Global Business Barometer

In this article from Econnomist, the author suggested that business people have become gloomier about future business conditions, according to a set of Econnomist/FT quarterly survey of over 1,500 executives.
As shown in the figure below, the balance of those who think global business conditions will improve over those who think it will worsen fell from +5.2 percentage points in the second quarter of this year, to +2.

Obviously the frame of the research included a list of executives, but we don't know how many people responded for each quarterly survey. Moreover, in order to have a meaningful quarterly survey result comparison, certain attributes of each sample need to be as consistent as possible, for example: 

1. Interviewee. Each interviewee as a person tends to have a subjective view on future business conditions. Some people can certainly be more opportunistic than those who are relatively more pessimistic. If there's a significant change on the group of interviewees between the quarterly survey, the results are also very likely to shift, reflecting the characteristic of the interviewees rather than the actually business condition.
2. Sampling approach. Particular survey modes put respondents into different frames of mind, referred to as a mental "script".[1] This can affect the results they give.

• Face-to-face surveys prompt a "guest" script. Respondents are more likely to treat face-to-face interviewers graciously and hospitably, leading them to be more agreeable and affecting their answers.

• Phone interviews prompt a "solicitor" or "telemarketer" script. Respondents may place less priority on telephone interviews, making them more likely to satisfice (answer questions with the least possible effort) in order to finish the interview sooner.

3. Interviewer. Survey methodologists have devoted much effort to determine the extent to which interviewee responses are affected by physical characteristics of the interviewer. Main interviewer traits that have been demonstrated to influence survey responses are race, [2] gender [3] and relative body weight (BMI) .[4]

By industry, the article indicated that retailers are the most bearish, with a balance of -23; while construction and car companies are most bullish by a similar degree.

However, we don't know the sample size in each industry, whether it can represent the population. Also maybe quote sampling is an effective way to conduct the survey, because we'll need respondents from different industries and different countries. For example, 5 executives from some Australian agriculture companies wouldn't be sufficient to represent the population, because of the differences in climate, economy, etc., which would be significant factors especially when conducting global research topics.

In summary, statistical surveys are designed to make a view based on statistical inferences about the population being studied, therefore, the results depend strongly on the sampling approach, survey methodologies, sample coverage, etc.

Economist Article:
http://www.economist.com/blogs/graphicdetail/2013/08/economistft-survey


References:
[1] Groves, Robert M. (1989). Survey Errors and Survey Costs, New York: Wiley-Interscience.
[2] Hill, M.E (2002). "Race of the interviewer and perception of skin color: Evidence from the multi-city study of urban inequality". American Sociological Review 67 (1): 99–108.
[3] Flores-Macias, F.; Lawson, C. (2008). "Effects of interviewer gender on survey responses: Findings from a household survey in Mexico". International Journal of Public Opinion Research 20 (1): 100–110. doi:10.1093/ijpor/edn007.
[4] Eisinga, R.; Te Grotenhuis, M.; Larsen, J.K.; Pelzer, B.; Van Strien, T. (2011). "BMI of interviewer effects". International Journal of Public Opinion Research 23 (4): 530–543. doi:10.1093/ijpor/edr026.