new option in crosstabs- comparing column proportions

[msherman]

Dear list:  Below is a crosstabs output and includes subscript letters which is new

Feature in spss. Compare column proportions which is presented in the Cell Display tab.

 If I look at Column Y 1 and read down the percent we have 62.5% and 37.5% and then for Column

Y 2 we have 25% and 75% but the subscripts for Column 1 Y are all a’s and for Column 2 Y the

Subscripts are all b’s. So does this mean that 62.5% is sign.  Different from 25.0% and that 37.5%

Is sign. Different from 75%.  ??

x * y Crosstabulation
			y		Total
			1.00	2.00
x	1.00	Count	50a	10b	60
		% within x	83.3%	16.7%	100.0%
		% within y	62.5%	25.0%	50.0%
	2.00	Count	30a	30b	60
		% within x	50.0%	50.0%	100.0%
		% within y	37.5%	75.0%	50.0%
Total		Count	80	40	120
		% within x	66.7%	33.3%	100.0%
		% within y	100.0%	100.0%	100.0%
Each subscript letter denotes a subset of y categories whose column proportions do not differ significantly from each other at the .05 level.

 

 

 

Martin F. Sherman, Ph.D.

Professor of Psychology

Director of  Masters Education in Psychology: Thesis Track

 

Loyola University Maryland

Department of Psychology

222 B Beatty Hall

4501 North Charles Street

Baltimore, MD 21210

 

410-617-2417

[hidden email]

 

[x p]

 My question is basic. If a data set of around 300 is normally distributed is it reasonable to use chi square tests and t-tests for different aspects of analysis. I have suggested that for consistency, only one approach should be followed ie parametric or non-parametric but not both. I would appreciate any comments and advice on this. 

Sorry I forgot to change the subject line

Rod

[Kornbrot, Diana]

This is an old fashioned and counterproductive format

Many people will want the numbers directly into spreadsheet. Its  a pain to remove subscripts

So a SEPARATE column for the subscripts a, b et would be a very useful option.

This problem also occurs in CORRELATION, where, sensibly there is option NOT to include superscripts for significance.

So strong recommendation, options:

Option a. Numbers should be numbers ONLY. Footnote subscripts and superscripts should have column to themselves.

Option b. Current ghastly format,  [still liked by some journals] so many will want to copy the SPSS tables direct into word-processor.

Best

Diana

____________________________________________

Professor Diana Kornbrot
email:  d.e.kornbrot@...    
web:    http://web.me.com/kornbrot/KornbrotHome.html
Work
School of Psychology
 University of Hertfordshire
 College Lane, Hatfield, Hertfordshire AL10 9AB, UK
 voice:   [phone]

Home
 19 Elmhurst Avenue
 London N2 0LT, UK
    voice:   [phone]
    mobile: [phone]

From: Martin Sherman <[hidden email]>
Reply-To: Martin Sherman <[hidden email]>
Date: Thu, 8 Sep 2011 22:21:07 +0100
To: "[hidden email]" <[hidden email]>
Subject: new option in crosstabs- comparing column proportions

Dear list:  Below is a crosstabs output and includes subscript letters which is new

Feature in spss. Compare column proportions which is presented in the Cell Display tab.

 If I look at Column Y 1 and read down the percent we have 62.5% and 37.5% and then for Column

Y 2 we have 25% and 75% but the subscripts for Column 1 Y are all a’s and for Column 2 Y the

Subscripts are all b’s. So does this mean that 62.5% is sign.  Different from 25.0% and that 37.5%

Is sign. Different from 75%.  ??

x * y Crosstabulation
			y		Total
			1.00	2.00
x	1.00	Count	50a	10b	60
		% within x	83.3%	16.7%	100.0%
		% within y	62.5%	25.0%	50.0%
	2.00	Count	30a	30b	60
		% within x	50.0%	50.0%	100.0%
		% within y	37.5%	75.0%	50.0%
Total		Count	80	40	120
		% within x	66.7%	33.3%	100.0%
		% within y	100.0%	100.0%	100.0%
Each subscript letter denotes a subset of y categories whose column proportions do not differ significantly from each other at the .05 level.

 

 

 

Martin F. Sherman, Ph.D.

Professor of Psychology

Director of  Masters Education in Psychology: Thesis Track

 

Loyola University Maryland

Department of Psychology

222 B Beatty Hall

4501 North Charles Street

Baltimore, MD 21210

 

410-617-2417

[hidden email]

 

[Maguin, Eugene]

Why wouldn't the statistical test follow from the question you wanted to answer given the data that you have?

 

Gene Maguin

---

From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of x p
Sent: Thursday, September 08, 2011 8:34 PM
To: [hidden email]
Subject: non-parametric and parametric tests together

 My question is basic. If a data set of around 300 is normally distributed is it reasonable to use chi square tests and t-tests for different aspects of analysis. I have suggested that for consistency, only one approach should be followed ie parametric or non-parametric but not both. I would appreciate any comments and advice on this. 

Sorry I forgot to change the subject line

Rod

[Antoon Smulders]

Hello list,

I haven’t seen an answer to this question. The footnote in the output is a bit ambiguous, but the online help gives a clear example:

http://127.0.0.1:56507/help/index.jsp?topic=/com.ibm.spss.statistics.help/syn_crosstabs.htm (look at the cells subcommand).

 

I have another remark about the same topic.

The syntax allows for /cells prob and /cells bprop.

Bprop uses the Bonferoni adjustment, while /prop doesn’t.

In the output however, it is not made clear if the adjustment has been made.

By the way: you can specify both, i.e. “ /cells prop bprop” without SPSS complaining (SPSS seems more tolerant then statisticians J) but nevertheless the program only gives one pair of subscripts and it remains unclear if the result is Bonferoni adjusted or not.

 

Antoon Smulders

 

 

 

Van: SPSSX(r) Discussion [mailto:[hidden email]] Namens Martin Sherman
Verzonden: donderdag 8 september 2011 23:21
Aan: [hidden email]
Onderwerp: new option in crosstabs- comparing column proportions

 

Dear list:  Below is a crosstabs output and includes subscript letters which is new

Feature in spss. Compare column proportions which is presented in the Cell Display tab.

 If I look at Column Y 1 and read down the percent we have 62.5% and 37.5% and then for Column

Y 2 we have 25% and 75% but the subscripts for Column 1 Y are all a’s and for Column 2 Y the

Subscripts are all b’s. So does this mean that 62.5% is sign.  Different from 25.0% and that 37.5%

Is sign. Different from 75%.  ??

x * y Crosstabulation
			y		Total
			1.00	2.00
x	1.00	Count	50a	10b	60
		% within x	83.3%	16.7%	100.0%
		% within y	62.5%	25.0%	50.0%
	2.00	Count	30a	30b	60
		% within x	50.0%	50.0%	100.0%
		% within y	37.5%	75.0%	50.0%
Total		Count	80	40	120
		% within x	66.7%	33.3%	100.0%
		% within y	100.0%	100.0%	100.0%
Each subscript letter denotes a subset of y categories whose column proportions do not differ significantly from each other at the .05 level.

 

 

 

Martin F. Sherman, Ph.D.

Professor of Psychology

Director of  Masters Education in Psychology: Thesis Track

 

Loyola University Maryland

Department of Psychology

222 B Beatty Hall

4501 North Charles Street

Baltimore, MD 21210

 

410-617-2417

[hidden email]