Correlation issues

[Samir Omerovic]

Hi to all,

 

I have a question I am not sure how to solve. I have asked bunch of respondents to evaluate certain
characteristics of a product on scale from 1 to 5 (1-I disagree completely, 5-I agree completely and
99-I do not know). Now I would like to find correlation between evaluations of certain products.
Pearson sounds just fine but I have an issue about 'I do not know' answers. One respondent could
evaluate one product with 1 to 5 value and some other product with 'I do not know' value. I am
thinking that I need some sort of adjustment for answers 'I do not know' but I am not sure about
what to do. Any ideas?

 

Regards

 

Samir

 

 

Samir Omerovic
Researcher
___________________________
GfK BH d.o.o.
Skenderija 44
71 000 Sarajevo
Bosnia and Herzegovina
Phone [phone]
Fax [phone]
[hidden email]
www.gfk.ba
Before printing this e-mail, please consider environment protection.

This e-mail (and any attachment/s) contains confidential and/or privileged information. If you are
not the intended recipient (or have
received this e-mail in error) please notify the sender immediately and destroy this e-mail.  Any
unauthorized copying, disclosure or
 distribution of the material in this e-mail is strictly forbidden.

Ova e-poruka sadrži povjerljive i/ili povlaštene podatke. Ako niste osoba naznačena kao primalac
molimo odmah upozoriti pošiljaoca
i uništiti primljenu e-poruku. Neautorizirano kopiranje, objavljivanje ili distribucija sadržaja iz
elektronske poruke striktno je zabranjeno.

 

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

[Matthias Spörrle]

Samir,

"I do not know" answers should not be included in your Pearson correlation
analysis concerning the relations between product characteristics (i.e. your
agreement questions). You should treat them as missing values (i.e., define
99 as missing value for your variables in the variable view). Run the
correlation analysis only for values 1-5.

[If you want you can make new dichotomous variables: For each variable,
recode every missing value into 0 and every agreement statement into 1. Then
run correlation for these new variables (Phi coefficient [Descriptives ->
crosstabs -> statistics]). The correlations obtained in this analysis will
tell you whether the tendency not to answer one question is associated with
the same tendency concerning another question (i.e., are there groups of
variables in your dataset sharing a [participant-specific] tendency to
induce "I do not know" answers?).]

HTH
Matthias







On Jan 29, 2008 11:25 AM, Samir Omerovic <[hidden email]> wrote:

> Hi to all,
>
>
>
> I have a question I am not sure how to solve. I have asked bunch of
> respondents to evaluate certain
> characteristics of a product on scale from 1 to 5 (1-I disagree
> completely, 5-I agree completely and
> 99-I do not know). Now I would like to find correlation between
> evaluations of certain products.
> Pearson sounds just fine but I have an issue about 'I do not know'
> answers. One respondent could
> evaluate one product with 1 to 5 value and some other product with 'I do
> not know' value. I am
> thinking that I need some sort of adjustment for answers 'I do not know'
> but I am not sure about
> what to do. Any ideas?
>
>
>
> Regards
>
>
>
> Samir
>
>
>
>
>
> Samir Omerovic
> Researcher
> ___________________________
> GfK BH d.o.o.
> Skenderija 44
> 71 000 Sarajevo
> Bosnia and Herzegovina
> Phone [phone]
> Fax [phone]
> [hidden email]
> www.gfk.ba
> Before printing this e-mail, please consider environment protection.
>
> This e-mail (and any attachment/s) contains confidential and/or privileged
> information. If you are
> not the intended recipient (or have
> received this e-mail in error) please notify the sender immediately and
> destroy this e-mail.  Any
> unauthorized copying, disclosure or
>  distribution of the material in this e-mail is strictly forbidden.
>
> Ova e-poruka sadrži povjerljive i/ili povlaštene podatke. Ako niste osoba
> naznačena kao primalac
> molimo odmah upozoriti pošiljaoca
> i uništiti primljenu e-poruku. Neautorizirano kopiranje, objavljivanje ili
> distribucija sadržaja iz
> elektronske poruke striktno je zabranjeno.
>
>
>
> 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
>