One type of observation, however, focuses on the physical traces left by the factor of interest. These traces are of two types – accretions left or erosion that has resulted – similar in character to the delta built up at the mouth of a river as compared to a canyon carved by the same river. To define the trading area of cities, researchers have found it useful to observe the formations where farm roads the main highways leading to a city. Where the corners of these intersections have been rounded off the most on the side toward a given city, it is safe to assume that the majority of the time farmers entering the main highway are turning toward that city and are, in fact, a part of the city’s trading area. At the point where the round off begins to be more in the opposite direction, the farmers belong to the trading area of another city. When flying over an area, this breaking point between two trading areas often be seen clearly.
Other accretion studies have involved the observation of liquor bottles in trash to estimate the liquor consumption in cities without liquor stores, and the determination of the best radio stations on which to advertise to car listeners by observing the stations at which radios were set in the cars brought to a car dealer for repair. Brand preference studies are often based on observation of brands on consumer pantry shelves, and relative store emphasis on different brands is noted by observing the size of the inventory displayed.
Erosion observations are less frequent in marketing research, but examples are a study of the relative popularity of different museum displays done by observing the relative wear on floor tiles around the exhibits, and a study of the relative readership of different sections of an encyclopedia by measuring the wear and tear on the pages.
Observation of the results of past actions will not bias the data if done on a one-time basis. Pantry audits can determine what products have been bought previously. If such audits are made regularly, however, as with some store audits, biased data may result. It is rumored, that manufacturers’ sales representatives attempt to find the stores which are audited and put special sales effort on these stores. In this way they hope to make it appear that they are building large stocks with retail store customers.
Observation of Records: Whenever researchers use data collected for another purpose, they are employing the observation method in a manner very similar in character to the observation of physical traces. In one sense, the records of previous activities for example, sales, inventories, newspaper accounts, population census, highway usage – are physical traces of previous periods. In recent years computerized warehouse inventory and movement data and retail store sales scanning data become major sources of information on brand shares.
If one wishes to find what people think or know, the logical procedure is to ask them. This has led marketing researchers to use the questionnaire technique for collecting data more than any other method.
It is not as easy as it might appear, however, to collect facts or opinions from people accurately. Unless the point of interest has been impressed on respondents’ minds very recently, they are apt to have trouble remembering it exactly. Another problem is the unwillingness of some people to answer questions from strangers. Different wordings of questions will often obtain different results yet there is no way of knowing what is the correct way to ask the question. When the survey method is properly used, these disadvantages can be minimized. The versatility of the questionnaire method, its speed, and its relatively low cost are important advantages relative to the observation method.
Interviews can be handled in various ways. The general purpose of the survey can be disclosed to respondents or it cam be disguised. Interviews can follow a formal list of questions that re asked as written, or interviews can be non-structured and proceed as the interviewer’s judgment dictates. The cross classification of these two characteristics establishes four classes of interview, each of which fits certain situations.