stats category order items=$mrsset_origOrderVars prefix=dorder specialvars=v19 v20 other=no
/options order=d categorylabels=varlabels specialsloc=after missing=exclude.

ctables /table $dorder_$mrsset_origOrderVars /titles title="Table3: Ordered (via STATS CATEGORY ORDER with Other/DK anchored)".

Jignesh Sutar

Re: Ascending-descending frequencies in mult response

PS. If you wrap syntax code with "Raw text" tag then it fails to come through in email

So here's the code for those not reading via nabble (but email):

dataset close all.
new file.
output close all.

set seed = 10.
input program.
loop #i = 1 to 500.
compute case = #i.

vector v(20).
loop #j=1 to 20.
compute v(#j)=rnd(rv.uniform(0,1)).
end loop.
end case.
end loop.
end file.
end input program.
dataset name sim.
execute.

descrip v1 to v20 /stat sum.

mrsets /mdgroup name=$mrsset_origOrderVars categorylabels=varlabels variables = v1 to v20 value=1 /display name=[$mrsset_origOrderVars].

ctables /table $mrsset_origOrderVars /titles title="Table1: Unodered".
ctables /table $mrsset_origOrderVars /titles title="Table2: Ordered" /categories variables=$mrsset_origOrderVars order=d key=count.

stats category order items=$mrsset_origOrderVars prefix=dorder specialvars=v19 v20 other=no
/options order=d categorylabels=varlabels specialsloc=after missing=exclude.

ctables /table $dorder_$mrsset_origOrderVars /titles title="Table3: Ordered (via STATS CATEGORY ORDER with Other/DK anchored)".

Jon K Peck

Re: Ascending-descending frequencies in mult response

In reply to this post by John F Hall

You can control the sort order of MR (or regular) variables in CTABLES. The output can be order by category value, category label, or category count. In syntax it would look like this

/CATEGORIES VARIABLES=$myMRvariable ORDER=D KEY=COUNT

Jon Peck (no "h") aka Kim
Senior Software Engineer, IBM
[hidden email]
phone: 720-342-5621

From: John F Hall <[hidden email]>
To: [hidden email]
Date: 06/01/2015 02:58 AM
Subject: [SPSSX-L] Ascending-descending frequencies in mult response
Sent by: "SPSSX(r) Discussion" <[hidden email]>

There doesn’t appear to be a facility in MULT RESP to arrange rows or columns in ascending/descending order of frequencies. Be nice if we could.

John F Hall (Mr)

[Retired academic survey researcher]

Email: johnfhall@...

Website: www.surveyresearch.weebly.com

SPSS start page: www.surveyresearch.weebly.com/1-survey-analysis-workshop

===================== To manage your subscription to SPSSX-L, send a message to LISTSERV@... (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

John F Hall

Re: Ascending-descending frequencies in mult response

Jon

Many thanks for this.

My version was:

mult resp groups=

Q2 'Quals working for..' (v109 to v112 (1,7))

/freq q2.

Q2 Frequencies
		Responses		Percent of Cases
		N	Percent	Percent of Cases
Quals working for..^a	GCSE	186	19.2%	23.1%
	BTEC	32	3.3%	4.0%
	A-levels	677	69.7%	84.2%
	CPVE	6	0.6%	0.7%
	City and Guilds	18	1.9%	2.2%
	RSA	23	2.4%	2.9%
	Other	29	3.0%	3.6%
Total		971	100.0%	120.8%
a. Group

It took me an hour to work out what to do and then write something in CTABLES, (only one error, used # instead of $ for setname: print in TFM too small) but I eventually got to:

MRSETS

/McGROUP NAME= $Q2 LABEL= 'Quals working for..'

VARIABLES= v109 to v112

/DISPLAY NAME= [$Q2].

ctables

/tab $q2

/CATEGORIES VARIABLES=$q2 ORDER=D KEY=COUNT.

