However the study was flawed in that it did not take account of the fact that interactional modifications resulted in a greater quantity of input than the premodified input. This discrepancy could be seen to explain the better performance of those learners who were able to negotiate. In 1989, Pica addressed this problem and carried out a similar study where both groups received the same amount of modified input. The results did not show a significant difference in the learners comprehension, but did suggest that negotiation of meaning could be most effective for learners with a low level of L2 competence.
Pica also extended the IH by illustrating the effect of social relationships between the participants in L2 interactions. She showed that interactions were more beneficial to learners if all participants were regarded as having equivalent social status. This theory was summarised by Ellis (1991) as follows; “A situation in which the conversational partners share a symmetrical role relationship affords more opportunities for interactional restructuring.” (pp. 8)
There is considerable evidence to support this part of the IH. Long and Sato (1983) showed that the social imbalance that results from teacher-dominated L2 classes, reduce the opportunity for learners to negotiate meaning. This research was furthered by Van Lier (1988) who proposed task based classroom activities designed to