How do I tell SPSS that numbers represent frequencies of categories and not of individual cases when performing chi-square test of independence?

[Faiz Rasool]

Hi all,

 

I have data in form of frequencies representing categories of two variables.  Instead of typing individual cases I would like to only enter frequencies of each category, and  request chi-square test of independence. Details of variables and frequencies of each category follows. Variable name equals awareness. highly aware 102, moderately aware 36, no awareness 12. variable name = conservation behavior. high conservation 90, moderate conservation 39, no conservation 21. I would like to type these frequencies in spss and tell it that these are frequencies of categories, and not individual cases. 

 

I tried to weigh cases by frequencies. But I was only  able to define one variable as frequency variable. The value I got after requesting chi-square test was 306.4. According to calculation done by hand the answer is not correct. I tried to find the answer of my problem on the web. But most of the tutorials have used pictures to explain data entry. Since I am blind, I was not able to solve my problem from those tutorials.

 

Any help is most appreciated,

 

Faiz.

 

[Marta Garcia-Granero]

Faiz Rasool wrote:

I have data in form of frequencies representing categories of two variables.  Instead of typing individual cases I would like to only enter frequencies of each category, and  request chi-square test of independence. Details of variables and frequencies of each category follows. Variable name equals awareness. highly aware 102, moderately aware 36, no awareness 12. variable name = conservation behavior. high conservation 90, moderate conservation 39, no conservation 21. I would like to type these frequencies in spss and tell it that these are frequencies of categories, and not individual cases. 

 

I tried to weigh cases by frequencies. But I was only  able to define one variable as frequency variable. The value I got after requesting chi-square test was 306.4. According to calculation done by hand the answer is not correct. I tried to find the answer of my problem on the web. But most of the tutorials have used pictures to explain data entry. Since I am blind, I was not able to solve my problem from those tutorials.

 

I am not really sure I understand you... What you need is a frequency variable for each combination of awareness and conservation behavior. Since awareness has 3 levels (highly, moderately and none) and conservation has 3 levels too (high, moderate and none), then you need 9 cells in your dataset, with only one frequency variable. You weight by it and that is it. I see you have an overall sample size of 150 (102+36+12, or, 90+39+21), but you only give marginal frequencies (totals by rows or columns) and you need to weight by the individual  frequencies (that add up to 150):

Awareness Conservation  Frequency
High         High                   N1
High         Moderate           N2
High         None                 N3
Moderate   High                 N4
Moderate   Moderate         N5
Moderate   None                N6
None         High                  N7
None         Moderate          N8
None         None                 N9

HTH,
Marta GG
===================== 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

[Marta Garcia-Granero]

Faiz Rasool wrote:

> thank you for your reply. The mental picture I have of data is
> follows. Since awareness is the independent variable, frequencies of
> its each category I.e. high, moderate, and no will be written in the
> first three cells of first column in the data editor. Since 102 people
> are highly aware 102 would be typed in the first cell of the first
> column, followed by 36 and 12 in cell two and three of the first
> column. Same would apply to conservation behavior variable, and 90,
> 39, 21 would be typed respectively in cell 1 to 3 of column two of the
> variable. Then crosstab,   awareness as the row variable and
> conservation behavior as the column variable.
>
> reading your response suggest to me that the way I am entering data is
> incorrect. So any help on how I should correctly enter the data I have
> mentioned in my emails would be appreciated.

Hi again:

I will adress my reply to the list, instead of privately. That way
everybody will know what information I am giving you (to avoid
redundancies).

Definitely, the layout you describe is incorrect. See if this example
helps you to understand the correct data layout:

DATA LIST LIST/Region (A5) WeightGrp (A10) Frequency(F8).
BEGIN DATA
South    Normal    130
South    Overweight    200
South    Obese    170
East    Normal    120
East    Overweight    170
East    Obese    210
North    Normal    100
North    Overweight    163
North    Obese    237
END DATA.
WEIGHT BY Frequency.

CROSSTABS
  /TABLES=Region BY WeightGrp
  /FORMAT=AVALUE TABLES
  /STATISTICS=CHISQ
  /CELLS=COUNT
  /COUNT ROUND CELL.


HTH,
Marta

> best regards,
> faiz.
>
>     ----- Original Message -----
>     *From:* García-Granero <mailto:[hidden email]>
>     *To:* [hidden email] <mailto:[hidden email]>
>     *Sent:* Monday, February 22, 2010 2:23 PM
>     *Subject:* Re: How do I tell SPSS that numbers represent
>     frequencies of categories and not of individual cases when
>     performing chi-square test of independence?
>
>     Faiz Rasool wrote:
>>     I have data in form of frequencies representing categories of two
>>     variables.  Instead of typing individual cases I would like to
>>     only enter frequencies of each category, and  request chi-square
>>     test of independence. Details of variables and frequencies of
>>     each category follows. Variable name equals awareness. highly
>>     aware 102, moderately aware 36, no awareness 12. variable name =
>>     conservation behavior. high conservation 90, moderate
>>     conservation 39, no conservation 21. I would like to type these
>>     frequencies in spss and tell it that these are frequencies of
>>     categories, and not individual cases.
>>
>>     I tried to weigh cases by frequencies. But I was only  able to
>>     define one variable as frequency variable. The value I got after
>>     requesting chi-square test was 306.4. According to calculation
>>     done by hand the answer is not correct. I tried to find the
>>     answer of my problem on the web. But most of the tutorials have
>>     used pictures to explain data entry. Since I am blind, I was not
>>     able to solve my problem from those tutorials.
>>
>
>     I am not really sure I understand you... What you need is a
>     frequency variable for _each_ combination of awareness and
>     conservation behavior. Since awareness has 3 levels (highly,
>     moderately and none) and conservation has 3 levels too (high,
>     moderate and none), then you need 9 cells in your dataset, with
>     only one frequency variable. You weight by it and that is it. I
>     see you have an overall sample size of 150 (102+36+12, or,
>     90+39+21), but you only give marginal frequencies (totals by rows
>     or columns) and you need to weight by the individual  frequencies
>     (that add up to 150):
>
>     Awareness Conservation  Frequency
>     High         High                   N1
>     High         Moderate           N2
>     High         None                 N3
>     Moderate   High                 N4
>     Moderate   Moderate         N5
>     Moderate   None                N6
>     None         High                  N7
>     None         Moderate          N8
>     None         None                 N9
>
>     HTH,
>     Marta GG
>
>     --
>     For miscellaneous SPSS related statistical stuff, visit:
>     http://gjyp.nl/marta/
>
>     ===================== 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
>

--
For miscellaneous SPSS related statistical stuff, visit:
http://gjyp.nl/marta/

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