I'll hide these tears I hope you'll never see

Working in an international organization, one gets to meet lots of people from around the world. My boss Andrew Shepherd in Rome usually says participants of regional workshops held in Africa are very quick to report the problems they are facing. On the other hand, he views Asians as much more reserved about pronouncing a judgment on their situation. They are masters at poker face.

This tendency makes it particularly tricky for marketing experts to understand customers and consumers in Asia because they will be reluctant to voice too much criticism or even to give any opinion at all. Peter Batt from Curtin University of Technology says he never uses a ranking scale with an odd number of ranks when working in Asia because a great majority of respondents then tick the middle rank. It is very difficult to get any trend from a data base where "average" is the most common entry.

The same poker face problem arises when trying to analyse participants' feedback surveys about a training session, workshop or conference (organizing meetings is one significant part of my work here). If the components of the event are ranked from 1 to 4 on an inverted scale with 1 being "excellent", 2 "very good", 3 "good" and 4 "fair", you seldom get responses below 3.

Rian Beise-Zee from the Asian Institute of Technology goes as far as saying that any respondent ticking 2 or "very good" is already showing dissatisfaction! Rian contends that we have to interpret in detail the variation of responses in the very narrow range of 1 and 2. I'd like to illustrate this with the participants' feedback from a large regional meeting I organized in Asia recently.

The Regional agro-industries forum for Asia and the Pacific was held in Yangling, China in November 2009. At the closing plenary, we asked participants to give oral feedback on how they had liked the meeting. The comments we received on the technical programme were excellent; there was relatively more negative oral feedback on the networking and consultation side of the event.

Now let's take a look at the results from the anonymous feedback questionnaires that participants completed before leaving the event. Without going into complicated statistical methods, the average ranking for the elements of the technical programme on the inverted scale 1 to 4 where 1 is "excellent" and 2 is "very good" are:
A - 1.677
B - 1.921
C - 1.875
D - 1.711
E - 1.938
F - 2.015
G - 1.935
H - 1.656
I - 1.825
J - 1.762
K - 1.65

By the looks of it, participants are very happy with the technical programme with all averages except one being between "excellent" and "very good". The last element K is actually the overall quality of the programme, which gets the best average. Success! Hiphip! Hurray! :-))

But if we follow Rian's advice to look more into the relative variation of responses in the narrow range of the more positive rankings, one does get a more nuanced picture. I've plotted the average of each element as a star on a horizontal scale ranging from 2 to 1.
I think it is quite clear here that the forum participants did not have such a good impression of elements E, F and G of the technical programme. So perhaps we organizers need to ponder how we could improve these components of the forum. I'm still struggling to come with a good visual representation of these results, which will allow me to report back to the participants. I'd be grateful for any feedback on this matter.

Advice telling me to cool down and to stop being paranoid also welcome.

Crying in the rain
Originally by The Everly Brothers, but I like better the version by:
A-ha, East of the sun West of the moon, Warner Bros