The study can be divided into three parts.
1. Survey study, correlation between frequency of play or training and MDORS sub scores.
2. Treatment study. Three groups (more play, more training, and control) and analysis of change in MDORS sub score.
3. Video analysis of dog-owner play and basic comand before and after treatment period.
1. Survey
2940 owners, off which 187 were men, completed the survey and met the requirements with only one dog/survey and at least 12 months ol
The survey consisted of two parts. The first part included questions about background information about both owner and dog as well as questions about the owner´s interaction with the dog in regards to play and training. Questions such as what type of play, how often they played, what they used during play etc were asked and similar questions were asked regarding their training interactions.
The second part of the survey included the Swedish version of the Monash Dog Owner Relationship Scale (MDORS) survey. The 28 questions of MDORS were presented on a 5-point Likert scale and aimed to measure three subscales of the dog-owner relationship: Emotional closeness, Dog-owner interaction, where high scores indicate a more positive relationship, and Perceived cost where higher scores indicate a negative view of the relationship.
A GLM was used to see wether frequency of play, frequency of training, sex of dog and owner or age of dog and owner was correlated to the different MDORS sub scores.
The last question of the survey was if the owner would like to take part in a longer study.
2. Treatment study
Participants for the experimental study were pseudo randomly divided into three groups, Play, Training and Control. The groups were balanced for age and sex of the owners, as well as age, sex and breeds of the dogs.
The experiment ran for four weeks where the Play and Training group were given weekly instructions on how to play and train with their dog which should be done 5 minutes extra each day on top of their regular training or play. All groups, including the Control group, were asked to fill in a weekly log of their interactions to motivate them to stay in the study and to make sure that they followed the instructions.
After the four week treatment period the owners answered the survey again but this time there were also questions included about the owners own experience of the study.
In order to test the effect of play and training treatments, only owners who reported that they had followed the treatment instructions for at least 8 days and had performed said interaction more than before the study were included in the analysis. This resulted in 137 owners in the Play group and 110 in the Training group. To ensure a true control group, only owners that reported that they had maintained the same level of play and training as before the study were included, resulting in 161 owners in this group.
A GLM was used to see wether group belonging, sex of dog or age of dog and owner was correlated to the change in MDORS sub scores.
For the question about the owners own experience chisquare and fisher exakt tests were used.
3. Video sampling
One hundred participants from each group were pseudo randomly selected and asked to video record a play interaction ending with a simple command, both before and after the treatment period. The participants were instructed to film themselves first playing with their dog for 30 seconds, then toss some treats on the ground and then ask the dog to perform a known trick (such as sit or lay down). The instructions also included to send their video recording through wetransfer for later behavioural analysis according to predetermined ethogram (Table 5).
49 owners provided useful videos before the treatment period started and before they knew which group they would belong in. After the treatment period 10 owners in the Play group, 6 in the training group and 6 in the control group sent videos, of which 9, 2 and 6 were matching owners.
Table. Ethogram for the video recordings.
| Behavior | Description | |
| Dog vocalisations | The dog barks, whines, howls or growls (during tug of war) | |
| Physical touch | The dog and owner touches for at least three seconds | |
| Human vocalisations | The owner laughs, talks to the dog, makes animated voices or noises | |
| Same level | The owner gets down to the dog’s level by sitting, kneeling or laying down | |
| Seeks out owner during play. | The dog stays close to owner and returns to the owner directly after obtaining the toy | |
| Type of play | What type of play dominates the session | |
| Repetition of cues | The number of time that the owner repeats a cue before the dog responds appropriate. | |