For future studies I would recommend performing the tests with the dogs in a standardized test arena with the experimenter present to ensure consistency in the tests, record all stress behaviours, and control for distractions. While this may result in less data being collected overall, the data may lead to a more accurate reflection of the actual cognitive capacities of rescue dogs. If the results of the tests done with the dogs in person are consistent with the current study, confirming its accuracy, then future studies could use citizen science methods for larger sample sizes. 

Longer delay intervals for the memory tests would be a more robust test of the dogs’ actual working memory capacity. Additional questionnaires about the background of dogs that provide information on resource guarding behaviour and how rewarding the dogs’ find food vs social contact would be useful for clarifying the results of the Simple Memory Test. 

Genetic factors could also be analysed to determine if there is an underlying genetic cause of the differences in cognitive abilites between rescue and non-rescue dogs.