We should be cautious about studies which measure the attachment of cat and dog owners to their pets. This is because questions used to assess attachment often include ones which are specific to interactions with dogs. This is partly why you will find that when news media report on the attachment of humans to their cats and dogs, dogs can come out on top.
A researcher and scientist, Zasloff, conducted a study, in 1994, on the various aspects of attachment to cats. She worked with another scientist whose name is Kidd. They surveyed 100 adult cat owners who had a strong attachment to their cats. In all they owned 267 cats. They stated that they preferred cats because of the ease of care combined with the affection and companionship provided.
They liked the behavior and appearance of cats and said that they felt comforted by them. They felt loved by their cat and had always had cats. They compared how people felt about their cats versus humans. They found that cats were ranked as being better than humans at (a) making people feel needed (b) providing companionship (c) something to care for and (d) something to watch. By contrast, cats were ranked as being worse than humans in making their cat owners feel safe or providing them with exercise.
However, in a subsequent study 2 years later they decided that the questions could have in-built bias. So for example, using the Comfort from Companion Animals Scale which measures attachment in terms of comfort received from a pet she found that there was no differences in attachment scores between cat and dog owners on 11 issues which pertain to the emotional nature of the relationship. However, if 2 items relating specifically to dogs were included in the study (they concerned exercise and safety), dog owners had significantly higher degree of attachment.
It is the age old problem. We should be cautious about the conclusion of studies especially those concerning emotional attachment issues.