Issue with weighting (SPSS 17 x Mac)

[Luca Meyer-3]

Something weird is happening with my SPSS 17.0.2 x Mac (*) output when I use weighting.

I have run this sample syntax:

NEW FILE. 

INPUT PROGRAM. 

- STRING gender (A6). 

- LOOP #I =1 TO 100. 

- DO IF #I <= 52. 

- COMPUTE gender='Male'. 

- ELSE. 

- COMPUTE Gender='Female'. 

- END IF. 

- COMPUTE AgeCategory = trunc(uniform(3)+1). 

- END CASE. 

- END LOOP. 

- END FILE. 

END INPUT PROGRAM. 

FREQUENCIES VARIABLES=gender AgeCategory. 

***create and apply weightvar***. 

***to simulate 49 males, 51 females***. 

DO IF gender = 'Male'. 

- COMPUTE weightvar=49/52. 

ELSE IF gender = 'Female'. 

- COMPUTE weightvar=51/48. 

END IF. 

WEIGHT BY weightvar. 

FREQUENCIES VARIABLES=gender AgeCategory. 

(font: Programming and Data Management for SPSS ®  Statistics 17.0 - page 90)

and I get the following output:

<--- before weighting --->

gender
		Frequency	Percent	Valid Percent	Cumulative Percent
Valid	Female	48	48,0	48,0	48,0
	Male	52	52,0	52,0	100,0
	Total	100	100,0	100,0
AgeCategory
		Frequency	Percent	Valid Percent	Cumulative Percent
Valid	1,00	40	40,0	40,0	40,0
	2,00	31	31,0	31,0	71,0
	3,00	29	29,0	29,0	100,0
	Total	100	100,0	100,0

<--- after weighting --->

gender
		Frequency	Percent	Valid Percent	Cumulative Percent
Valid	Female	51	51,0	51,0	51,0
	Male	49	49,0	49,0	100,0
	Total	100	100,0	100,0
AgeCategory
		Frequency	Percent	Valid Percent	Cumulative Percent
Valid	1,00	40	39,7	39,7	39,7
	2,00	31	31,3	31,3	71,0
	3,00	29	29,0	29,0	100,0
	Total	100	100,0	100,0

Look at the second AgeCategory Frequency: it is unchanged. But now look at the corresponding Percent column: it does not correspond.

Can someone help me to understand what's going on here?

Thanks,

Luca

(*) For the time being I am using the trial version I have downloaded from the SPSS support site

Mr. Luca Meyer

www.lucameyer.com

[ViAnn Beadle]

Sure looks to me like you have 39.7 cases for agecat = 1 and 31.3 cases for agecat=2.  Frequencies are displayed with no decimal digits

 

From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of Luca Meyer
Sent: Saturday, April 25, 2009 10:30 AM
To: [hidden email]
Subject: Issue with weighting (SPSS 17 x Mac)

 

Something weird is happening with my SPSS 17.0.2 x Mac (*) output when I use weighting.

 

I have run this sample syntax:

 

NEW FILE. 

INPUT PROGRAM. 

- STRING gender (A6). 

- LOOP #I =1 TO 100. 

- DO IF #I <= 52. 

- COMPUTE gender='Male'. 

- ELSE. 

- COMPUTE Gender='Female'. 

- END IF. 

- COMPUTE AgeCategory = trunc(uniform(3)+1). 

- END CASE. 

- END LOOP. 

- END FILE. 

END INPUT PROGRAM. 

FREQUENCIES VARIABLES=gender AgeCategory. 

***create and apply weightvar***. 

***to simulate 49 males, 51 females***. 

DO IF gender = 'Male'. 

- COMPUTE weightvar=49/52. 

ELSE IF gender = 'Female'. 

- COMPUTE weightvar=51/48. 

END IF. 

WEIGHT BY weightvar. 

FREQUENCIES VARIABLES=gender AgeCategory. 

 

(font: Programming and Data Management for SPSS ®  Statistics 17.0 - page 90)

 

and I get the following output:

 

<--- before weighting --->

 

gender
		Frequency	Percent	Valid Percent	Cumulative Percent
Valid	Female	48	48,0	48,0	48,0
	Male	52	52,0	52,0	100,0
	Total	100	100,0	100,0

[Hector Maletta]

SPSS rounds weighted frequencies. Your weights are fractional, and need rounding before getting displayed. So a small difference arises. Percentages are computed on the unrounded weighted frequencies.

 

Hector

 

---

From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of Luca Meyer
Sent: 25 April 2009 13:30
To: [hidden email]
Subject: Issue with weighting (SPSS 17 x Mac)

 

Something weird is happening with my SPSS 17.0.2 x Mac (*) output when I use weighting.

 

I have run this sample syntax:

 

NEW FILE. 

INPUT PROGRAM. 

- STRING gender (A6). 

- LOOP #I =1 TO 100. 

- DO IF #I <= 52. 

- COMPUTE gender='Male'. 

- ELSE. 

- COMPUTE Gender='Female'. 

- END IF. 

- COMPUTE AgeCategory = trunc(uniform(3)+1). 