		Count
Quals working for..	A-levels	677
	GCSE	185
	BTEC	32
	Other	29
	RSA	23
	City and Guilds	18
	CPVE	6

Will play with CTABLES to get % and TOTAL displayed, then see if I can decipher Jignesh’s fearsome code.

Nice to see /FORMAT CONDENSE still available in MULT RESPONSE: can’t see when I’d use it, but why was it ever removed from FREQUENCIES?

John

From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of Jon K Peck
Sent: 01 June 2015 14:21
To: [hidden email]
Subject: Re: Ascending-descending frequencies in mult response

There doesn’t appear to be a facility in MULT RESP to arrange rows or columns in ascending/descending order of frequencies. Be nice if we could.

John F Hall (Mr)

[Retired academic survey researcher]

Email: [hidden email]

Website: www.surveyresearch.weebly.com

SPSS start page: www.surveyresearch.weebly.com/1-survey-analysis-workshop

John F Hall

Re: Ascending-descending frequencies in mult response

In reply to this post by Jignesh Sutar

Jignesh

My variables are MCGROUP, not MDGROUP. I've modified your syntax, but it effectively comes out the same as Jon's.

mrsets

/mcgroup name=$q17 label = 'Newspapers read' variables = v207 to v213

/display name=[$q17].

ctables /table $q17 /titles title="Table1: Unordered".

ctables /table $q17 /titles title="Table2: Ordered"

/categories variables=$q17 order=d key=count.

Table2: Ordered
		Count
Newspapers read	Local evening	206
	Mail	135
	Mirror	124
	Tele- graph	105
	Express	101
	Sun	90
	Independ-ent	76
	Guardian	76
	Times	69
	Local morning	51
	Today	37
	Financ. Times	30
	Star	19
	Too many	5
	Morning Star	1

Table2: Ordered
		Count
Newspapers read	Local evening	206
	Mail	135
	Mirror	124
	Tele- graph	105
	Express	101
	Sun	90
	Independ-ent	76
	Guardian	76
	Times	69
	Local morning	51
	Today	37
	Financ. Times	30
	Star	19
	Too many	5
	Morning Star	1

Even after RTFM I can't get any %% displayed, just error messages. For some reason COLPCT.COUNT and others cause errors.

John

-----Original Message-----
From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of Jignesh Sutar
Sent: 01 June 2015 12:55
To: [hidden email]
Subject: Re: Ascending-descending frequencies in mult response

PS. If you wrap syntax code with "Raw text" tag then it fails to come through in email

So here's the code for those not reading via nabble (but email):

dataset close all.

new file.

output close all.

set seed = 10.

input program.

loop #i = 1 to 500.

compute case = #i.

vector v(20).

loop #j=1 to 20.

compute v(#j)=rnd(rv.uniform(0,1)).

end loop.

end case.

end loop.

end file.

end input program.

dataset name sim.

execute.

descrip v1 to v20 /stat sum.

mrsets /mdgroup name=$mrsset_origOrderVars categorylabels=varlabels variables = v1 to v20 value=1 /display name=[$mrsset_origOrderVars].

ctables /table $mrsset_origOrderVars /titles title="Table1: Unodered".

ctables /table $mrsset_origOrderVars /titles title="Table2: Ordered"

/categories variables=$mrsset_origOrderVars order=d key=count.

stats category order items=$mrsset_origOrderVars prefix=dorder

specialvars=v19 v20 other=no

/options order=d categorylabels=varlabels specialsloc=after missing=exclude.

ctables /table $dorder_$mrsset_origOrderVars /titles title="Table3: Ordered (via STATS CATEGORY ORDER with Other/DK anchored)".

View this message in context: http://spssx-discussion.1045642.n5.nabble.com/Ascending-descending-frequencies-in-mult-response-tp5729660p5729667.html

Sent from the SPSSX Discussion mailing list archive at Nabble.com.

