Statistical tests for SPSS - which is the most appropriate?

#1
I'm a little confused on what statistical test(s) I should use in SPSS, and hope that someone could enlighten me.

Research is looking into the relationship between breed size and how an owner manages certain behavioural problems.

A questionnaire has been used to collect behavioural data from 500 owners. 20 Closed questions were used.

Purebred dogs were only used in this study to measure breed size more effectively. A list of purebred dogs were listed in the questionnaire for the respondents to pick from. This list was took from the UK Kennel club (which places each breed into size categories: small, medium, large).

For analysis, I am placing breeds Into size categories (small, medium and large).

Number of responses per size category:
Small: 201
Medium: 133
Large: 166

There are 10 questions to in my survey asking the respondent if their dog has a behavioural problem such as urination and defection in the home as seen below in “example layout”. Following each of these questions, there are “If yes, how do you manage this?” questions. The respondent is able to select from a list of a minimum 5 to 7 choices. The respondent can select as many as they deem appropriate to their circumstances.

Example layout of question from questionnaire:
7. Does your dog urinate/defecate in the home?
Yes
No
7a. If yes, how do you manage this? (Closed question with several choices to pick from)

I hypothesise that behavioural problems in small breed dogs are tolerated more.
 
Last edited:
#3
How was bread size actually measured? Did owners respond with a weight or did they respond by choosing one of S/M/L?
Purebred dogs were only used in this study to measure breed size more effectively. A list of purebred dogs were listed in the questionnaire for the respondents to pick from. This list was took from the UK Kennel club (which places each breed into size categories: small, medium, large).

Once I received 500 respondents, I split respondents into three size categories, all dependent on the breed they selected.
 

Karabiner

TS Contributor
#4
Dog size is an ordinal scaled variable, therefore, for the yes/no responses you could
use Mann-Whitney U test in order to analyse whether those with “yes“ responses tend
to have smaller (or bigger, respectively) dogs than “no“ respondents.

With kind regards

Karabiner
 
#6
For each answer, there are 3 responses to be recorded: doesn‘t apply ( no problem behavior) / manage it this way / do not manage it this way. You could either use Kruskal-Wallis tests, or maybe U-tests (if “ doesn‘t apply“ category is left out). Or maybe you want to aggregate several kinds of management?

With kind regards

Karabiner