This paper examined the performance of dingoes (Canis dingo) on the rope-pulling task, previously used by Miklo´si et al. (Curr Biol 13:763–766, 2003) to highlight a key distinction in the problem-solving behaviour of wolves compared to dogs when in the company of humans. That is, when dogs were confronted with an unsolvable task, following a solvable version of the task they looked back or gazed at the human, whereas, wolves did not. We replicated the rope-pulling task using 12 sanctuary-housed dingoes and used the Miklo´si et al. (Curr Biol 13:763–766, 2003) definition of looking back behaviour to analyse the data. However, at least three different types of look backs were observed in our study. We, then developed a more accurate operational definition of looking back behaviour that was task specific and reanalysed the data. We found that the operational definition employed greatly influences the results, with vague definitions potentially overestimating the prevalence of looking back behaviour. Thus, caution must be taken when interpreting the results of studies utilising looking back as behaviour linked to assistance seeking during problem solving. We present a more stringent definition and make suggestions for future research.