Summary

Group brainstorming is what the authors believe to be effective for idea generation. It resolves the cognitive bottleneck of idea generation through social means. The collaboration between individuals broadens the knowledge base and allows for aggerated creative effort. Yet, the effectiveness of group brainstorming varies by individual differences. This is because empirical studies have shown that too much similarity inside a collaborating group makes idea generation worse. Therefore, it is important to ensure enough amount of diversities in terms of concepts in a group. The authors define two type of diversity: internal diversity (refers to background differences between group members that originate from long-term learning and socialization) and external diversity (refers to stimulation supplied by external agents). The authors examine how one of each of these diversities (cultural difference and picture stimuli) affect the productivity and diversity of ideas generated in group brainstorming. They find that picture support in general enhanced brainstorming outcomes, and that intercultural groups especially benefited from this external stimulation. The results suggest that technological design grounded in a systematic understanding of cultures can play a valuable role in converting cultural diversity into creative outcomes.

Reflection

First, I like how the authors implement the picture support. At first, I thought they just script the brainstorming ideas and choose images on their discretion. Then I will probably question about the validity of the congruence and stimulus pictures because: 1^st they could be biased; 2^nd they might work differently in the participant’s mind. To have a picture retriever and selector based on the user’s choice of words is smart. In this way, conceptually there should not be any bias towards the selection of pictures. On the other hand, since the authors did not report it, I am wondering how large is the language pool. And how large is the picture pool. Then the authors help answering the second question as well. The t-test suggests the effectiveness of the picture selector. Which is a little bit surprising for me because in my own experience dealing with TF-IDF, the accuracy was not as nearly good as theirs.

Another thing about the paper is the participants. As the authors documented, two of the three questions in their main tasks were “what if you have an extra thumb or extra eye?”. These two questions have already been used in other studies before. Were the participants presented/encountered with these questions before? I think this is important to say because if by any chance the participants knew about the questions, the data gathered would definitely be biased.

Question

When different cultural background is considered, what about cultural differences between the same nationalities? For example, participants from China might be somehow different than participants from HongKong or Taiwan.