Principal component analyses of the training techniques
did not result in easily interpretable components. For
instance, in a three component solution, the item ‘‘ignore
the dog/time-out’’ (negative punishment) had a low
loading on the component consisting of the techniques
of positive punishment and the component including all
forms of rewards. The third component consisted of
‘‘comfort the dog by speaking/petting’’ and ‘‘distract the
dog with food/play’’ but ‘‘distract with food/play’’ also had
a low loading on the component including all forms of
rewards. Therefore, training techniques were grouped
according to learning theory (punishment and rewards) and
the intention the owner pursues when applying each type
of training technique (reward-based responses to unwanted
behaviour) (Table 1). The subscale scores were calculated
using the same principles as applied to the subscales
resulting from principal component analyses.
The questionnaire items relating to owner consistency,
again, were analysed using principal component analysis.
One component, named inconsistency, was isolated (Table
2). The item ‘‘Sometimes my dog is allowed to do
something usually forbidden’’ had a negative loading
and was reverse scored so that a high score indicates high
inconsistency.