When we designed our study of understanding the reception of biological evolution amongst Muslim physicians and medical students, we used prayer frequency as one of the measured of religiosity (we also used self-designation and if the person has read the Qur'an or not). Now our pilot study was amongst Pakistanis, and the question of prayer frequency made sense and it worked well (we recovered the full range, from 5 times a day to never or once or twice in life). When we did the same study in Turkey, it still worked, but the self-designation of "religious" corresponded to a lower level than Pakistan. Such variations are expected across countries and all we have to do it is to normalize the difference (and perhaps use other surveys, such as Pew study on prayer frequency), and take note of the difference when doing cross-country analysis.
Ah - but then we did interviews in Malaysia - and this particularly category completely collapsed. Almost everyone we interviewed prayed at least 5 times a day. There was one exception. He called himself "not very religious", but he prayed 4 times a day! Okay - so it seemed a little amusing that this guy is calling himself not very religious. However, the broader lesson was that since everyone prays 5 times a day, the prayer frequency may not tell us much about religiosity (for example, if everyone is taking hijab, then the taking of hijab may not symbolize more piety).
Well, I have been finding similar results about prayer frequency here in Indonesia. However, I interviewed a very thoughtful young student today, who provided a clear explanation for this. He picked the category of "Religious but Not religiously observant", and defined himself not religious. However, he prayed 5 times a day. When I questioned about it, he said that he is only doing things that are "obligatory" in Islam. That doesn't make him religious. He doesn't do anything extra that would separate him from others who are doing things more or less as part of culture. I this is a fantastic explanation and now I find the answer of Malaysian exception completely reasonable. Very cool. And it is a good reminder to be extra careful in analyzing big cross-country surveys.
The interviews here are going a bit slow as english is a bigger problem that we initially thought. We thought that since English is taught in schools, and medical students encounter it in their medical studies. However, the ground reality is a bit different and that limits the choice of who we can interview.
Okay - tomorrow, I have to get up early to check out the fantastic Prambanan Temple. More on that later.