- END CASE. 

- END LOOP. 

- END FILE. 

END INPUT PROGRAM. 

FREQUENCIES VARIABLES=gender AgeCategory. 

***create and apply weightvar***. 

***to simulate 49 males, 51 females***. 

DO IF gender = 'Male'. 

- COMPUTE weightvar=49/52. 

ELSE IF gender = 'Female'. 

- COMPUTE weightvar=51/48. 

END IF. 

WEIGHT BY weightvar. 

FREQUENCIES VARIABLES=gender AgeCategory. 

 

(font: Programming and Data Management for SPSS ®  Statistics 17.0 - page 90)

 

and I get the following output:

 

<--- before weighting --->

 

gender
		Frequency	Percent	Valid Percent	Cumulative Percent
Valid	Female	48	48,0	48,0	48,0
	Male	52	52,0	52,0	100,0
	Total	100	100,0	100,0
AgeCategory
		Frequency	Percent	Valid Percent	Cumulative Percent
Valid	1,00	40	40,0	40,0	40,0
	2,00	31	31,0	31,0	71,0
	3,00	29	29,0	29,0	100,0
	Total	100	100,0	100,0

 

<--- after weighting --->

 

gender
		Frequency	Percent	Valid Percent	Cumulative Percent
Valid	Female	51	51,0	51,0	51,0
	Male	49	49,0	49,0	100,0
	Total	100	100,0	100,0
AgeCategory
		Frequency	Percent	Valid Percent	Cumulative Percent
Valid	1,00	40	39,7	39,7	39,7
	2,00	31	31,3	31,3	71,0
	3,00	29	29,0	29,0	100,0
	Total	100	100,0	100,0

 

Look at the second AgeCategory Frequency: it is unchanged. But now look at the corresponding Percent column: it does not correspond.

 

Can someone help me to understand what's going on here?

 

Thanks,

Luca

 

(*) For the time being I am using the trial version I have downloaded from the SPSS support site

 

 

Mr. Luca Meyer

www.lucameyer.com

 

 

 

[Luca Meyer-3]

Yes, that was it. 

Thank you,

Luca

Il giorno 25/apr/09, alle ore 18:37, ViAnn Beadle ha scritto:

Sure looks to me like you have 39.7 cases for agecat = 1 and 31.3 cases for agecat=2.  Frequencies are displayed with no decimal digits

 

From: SPSSX(r) Discussion [[hidden email]] On Behalf Of Luca Meyer
Sent: Saturday, April 25, 2009 10:30 AM
To: [hidden email]
Subject: Issue with weighting (SPSS 17 x Mac)

 

Something weird is happening with my SPSS 17.0.2 x Mac (*) output when I use weighting.

 

I have run this sample syntax:

 

NEW FILE. 

INPUT PROGRAM. 

- STRING gender (A6). 

- LOOP #I =1 TO 100. 

- DO IF #I <= 52. 

- COMPUTE gender='Male'. 

- ELSE. 

- COMPUTE Gender='Female'. 

- END IF. 

- COMPUTE AgeCategory = trunc(uniform(3)+1). 

- END CASE. 

- END LOOP. 

- END FILE. 

END INPUT PROGRAM. 

FREQUENCIES VARIABLES=gender AgeCategory. 

***create and apply weightvar***. 

***to simulate 49 males, 51 females***. 

DO IF gender = 'Male'. 

- COMPUTE weightvar=49/52. 

ELSE IF gender = 'Female'. 

- COMPUTE weightvar=51/48. 

END IF. 

WEIGHT BY weightvar. 

FREQUENCIES VARIABLES=gender AgeCategory. 

 

(font: Programming and Data Management for SPSS ®  Statistics 17.0 - page 90)

 

and I get the following output:

 

<--- before weighting --->

 

gender
		Frequency	Percent	Valid Percent	Cumulative Percent
Valid	Female	48	48,0	48,0	48,0
	Male	52	52,0	52,0	100,0
	Total	100	100,0	100,0

AgeCategory

 

Frequency

Percent

Valid Percent

Cumulative Percent

Valid

1,00

40

40,0

40,0

40,0

2,00

31

31,0

31,0

71,0

3,00

29

29,0

29,0

100,0

Total

100

100,0

100,0

 

<--- after weighting --->

 

gender
		Frequency	Percent	Valid Percent	Cumulative Percent
Valid	Female	51	51,0	51,0	51,0
	Male	49	49,0	49,0	100,0
	Total	100	100,0	100,0

AgeCategory

 

Frequency

Percent

Valid Percent

Cumulative Percent

Valid

1,00

40

39,7

39,7

39,7

2,00

31

31,3

31,3

71,0

3,00

29

29,0

29,0

100,0

Total

100

100,0

100,0

 

Look at the second AgeCategory Frequency: it is unchanged. But now look at the corresponding Percent column: it does not correspond.

 

Can someone help me to understand what's going on here?

 

Thanks,

Luca

 

(*) For the time being I am using the trial version I have downloaded from the SPSS support site

 

 

Mr. Luca Meyer

www.lucameyer.com