I find your pre-print very interesting and I will try to answer your feed-back request. Your work is interesting in that it intends to scientifically investigate the efficiency and the effectiveness of a traditional tool (Brainstorming, in short BS) aimed to organizational development and creativity skill improvement (also introducing and testing a possible variation of the traditional use). An approach like this is becoming more and more necessary in order to continuously improve the quality levels of the organizations (research teams as well as enterprises) and of their products and results. What is more, such investigation deals with training and learning processes, that are crucial in scientific research.
The reason why I think I can give you some possibly useful feed-back is that I have been working for long time with BS and similar instruments, mainly in enterprise contexts. I have carefully read your pre-print and, on the basis of my experience and of my methodological reflections and studies, I have the impression that, in its present form, it could be improved with regards to two aspects.
First aspect: The TOOL – BS, as a creativity technique, was borne and developed in enterprise context (Osborne has been an ad officer and an entrepreneur). Eventually, it has become a sophisticated instrument with a lot of variants; the most acknowledged scholar, at least in the enterprise environment, is Edward de Bono, that I do not find in your references. de Bono’s thought about creativity should be known, in order to have a complete reference horizon. He not only gives precise definitions of BS and precise instructions about how to carry it out; through his theory of the “lateral thinking”, he also gives an explanation of WHY these techniques can work (his idea of the mind as a “passive” system is worth being taken into account). One of his most famous works is Lateral thinking – A textbook of creativity, Penguin Group 1990 (from first edition, 1970, published by Ward Lock Education; I have also heard about a new 2007 edition but I’m not certain of this). Other references can be easily retrieved on the Internet.
There is a couple of specific topics, linked to de Bono’s thought and to my trainer experience, that is worth deepening. The first topic regards the non-systematic approach to BS that, as I read in your pre-print, has gone spreading among the users. About this, I think I can reasonably state that, by one hand, I have no news of such a trend in enterprise environment; certainly, the application accuracy can vary very much from a trainer to another, but this is a diverse story. By the other hand, the bypassing of the prescribed rules can jeopardize the conclusions about the instrument; for example, a careless, inattentive application prevents from deciding if the unsatisfactory results depend on the instrument itself or on the way it is used. Anyway, as far as I know, BS cannot be equated with a totally free discussion in a group. One can introduce new variants, but the tool must be precisely structured, in order to be effectively tested. Main of all, there are two basic principles that should always be followed: separation between production of ideas and assessment / selection of them; aim to the maximum possible amount of produced ideas.
The second topic regards the productivity loss during BS that is claimed by several researches. I have not (yet) read the literature you referenced but what I’m reading about this in your pre-print suggests me possible technical or relational problems. For example, about the delay between generation and verbalization of ideas, in order to avoid forgetting them, a paper notebook and a pencil could be enough (this is a possible technical solution); in addition, we must consider that the problem of ideas that shyly appear and then vanish, does exist in individual BS, too. About the relational problems, described behaviours like own ideas’ self-suppression or the fear of negative evaluations, seem to be typical effects of an insufficient relationship quality inside the group. The same is for “free-riding” behaviour. All this opens a great question, that should be carefully taken into account in the designing of your planned full research: beyond the rules to govern the sessions, what are the pre-conditions required for effective BS? If, in the groups, the internal relationship system has a low quality level, it will be difficult to assess BS as a tool in itself. The same if the groups to be compared present very different levels in the quality of their internal relationships.
Second aspect: The FULL RESEARCH – From a methodological slant, you have chosen to approach the matter through an action case study. It is an effective approach, widely accepted in enterprise contexts, possibly less welcome in certain scientific research environments; exactly for this, the research design should be extremely rigorous and the data collection should be tightly controlled. About such aspects, I will synthesize here below my feed-backs.
1. The composition of the groups to be compared: 4 or 5 persons (that is the composition reported on your pre-print, if I am not wrong) are too few to lead to some meaningful conclusion. The standard amount of BS group members is about 12; my experience leads me to conclude that a group of less than 8 persons cannot provide an acceptable level of variability (remember that one fundamental principle of BS is to gather the maximum possible ideas’ amount, before selecting them).
2. More than two groups should be employed, if possible.
3. All the efforts should be done to make uniform all the group features not involved in the research: age, sex, background experience… of the members and, measuring it in a way to be studied, the quality of the relationships inside each group.
4. One crucial point, in works like this, is to quantify what is qualitative. Expressions like “the discussion was lively, interactive and energetic” should be replaced by indicators capable to express such concept through numbers. It is not easy but it is certainly possible, within certain limits.
5. Another crucial point is the detailed design of the two different experiences the couples of groups would undergo. In this case, too, expressions like “the other group followed the normal approach” should be replaced by structured descriptions of the session sets and scripts (or even true screenplays).
6. In the end, the question of the group results’ comparison must be tackled: how will we be capable to decide which of the two methods is better? A general indicator must be found (even through the combination of the other indicators already introduced) or a decision grid must be designed; the important is aiming to the maximum possible of objectivity, otherwise the work could be easily criticised.
Well, I hope my hints will be useful to you and I am ready to interact again with you, if necessary.
With my best regards and wishes of success