=====================

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

John F Hall

Re: Ascending-descending frequencies in mult response

In reply to this post by John F Hall

Jignesh

Well spotted! I really don’t have time to study all this syntax: it took me the whole morning to get that far!. Undergrad students would have gone home hours ago. Why can’t they just have /FOR DVAL in MULT RESP?

It’s not for me but I was checking the data from a 1989 survey in FE colleges and Sixth Forms of Attitudes to Proposed Student Loans (conducted by the National Union of Students) which Edinburgh have agreed to curate, so I can clear it from my machine.

(See: http://stats.datalib.edina.ac.uk/sdaweb/analysis/?dataset=nus89)

Essex Re-share is too bureaucratic (I don’t have authorisation for the data and no-one answers repeated requests, but I’m loath to destroy them). At least Essex can store and distribute the SPSS *.sav file: with Edinburgh you have to re-create it (in Stata and SAS as well as SPSS) but why anyone would want to use Stata or SAS on small surveys is beyond my comprehension.

John

From: [hidden email] [mailto:[hidden email]] On Behalf Of Jignesh Sutar
Sent: 02 June 2015 13:09
To: John F Hall
Subject: Re: Ascending-descending frequencies in mult response

You don't want "Other" floating around there in the middle! Have no fear, start studying! Happy to help out if you get stuck...

On 2 June 2015 at 11:00, John F Hall <[hidden email]> wrote:

Jon

Many thanks for this.

My version was:

mult resp groups=
Q2 'Quals working for..' (v109 to v112 (1,7))
                       /freq q2.

Q2 Frequencies

Responses
Percent of Cases
N
Percent
Quals working for..^a
GCSE
186
19.2%
23.1%
BTEC
32
3.3%
4.0%
A-levels
677
69.7%
84.2%
CPVE
6
0.6%
0.7%
City and Guilds
18
1.9%
2.2%
RSA
23
2.4%
2.9%
Other
29
3.0%
3.6%
Total
971
100.0%
120.8%
a. Group

It took me an hour to work out what to do and then write something in CTABLES, (only one error, used # instead of $ for setname: print in TFM too small) but I eventually got to:

MRSETS
/McGROUP      NAME= $Q2 LABEL= 'Quals working for..'
VARIABLES= v109 to v112
/DISPLAY         NAME= [$Q2].

ctables
/tab $q2
/CATEGORIES VARIABLES=$q2 ORDER=D KEY=COUNT.

Count
Quals working for..
A-levels
677
GCSE
185
BTEC
32
Other
29
RSA
23
City and Guilds
18
CPVE
6

Will play with CTABLES to get % and TOTAL displayed, then see if I can decipher Jignesh’s fearsome code.

Nice to see /FORMAT CONDENSE still available in MULT RESPONSE: can’t see when I’d use it, but why was it ever removed from FREQUENCIES?

John

From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of Jon K Peck
Sent: 01 June 2015 14:21
To: [hidden email]
Subject: Re: Ascending-descending frequencies in mult response

You can control the sort order of MR (or regular) variables in CTABLES. The output can be order by category value, category label, or category count. In syntax it would look like this

/CATEGORIES VARIABLES=$myMRvariable ORDER=D KEY=COUNT

Jon Peck (no "h") aka Kim
Senior Software Engineer, IBM
[hidden email]
phone: 720-342-5621

From: John F Hall <[hidden email]>
To: [hidden email]
Date: 06/01/2015 02:58 AM
Subject: [SPSSX-L] Ascending-descending frequencies in mult response
Sent by: "SPSSX(r) Discussion" <[hidden email]>

There doesn’t appear to be a facility in MULT RESP to arrange rows or columns in ascending/descending order of frequencies. Be nice if we could.

John F Hall (Mr)
[Retired academic survey researcher]

Email: [hidden email]
Website: www.surveyresearch.weebly.com
SPSS start page: www.surveyresearch.weebly.com/1-survey-analysis-workshop

===================== 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