Survey results should be taken with a pinch of caution

livemint.com:

Carrying out any survey is a challenge not just because its sampling process must be right, but there is often an ‘information and understanding’ asymmetry between surveyors and respondents. Asking a man on the street what he thinks of the economy, for example, is an exercise fraught with trouble since only a tiny proportion of people would have studied economics and perceptions can vary widely. If one were asked the quantity of a certain product (say, spinach) consumed in the last 7 or 30 days, can one really remember how much? This is a problem with such consumer surveys.

Most responses are broad approximations. Also, there may be a tendency to overstate or understate consumption. Understating it for a government survey could entitle one to future benefits, while overstating it helps one look good.

Interestingly, this problem of perception is not just with individuals, but also companies, even though corporate respondents are better informed of business conditions. The purchasing managers’ index (PMI) is a popular leading indicator of industrial or service-sector activity. But when this is compared with the index of industrial production (IIP), we find a weak link.

At times, when the PMI presents a bright picture, the IIP indicates stagnation. This is not surprising because the PMI surveys 400 respondents and makes a comparison over the previous month on five parameters, with responses based not on specific numbers, but on whether there has been improvement or not. The IIP, on the other hand, looks at production numbers and growth over the previous year for a more comprehensive set of goods.