How to add a line to a table with a not weighted sum

[fjmenendez]

Hi,

I have been asked about how to get a simple sum of mean (not weighted), such as the one in bold font. I don´t fully understand the purpose of that sum from a data analysis point of view. Despite that, I d´appreciate help to do it.

		VARX1 Mean1	VARX2 Mean2
VARX3 (categorical)	aaa	48	6000
	bbb	40	7000
	ccc	30	10000
	ddd	40	30000
	Sum of means (simple, not weighted)	158	53000

Thanks in advance :)

Jorge.

===================== 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]

You can use custom tables (CTABLES command), provided you have that add-on module, to generate the weighted and unweighted means. Since version 18 there has been an introduction of PCOMPUTE within CTABLE to be able to post compute additional statistics from initial aggregated statistics i.e. in your cases taking the sum of means of VARX1 across all categories of VARX3.

Here is an example using the employee data shipped with SPSS:

ctables
 /pcompute &SumMean=expr([1]+[2]+[3]) /pproperties &SumMean label="Sum of means (simple, not weighted)"
  /table jobcat[c] by (salbegin+salary)[s] [mean 'Weighted mean' umean 'Unweighted mean']
  /categories variables=jobcat [1,2,3,othernm, &SumMean]
  /slabels position=row.

This is far as I can get as close to your desired output. You will notice that there is a row for both weighted and unweighted for each category, when only weighted is desired for valid categories and unweighted for the sum. There is no direct way of changing this behaviour in CTABLE (would be good if it was possible) but you can do so using extension command that modify output tables. Or alternatively if you are exporting to excel anyway you could delete these redundant rows using a VBA macro or if it is just a one off then delete manually.

Hope that helps,

Jignesh

On 15 November 2014 22:52, Florentino Jorge Menendez <[hidden email]> wrote:

Hi,

I have been asked about how to get a simple sum of mean (not weighted), such as the one in bold font. I don´t fully understand the purpose of that sum from a data analysis point of view. Despite that, I d´appreciate help to do it.

		VARX1 Mean1	VARX2 Mean2
VARX3 (categorical)	aaa	48	6000
	bbb	40	7000
	ccc	30	10000
	ddd	40	30000
	Sum of means (simple, not weighted)	158	53000

Thanks in advance :)

Jorge.

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

[Rich Ulrich]

You do want to get that sort of mean in a report of (say) sales for
different stores.  The format that you suggest is one sort of default
for the Report Generator.

--
Rich Ulrich

---

Date: Sat, 15 Nov 2014 20:52:29 -0200
From: [hidden email]
Subject: How to add a line to a table with a not weighted sum
To: [hidden email]

Hi,

I have been asked about how to get a simple sum of mean (not weighted), such as the one in bold font. I don´t fully understand the purpose of that sum from a data analysis point of view. Despite that, I d´appreciate help to do it.

		VARX1 Mean1	VARX2 Mean2
VARX3 (categorical)	aaa	48	6000
	bbb	40	7000
	ccc	30	10000
	ddd	40	30000
	Sum of means (simple, not weighted)	158	53000

Thanks in advance :)

Jorge.

===================== 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]

It is trivial to calculate these using AGGREGATE!

fjmenendez wrote

Hi,

I have been asked about how to get a simple sum of mean (not weighted),
such as the one in bold font. I don´t fully understand the purpose of that
sum from a data analysis point of view. Despite that, I d´appreciate help
to do it.


  VARX1
Mean1
VARX2
Mean2VARX3
(categorical)aaa486000bbb
407000ccc3010000ddd4030000 *Sum of means (simple, not weighted)**158**53000*

Thanks in advance :)
Jorge.

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