I am not a stranger to the perils of bad data and the analysis, conclusions and recommendations based on it. Last week and the next couple, our students will be presenting and defending their research proposals (one group) and research outcomes (another group), respectively. Many of these students conducted secondary data collection and/or depended on online surveys for their primary data needs including interviews for their respective topics. Much of the secondary data are from past studies including the MMUTIS Update and Capacity Enhancement Project (MUCEP), which was the most recent comprehensive transport planning study for what we basically refer to now as NCR Plus (MUCEP covers Mega Manila, which includes parts of Bataan, Pampanga, Batangas and Quezon provinces aside from Bulacan, Cavite, Laguna and Rizal).
I share the following quote from one who is in the know or has inside information about what went about during the data collection for a major project that sought to update the Metro Manila Urban Transportation Integration Study (MMUTIS):
“The MUCEP data can not be trusted. A major part of the survey was done by DoTr (via a contracted group) – not by JICA-supervised surveys. Its results revealed a large portion of walking per day, because the surveyor filled up the forms and/or disregarded sampling design as most car-owning HH were not available (working or declined to participate) during the survey.
However, citing it is for convenience (aura of credibility). Its (mis)use is another matter.”
The stories are not at all new and I have heard this from various sources including surveyors and survey supervisors themselves whom we also engage for our own data collection (it’s a small world after all – transportation practice in the Philippines). Whether these are factual or not, should be obvious from the data and whether it is consistent with past studies or presents an abrupt change in matters such as mode share and vehicle ownership.
Of course both the DOTr and JICA will deny there was any error in data collection at the time and the weights of their statements will definitely make these the more accepted even if there are reasonable doubts about the assumptions and the survey implementation. But infallibility claims aside, what if the assertion in the quote was correct? What are the implications to activities such as forecasting, policymaking or master planning? Are we not surprised or dumbfounded that despite what is being reported as lower vehicle ownership for Mega Manila, it seems that people do have the motor vehicles and are opting to use them as public transport reliability and safety perceptions are still at low points. Mode choice after all is not as simple as some people want to make it appear to be. And if the assumptions including vehicle ownership are off then any modeling or analysis will end up with erroneous results.